Good Robot Postmortem #2: Gameplay

By Shamus Posted Tuesday Jul 19, 2016

Filed under: Good Robot 139 comments

My original design for the game was “2D Descent”. That was the core inspiration for the project: Bright colors, electronic music, shooting lasers at crazy robots in caves. But this design wasn’t working. I didn’t know how to fix it and I was worried I was too close to the project to be able to analyze the thing objectively.

Lots of people offered advice. I had a dozen different, completely contradictory suggestions on how the game might be fixed. Everyone was happy to tell me how to fix the game by making a completely different game, but if I couldn’t realize my own vision then I probably wasn’t qualified to build theirs either. My original vision didn’t work out, and I didn’t know where to go next.

One of the key problems was that I’d never really worked out a proper fail state for the game. I had the moment-to-moment stuff down, but the win / lose states were undefined. If you died you’d just respawn, BioShock-style. It’s not that I thought having a game with no lose state was a good idea, it’s that I’d just never settled on a particular system.

How should death work? Maybe like Borderlands, where you’re taxed 10% of your money / XP when you respawn? Maybe like Dark Souls, where you drop all of your money / XP when you die, but can recover it? Maybe like Pac-Man, where you have a fixed number of deaths before a game over? Maybe like a modern game, where death reloads the last checkpoint? I fiddled with all of these and a few other ideas, but none of them really jumped out as the “correct” solution.

I guess early in the project I sort of assumed the answers to these sorts of questions would become obvious as the game took shape. I expected mechanics to “click” into place, but that never happened. I had a dozen different directions the game could go in and I didn’t have a good sense of which way was best. I was paralyzed by choice. I realized I could spend the rest of my life prototyping stuff without ever moving any closer to my goal.

And so the project stalled.

Man, that recoil.
Man, that recoil.

Then later when I teamed up with Pyrodactyl, we returned to the question of “What does this game need to be fun?” As we began working together the game began moving in a very “roguelike” direction.

Lots of people have said, “I’m surprised you made a game with feature X, Shamus, since you’ve spoken out against it in the past.” But I hope it’s clear at this point that I wasn’t the only person who contributed to the final product. My original idea had failed to produce a viable game. Sure, I could have put my foot down and refused to let the team implement features that fell outside of my personal preferences. Here is how that conversation would go:

Shamus: No! We can't have escalating difficulty mixed with permadeath in my game!

Everyone else: Okay. So what's your suggestion?

Shamus I literally don't have one. That's why I joined up with you guys.

Everyone else: So what are we supposed to do?

Shamus: I don't know what I want but I'll know it when you suggest it. Just keep suggesting ideas for me to shoot down.

I’ve worked with people who behaved this way. It was not fun. More importantly, they didn’t tend to carry projects to victory. Part of being on a team is trusting that the other people know what they’re doing. Whatever faults people may find with our final design (lots of people have offered insightful critiques in the comments) it’s still better than the dead-end shooting gallery I’d put together on my own.

I’ll let Ross talk about how the gameplay took shape…


Note the lack of quicktime events. That was a deliberate design choice!
Note the lack of quicktime events. That was a deliberate design choice!


Coming into Good Robot as a game with its core systems basically completed was an interesting experience. After Unrest and Will Fight For Food it would be the third time starting my job on a game part-way through development, but among them only Good Robot had so much of the base game complete before I got my hands on it. This wasn't so much a project to finish putting together a video game as one to take a game apart and reassemble it, adding on as few new pieces as possible.

Firstly, as might surprise you, I played the game. I played the game a lot. Simple enough, but often underestimated – especially by me, and doubly so in the midst of development. My first thought, after noting and instantly enjoying the thumping, booping soundtrack, was that this game was solid. Levels generated in a reasonably diverse ribbon shape (as Arvind already elaborated on); bullets had punch to them and shooting felt good; power-ups were easily understood; the robot's movement and aiming were precise and challenging enough to feel rewarding. Picking up the Good Robot for the first time was a breeze and a joy. The only immediate lack I felt was variety, with only a single weapon to my name.

Variety is hard to call an easy fix, since the result is cracking down and creating more types of things, but it's certainly a simple one. More weapons! More robots! More levels! It sounds like a mistake in the making, but this wasn't actually wrong – more weapons made more styles of gameplay and gave us another variable to randomize, more enemies extended the shelf life of our shooting mechanics even further, and more levels made good on that promise. Sometimes bigger really is better.

But in spite of all this, a sort of listlessness still hovered over the game after a short bit of play. Even though we had more options than ever, the drive to use or discover them was slipping away once the player was familiar with the game's fundamentals.

So where was the interesting-ness going?

Rhythm. It's not a perfect analogy, but it's close. Every game has a series of core decisions that form a loop which makes up the moment-to-moment play. Survey the environment, see a demon, dodge its fireball, shoot, collect a powerup, repeat. Head towards un-eaten dots, reach branching path, evaluate threat of ghosts vs. potential score, repeat. Depending on the game, these loops can be long or short. Most have several types of loop strung together, or a core loop with wildly divergent branches. Put all of them together and you have a series of repeating actions forming a larger whole – a rhythm. This is where our problems stemmed from.

If you were to somehow graph ten minutes of gameplay in terms of the number, type, and weight of decisions made within that time, the earliest build of Good Robot I played would closely resemble a straight line. You moved horizontally for five screens, shooting the level's enemy type and collecting their score nuggets; as your score increased you bought simple upgrades (damage, speed, rate of fire, shields… and “energy”, but we don't talk about energy); soon enough you dropped down and looped back the other way for another level. The line ticks up once in a great while as a boss appears, but it's not that major a change – and only a brief interlude before the player descends another shaft, reverses direction, and enters the next level. As mad as it sounds, even seamless level transitions were working against us.

What we needed were moments of tension and uncertainty, and instances of downtime where the player could take a breather to juxtapose them with. Rather than a consistent, predictable simmer, our line needed its peaks and valleys. The same principle goes into just about every form of entertainment, from composing music to writing stories. It's amazing how far good rhythm can carry a game – I strongly suspect it's the secret sauce behind a game like Diablo II's success, and the mixed results of its imitators.

With this as a guidepost, it's easy to explain a lot of the changes Good Robot underwent before release: more lethal combat, meaningful death states, hand-designed levels and profiles for the types of robots that could appear in them, weapons as powerups, warranties, randomly-chosen shop types with randomized contents. The goal was to make it possible for a player to fluctuate between being on the ropes and relaxing safely behind their contingency measures over the course of a game; to let them to walk into a seriously bad situation now and again, or catch a lucky break. We engineered breather periods by breaking the game into distinct levels with shops at the start and end, and built bosses and enemies which played with the player's expectations and altered their tactics.

In so many words, we changed up the rhythm to keep it from growing stale or predictable, and this time the interesting-ness hung around for another go.

No Plan Survives The Meta

The part of the game with the shooting.
The part of the game with the shooting.

In what's certain to be the second time I lead an article with a piece of obvious-seeming insight, I offer you this: no game designer in mankind's colourful history has ever been entirely correct about how another human being will play his game.

I stressed the importance of rigorously playing one's own game before (and chided myself for not doing it enough – bad Ross!), and this is the main reason why. No matter how far ahead you think, no matter how predictive or canny your reasoning, the meta will outwit you. It is the gestalt of a hundred thousand minds, pulling in different directions – none of them quite where you intended – and it will eat your carefully-laid strategies for breakfast.

As it turns out, predicting humans is hard. Take a recent game of Pathfinder I ran: a five-man party was working its way through the hallowed crypt of their nation's royal mausoleum, leagues under the earth, having just pulled themselves from a tricky encounter with a legion of ghostly soldiers. They stopped to rest, and the session ended. My dark work began.

I built a long hallway full of corners and alcoves that emptied into a dark, maze-like chamber full of standing pillars. Like any sensible person, I reasoned they would see the superfluity of hiding places and enter with extreme caution, checking every recess… but two dozen fruitless searches later, their guard would drop enough to miss a skulking assassin pressed into the shadows. To extend this time and build out my world a little more, the spirits of two small children would draw the party's attention, and make them doubt the possibility of combat so close by. One of the pair would flee into the darkness at the sight of them, naturally leading the group in that direction. A body with an important item lay squirreled away in an out-of-the-way passage far to the south – something they would be asked to come back for in a future encounter. The stage was set.

As my group assembled for the night, I played out in my head the way this ruse would go down: after confirming nothing was hiding, the party would find the first of the two ghosts and stop to interact, leading into the second's retreat. The band would follow, passing enough of the pillars to get a sense of the place before stumbling upon the little ghost hiding behind one… and the assassin would strike from behind. This would lead them into a cautious advance, fighting through ambushes and uncovering bits and pieces of the room's story. Caution and wit would come to the fore – a stark contrast to the running bedlam of the previous combat encounter. I knew how people thought, you see, and I would use that to build for them the game of their lives.

The fighter – a towering half-demon juggernaut with a maul – entered my first passage, took a single look at the terrain, and proceeded to roll his Perception to look for danger. He repeated this roll every ten steps. The others followed in a tight formation. Peering playfully out from behind a hanging tapestry, the spectre of a small girl beckoned the man's attention. I silently grinned. And then my plans disintegrated.

“I ignore the ghost.”
“I do the same and follow him.”
“I talk to it.”
“I'll listen in.”

Game footage unrelated to current D&D anecdote.
Game footage unrelated to current D&D anecdote.

Three of my party remained behind – the other two barreled right into the second ghost immediately and led off in pursuit. Careful pursuit. The most deliberate pursuit Torvus had ever seen, featuring torches and night vision and dutiful Perception checks. The pair branched south a corridor later, and branched again. They missed my assassin; they missed the assassin after my assassin. Preternaturally they dodged every threat – every trap – for a solid five minutes, without realizing a thing.

Hundreds of feet from their eagerly-chatting comrades, they found the rare thrice-nested assassin. By rolling max dice on a Perception check. Which they were still doing. They killed him in a single round, never realizing he was an assassin, then walked smack dab into the quest-hook body before knowing why they needed it. They thought this session was even more fun than the last.

Putting aside my pedestrian GMing skills, this is something that happens to every game designer at every level, from multiplayer FPS titles to turn-based card games. You can see it every time Hearthstone prints an Alarm-O-Bot, or a tooltip in Overwatch advises you to shoot Mei's ice walls. Even in games with no multiplayer component like Good Robot, how one expects the game to be played does not resemble the way others will play it. It's only natural – we're not psychic; most of the time we're passing crafty at best. Our planning serves to improve the odds, not eliminate them.

Valve has had this worked out for years – they know that they don't know, so they playtest rigorously and early in development, making changes to their games in progress based on the behaviour they see forming rather than anticipating the reactions of the players in their minds. It shows in their games (now of yore?), that almost every mechanic does what it's intended to… because if it ever didn't, it was probably noticed early on and altered. Even then, their games see regular balance patches and variable tweaks to adapt to the developing meta.

Of course, not everyone can be Valve, and as my anecdote hopefully showed, not every break from the designer's intentions is a bad experience for the players. Whole new genres of game have come to be through some hapless designer's ruined plans, and others are swiftly patched without much issue. It's something that has to be considered on a case-by-case basis.

On the subject of which: Good Robot's meta maladies were enlightening to study. As Shamus has written about before, we considered homing missiles too strong before I started to dodge them laterally rather than break their line of sight. Before release none of us put much weight on the range of weapons as a balancing factor, and after realized that firing from the safest position was literally the most important aspect of the game at high levels, especially in the environment of a no-hit run that we had just constructed. Bouncing projectiles struck me as fun and skillful when I included them on a handful of weapons; they became vital survival tools to kill enemies without danger in the eyes of the playerbase.

Some of these revelations have led to changes in the rules – others have become canonized and remain core parts of the game. The majority of my work on Good Robot post-launch has been dedicated to exactly this: playing a strange game of chicken with the meta, a move at a time – forever trying to outwit the human race.


From The Archives:

139 thoughts on “Good Robot Postmortem #2: Gameplay

  1. Da Mage says:

    I feel the mechanics of Good Robot actually work against each other at times and force a certain playstyle. Death is really harsh, as dying without a warranty is an instant game-over, and buying warranties means falling behind on upgrades…which leads to more dying. It’s a death spiral where dying puts you behind, and getting behind makes you die more.

    In the end most players quickly figure out they need to go slow and shoot from off-screen, then even when they are better at the game, that strategy still works and they never try anything more risky…..since there is no benefit to rushing.

    How do you fix that? Damn that’s a hard question, and I don’t think there is an easy answer.

    1. Robyrt says:

      What if loot drops scaled to your distance from the target? A linear scale would be too aggravating, so just declare a three-point line where if you are more than X% of the screen away, you get Y% of the money.

      1. Jsor says:

        You could always lift a mechanic from a couple old bullet hell games/Gradius clones: dodging enemy projectiles very closely builds up some important resource you expend to do other things. This is sort of the cousin to Bayonetta’s “dodging at the last second gives you slowmo”.

        Bullet Hell games are a little different since they’re autoscrollers so you’re forced into contact with the bullets anyway, so you could make the Good Robot resource more integral, though not so integral you can’t play it a little safe if you need to. E.G. the resource could be some sort of resource that makes your upgrades work, so you don’t get bouncy bullets and such if you’re not building that resource, but even if you don’t have it you can still try to snipe.

    2. SomeGuyInABikini says:

      I actually speedrun the game most of the time as I find frenetic combat is more interesting/fun than cautious shooting from off-screen, which can certainly become boring, quickly, but it means I often get one-shot killed. The curious thing about Good Robot as far as a ‘bullet hell’ goes is the utter lack of invincibility frames. This means if there’s a tight pack of bullets you will be hit by literally every one. Most of these types of games give you half, one or even two seconds of invincibility after being hit and the lack of that means a single bad decision/unfortunate scenario can result in insta-death and game over, irrespective of your shield level. Perhaps the inclusion of short-term invincibility would let people be more adventurous and less shoot-off-screeny.

      Incidentally, I made a single-file mod that increases the number of bosses in the boss fights as simultaneously fighting multiple worms or Garbage Supervisors is all kinds of crazy fun, especially for people who like chaotic fights :)

      1. Confanity says:

        Thank you for inadvertently answering a question I had… I had noticed that some hits damaged me even when I had a hat on, and was annoyed that the hat’s “invulnerability” wasn’t performing as advertised. I’m guessing now that what I interpreted as a single shot was actually a spray of bullets, or particles from an explosion, or the like, so the first one knocked the hat off and then immediately twenty more hit me and chipped off a big chunk of health.

        So maybe you’re on to something there… a split-second of invulnerability after taking a hit wouldn’t just make aggressive play less costly, it would also make hats feel like less of a rip-off at higher levels.

    3. Confanity says:

      I feel like there are at least four responses to the “slow down and shoot from off-screen,” all operating on different levels.

      First, the money. I’m under the impression that it degrades over time, but even without that I find myself unconsciously jumping forward to collect money, only to run into the next bad robot. While the Cash Magnet obviates this, I’m guessing it was an intentional design decision that encourages aggressive play while you’re still at lower levels.

      Second, the score system. More killing more faster keeps your score multiplier up, and you get more points. There’s even a “score above X” Accomplishment, although since I got it through relatively conservative play, maybe they should add another one that’s higher?

      Third, the limited availability of guns. Sure, the long-range bouncy shots are amazing, but if the best weapon you can find is the trash compactor or a shotgun-type, then you’re going to need to change up your play style, for a while at least, which can help break you out of an “all-sniping all the time” mentality.

      Fourth, those [expletive deleted] worms in some of the higher levels. If you try to turtle up and cautiously snipe, they’ll hunt you down and swarm your start zone. In my experience the only escape, unless you’re *really* good at dodging (in which case it’s easy for you to break out of a sniper-type play style anyway), is to open up some space so you have ground to fall back to before they arrive.

      And finally, the play examples we see in the videos often feature high-skill, up-close robot killing. There’s bound to be some effect from that, in terms of people trying to emulate such a style of play.

      1. Ninety-Three says:

        First, the money. I'm under the impression that it degrades over time, but even without that I find myself unconsciously jumping forward to collect money, only to run into the next bad robot. While the Cash Magnet obviates this, I'm guessing it was an intentional design decision that encourages aggressive play while you're still at lower levels.

        Nope, money lasts forever. Or at least, for as long as I had the patience to test. In light of that, it seems weird that money drops exist at all, why not just send them straight to the player’s wallet (which you essentially do once the player gets Cash Magnet)?

        Is this a relic of some past version of the system with expiring money?

        1. Naota says:

          The earliest version I played actually had money/exp persist forever, but there was no magnet and picking it up was something you did in very close proximity to the loot. I never actually inquired why it worked this way, but I can see a few advantages:
          -Larger, greener plumes of the stuff easily convey how much an enemy is worth at a glance, where a persistent counter increasing does not.
          -Picking up money is more involving than simply receiving it invisibly, for good and ill (see Borderlands for both examples at once).
          -Moving towards the money to pick it up draws the player through the level, often putting them in range of more enemies to fight in ways they didn’t necessarily plan.

          1. Ninety-Three says:

            I was thrown by the system because it felt like a case of cargo cult design missing the point: most of these topdown shooters use cash pickups as a way to punish exactly the careful sniping playstyle Good Robot suffers from: if the pickups expire, you have to get close to your kill fairly soon after killing it. As it plays now, picking up cash is something to prioritize after offscreen-sniping everything in the area.

            I haven’t played Borderlands (well, the first one for a couple hours, ages ago), but if you were going to axe the cash pickups, you obviously wouldn’t want to make them completely invisible, maybe have exploding robots put up a green floating “$100!” like they put up damage numbers, only more noticeable.

          2. Shamus says:

            Small correction: At that point in the project we had powerups. These were random drops from bosses. Twin fire, bouncy bullets, wall slide, etc. One of these was cash (then XP) magnet.

            Most of these were made obsolete when we added the weapon drops, and the rest were turned into vending machine upgrades.

      2. I’m sure I’ve seen money go *poof* in early zones before I’ve got the Cash Magnet. Was I imagining that? o_0; (It does take a while, mind.)

    4. Matt Downie says:

      There are lots of things you could theoretically do to encourage risk-taking. (I’ve heard it said the new Doom handles this well.)
      For example: Scoring/money incentives. Limited ammo. Giving all guns limited range. Enemies who gather together and then hunt you down if you take too long to find them.

      1. Syal says:

        Tie bullet range to vision maybe, so you can’t kill things offscreen. Bounce weapons bounce off visible walls but not ones outside range.

    5. Kylroy says:

      “It's a death spiral where dying puts you behind, and getting behind makes you die more.”

      I personally *hate* this style of gameplay. It makes me feel like I ought to restart the moment I take damage, since my mistake on the current challenge will make me weaker for all of the coming, presumably harder challenges. Tying XP and health regen to the same (limited) currency has the effect of making every bit of damage I take a permanent scar on my character advancement. No bueno.

    6. The Other Matt K says:

      In particular, the game has no real incentive to reward aggressive gameplay. The super-cautious / long-range / bouncing-projectile strategy is boring, but effective. Short-range aggressive gameplay seems very dangerous, but without a corresponding benefit to taking that approach.

      I still think the biggest issue is the sheer punishment of death, and the idea that forcing players to grind back through early stages, over and over again, is somehow a good thing. I’d be totally ok with a dangerous and difficult game if I felt I was able to make actual progress, rather than constantly going back to square one. Something like Super Meat Boy, where you get to die over and over again, but respawn quickly and right back near the action. Combine that with some sort of benefit (like a bonus drop if you complete the level in a certain timeframe), and suddenly there is a reason to pursue a strategy of aggression and speed.

  2. John says:

    I have to say that the comments following the last post in this series really took me by surprise, because I honestly don’t think Good Robot is all that difficult. I mean, I beat it, and I suck at these kinds of things. (And by these kinds of things, I mean games requiring reflexes. Or sometimes just games, period.) Still, I don’t want to tell anyone that their experience didn’t happen, doesn’t count, or was somehow wrong. I very emphatically do not want to be that guy. Instead, I will count myself fortunate that my default play-style–go slow, kill enemies at extreme range–happened to be a good fit for the game.

    I think I was sort of mentally prepared–conditioned, perhaps–for Good Robot by some of the games that I had played in the year or so prior. First, Descent. Descent taught me that, yes, there is a bad robot around that corner, of course there is. Why did I even ask? Second, Bastion. Bastion and Good Robot have pretty much identical keyboard & mouse controls. Use keyboard to move and dodge enemies. Use mouse to target and fire weapons. Bastion plays a little differently, since you can use a combat roll to get away from enemies or the shield to block many attacks, but they really are quite similar. (Seriously.) Finally, Crypt of the NecroDancer. Crypt of the NecroDancer is much, much harder than Good Robot, as it is–rhythm aspects aside–a fairly traditional rogue-like. (A much more tradtional rogue-like than Good Robot, for damned sure. I don’t really understand why Good Robot gets called a rogue-like, to tell the truth.) I died a lot playing Crypt of the NecroDancer. I got used to dying a lot playing Crypt of the NecroDancer.

    Finally, I have some questions for those who thought that the game was too difficult, as “difficult” is a none-too-precise word. Was the game difficult because what you were doing didn’t work? Was it difficult because nothing you did worked? Was it difficult because what did work wasn’t fun? Or something else? I am deeply curious.

    1. krellen says:

      The original prototype encouraged close-range, dodge-the-missiles play, of the sort shown in all the gifs and trailers. That was my expectation for the finished game as well, and “creep along and snipe” is an incredibly unsatisfying playstyle after the hours I poured into the prototype. The game is difficult because it presents a particular playstyle and does everything in its power to discourage that playstyle.

      (A proposed fix would be to bring back the regenerating shields – with or without the non-regenerating armour/health – to allow the player to take some hits without long-lasting penalty.)

      1. Matt Downie says:

        Regenerating shields tend to reward defensive play even more. Hanging back allows you to recover safely between hits.

        1. krellen says:

          Regenerating shields encourages a peaks-and-valley playstyle. Rush in, take some shield hits while skirmishing, then resting a bit for recharge.

          1. Matt Downie says:

            Only if rushing in provides a significant advantage over sniping.

            Would the bad robots have regenerating shields too? If so, that would make picking them off by firing wildly a lot more difficult.

            1. krellen says:

              Regenerating shields does no more to incentivise sniping than the game already does. However, they incentivise more aggressive playstyles (such as running in and dodging attacks) because a few mistakes are forgiven, instead of each mistake/hit being an unforgiving punishment.

              Note that, barring the last one, every gif on this page shows the Good Robot taking a single hit while doing its antics. A regenerating shield that absorbs a few hits (upgradable, probably) would help encourage this sort of play.

              1. Ninety-Three says:

                As long as we’re talking game design, it’s important to get our terms clear. Regenerating shields would not incentivize aggressive playstyle, it only permits aggression. There’s still no reason for an optimal player to be aggressive, other than “I’m not patient enough to snipe”. An example of a system that actually incentivized aggressive play would be “You get more cash for killing robots close to you” or “killstreaks earn a combo multiplier to cash” (not that I’m suggesting those should be implemented, they’re just easy examples).

                1. krellen says:

                  Killstreaks do earn a combo multiplier to score (which is not cash).

                  1. Matt Downie says:

                    If I cared about score, and was aware of there being combo multipliers in the game, that would provide a fairly good incentive (in combination with me being able to survive fighting at close range, or at all).

                  2. It took me an embarrassingly long time to even notice that. And when I finally did, my attitude was along the lines of, “Well, that’s a whole bucket of Don’t Care.” :/

                    ETA: Now if score had an impact on anything in game, such as restoring health or increased damage or even earning a warranty, then I would start caring.

                2. Loonyyy says:

                  Unless that becomes in some way a playstyle that competes with the existing one. It might earn a higher score, or be simply quicker.

                  I think krellen is exactly right about regenerating health or shields. You can take the risk of moving through fire if you can at least recover from a minor hit. When the effects are permanent and severe, you just can’t play that aggressively.

              2. Mephane says:

                Note that, barring the last one, every gif on this page shows the Good Robot taking a single hit while doing its antics. A regenerating shield that absorbs a few hits (upgradable, probably) would help encourage this sort of play.

                One can even see what I assume is a shield animation. But the final game had not form of even partially regenerating health, and instead healing up becomes

                a) More expensive each time you do it.
                b) (IIRC) Each repair has a price regardless how heavily hurt you are.
                b) It scales up between repairs not by how much health you heal, but just by how often you use the repair function.
                c) Thus, the game encourages to repair as little and late as possible, otherwise it double dips into your pockets by having you pay full price for repairing from 95%, and making the next repair more expensive, too.

                And in between all that, we are expected to buy warranties and weapons. Actually, buying weapons turned out to be a bad choice in almost any circumstance.

                1. Naota says:

                  Repairs are actually calculated from three different things: a static base price of 500, an additional amount based on how damaged the player is (from 0-100%), and another flat number multiplied by the number of repairs previously bought that’s added on to the sum of the previous two.

              3. Decius says:

                What about a shield that regenerated, tolerated a small number of hits before regenerating (perhaps regen requires being still and not firing), provides a period of invincibility after activation, and prevents attacking during that time?

                That would enable “jump in, dodge the bullets and use your knife-range gear, then get out quick if you get tagged” as a core loop, which is what the short videos promise.

                Shield-piercing projectiles would be an interesting complication to that.

      2. Daemian Lucifer says:

        Wouldnt a better solution be for enemies to drop quickly disappearing health pellets?That way,youd be encouraged to stay close to them in order to pick them up,and wouldnt be penalized for getting hit a few times while at that range.

        1. Syal says:

          Not quite the same. With regenerating shields you can tank two shots, but three requires using money on a health regenerater to get back up to full. With health drops, tanking three shots in a fight just means if you only tank one in the next fight you’re back to full. There’s also the question of how many enemies in the group will drop health; if you nuke a spawner, should they all drop health?

          1. Daemian Lucifer says:

            Regenerating shields give you no incentive to go forward.And they make all the encounters static,samey.In shooters that have regen you still have limited bullets,so there is a limited resource you have to hunt in every battle.But here you have infinite ammo,so you dont have to hunt even for that.Regenerating shields would just make the game easier and more static.

            With health drops,youd have a limited resource that needs constant refilling,and having it be refilled just by killing enemies would nudge you towards killing them.Doom has already shown that such a system works,and it works well.

    2. Jonathan says:

      For me, it was because I could be flying along safely, and then suddenly a factory would launch several death-wheel robots (with explosions or a cloud weapon, can’t really tell) at me that would knock off 75% of my shields instantly. No warning, no chance of dodging.

      And because I couldn’t reliably get a decent weapon… or tell what does what at the vending machine unless I’d seen it before (and most drops were weapons I didn’t want or already had).

      The repetitive grind through levels 0 & 1 only to die at whatever point the game decided to screw me over by firing high-speed robots at me or leaving me stuck with bad weapons got pretty boring too.

      The laser beam weapons look really neat in the animations and trailers. In reality, they’re pencil-thin, so they are very easy to miss with, and they fire too slowly to track onto targets and do damage.

      1. If factories are still glowing, then they have a chance to spawn stuff. As far as I can tell, each zone has X pre-existing enemies and the capacity to spawn up to Y more enemies through factories. You can reduce Y by shooting out factories before anything can spawn from them, but you risk cutting into your $ haul from that zone if you do, so it’s a tricky call. In fact, if you’re paying attention and you have decent weapons to hand, you can take things out as they’re spawning and all clustered up, which is pretty sweet when it works out.

        1. Jonathan says:

          Every factory lights up at first. The only way to be sure would be to fly for a bit, stop, fly for a bit, stop, etc. Jerky uber-cautious flying is boring.

          1. Trix2000 says:

            Or shoot EVERY SINGLE ONE.

            1. I started off this way! Then I realised that I was losing out on much-needed $ and that I couldn’t afford to shoot them out. Not to mention that some spawn stuff before you actually have them in view.

  3. Ninety-Three says:

    I’m surprised to learn that the source of Good Robot’s balance problems is as simple as “We didn’t realize long-range was good.” In retrospect, I suppose it was the obvious explanation, that’s why the game has all those terrible shotguns in it.

    I figured out the value of range pretty much immediately, and of course I can’t know how many other people think like me, but it seems like this should’ve been uncovered by the dev team fairly easily. In Part 1 of the retrospective, Naota mentioned that the harsh death penalty was a fairly late balance change, I wonder if what happened is that the dev team got used to playing like berserkers with a merciful death system, then got set in their ways and didn’t reevaluate after the change.

    How many testers did you show the game to before release, and per the above speculation, how many of them had their first exposure be to the harsh deaths?

    1. Matt Downie says:

      When I bought the game, the first couple of times I tried it, it seemed way too difficult. Then I read someone here saying that the best way to play was to shoot at enemies while they were still off-screen, and it became way easier.

    2. Supah Ewok says:

      “In retrospect, I suppose it was the obvious explanation, that's why the game has all those terrible shotguns in it.”

      Don’t you be sassing my Blunderbuss.

      “Yea, though I fly through the tunnel of the shadow of death, I will fear no robot: for thou art with me; thy tightly grouped and piercing projectiles, they comfort me.

      1. Syal says:

        Of course the primary use of Blunderbuss is dodging with the ridiculous recoil.

        1. Supah Ewok says:

          Not to mention each shot is practically a screen for missiles!

        2. 4th Dimension says:

          True friend was minigun or something similar that allowed you to hide behind a constant stream of lead while the lead itslef propels you backward.

    3. Once you start looking for it, you can see there’s a lot of games that go to a surprising amount of effort to eliminate the “range” advantage, because it is large, and natural to any even remotely “normal” space (even a 2D plane like Good Robot).

      If you’re watching the current Spoiler Warning on Fallout 4, or have played any Fallout past 3, consider the absurdity of Josh casually charging a machine gun turrent from 60 feet away, while it is pointed at him and firing, and beating it up with what is basically a fancy hunk of wood. Would you even consider that choice in real life? I don’t care what armor you’re in, you’d still be worried, and would seek another option. In real life, the machine gun has an enormous range advantage, and by the time you can close enough to hit it with your range-zero weapon, you’re too likely to be dead.

      Turn-based JRPGs often eliminate the question by making the battles, despite how they may be animated, take place on a single point. Everybody is equally at zero distance from everybody else, so “gun” and “sword” are just two flavors of “things that deal damage” with little meaningful difference between them. SRPGs often weaken the range advantage by taking advantage of the turn-based nature to make it so that in your “move”, you automatically cover 50-100 meters in an indivisible time instant. It doesn’t eliminate the effects of range, but it greatly weakens it by making “covering fire” impossible, and isolating any “in between” states to a single moment of vulnerability instead of a long period of vulnerability. Compare with the XCOM approach, that does a reasonable job of preserving that concept while still being turn based.

      You also, across the board, often see a weakening of ranged weapons to the point of absurdity. In a realistic setting, guns ought to really hurt. In real life, guns past the musket stage uniformly defeated all melee weapons. By all rights, an assault rifle against a Super Mutant ought to not merely be “better” than hitting them with a stick, but so much better, so obviously better, that “hit them with a stick” doesn’t even cross your mind. In fantasy settings, observe in things like Skyrim just how much the ranged spells have been progressively nerfed to be little more than gentle fire massages, because you’re expected to be able to abuse the heck out of range to defeat anything, and handing you an effective ranged weapon is basically a “win” button in their world because the AI can’t handle it… and they can’t give the AI very many effective ranged weapons for the complementary reason! People don’t like it when AIs have “win” buttons.

      Range is huge. If you have a weapon that can hit your enemy while they can’t hit you, and you can preserve that state well enough, it’s almost an automatic win, especially in contexts that have no logistic issues. I’d also observe that a lot of the spread stuff in Good Robot fell prey to the natural inverse square law; what is a devastating, game-breaking attack at point-blank range becomes love taps very quickly as the range grows, especially for omni-directional attacks.

      1. Kylroy says:

        “You also, across the board, often see a weakening of ranged weapons to the point of absurdity.”

        Worst offender – Dead Rising. Pistol shots hurt less than punches. Fighting one of the (human) bosses, I’d do as much damage with 10 sniper rounds as one solid swing with an axe.

        1. MichaelGC says:

          “Heh, Shamus has written about that before,” I thought, but it took me a short while to find it, as the article I was thinking of was actually about Dead Island. So, either it was Island you were thinking of, or both games have that problem – anyway, no matter! The article’s the main thing, o’ course, and that’s here:

      2. Fix to myself: In the Good Robot case, it would just be the inverse law. Inverse-square is for 3D spreading. But even the inverse law makes it very difficult to balance the weapon to have a good “adequate/correct” damage range without having large “useless” ranges and large “gamebreakingly-powerful” ranges.

        And as long as I’m adding a comment, let me also add that being able to see this is supremely frustrating. It’s crazy that in those JRPGs that take place on single points, the guns are often not only not powerful, but the weakest weapons around, as if the designers are compensating for the non-existent range issues. (In those JRPGs whose combat systems do have a concept of range, such as Tales of …, I at least understand the game balance issues.) Fallout and the Elder Scrolls are getting more and more frustrating for me, because I want to play a ranged character, and I can’t help but notice how I must pump a lot of very expensive ammo into mid-grade targets or pour all my mana, then recharge (potentially at cost of potions), then do it again, whereas melee characters walk right up to the target and down it in two hits. It’s backwards; I’m playing a high-input play style, incurring expenses the melee characters mostly aren’t (I still get hurt too), but I get pathetic output compared to what they get for nearly free.

        I remember Fallout 2 talking about making melee builds “viable”… if you ask me, they’ve gone waaay too far past that.

        I haven’t played it as much as I really should have, but what I’ve played of Wasteland 2 seems to get this right. If you can make it to melee range, you can really mess people up, which is true because in real life, ranged weapons generally have more issues with accuracy than they do in video games. But getting to that range is dangerous, and so is staying there if there’s multiple enemies.

      3. Echo Tango says:

        “Range is huge.”
        This shows up a lot in RTS games too. Even if you’ve got a pretty strong army of medium-range units, it’s still going to lose to a swarm of weak-but-long-range bullets from some kind of sniper unit. Sure, my weak units die in one hit, but if your bigger guys never get a shot off, my things’ lack of hitpoints doesn’t matter. :)

      4. evilmrhenry says:

        It gets worse once you’re dealing with multiple enemies. Consider a room with a dozen weak (killed in one hit) enemies in it. You can use a gun or a sword, but if there’s any sort of travel time, it’s going to take you longer to travel to each enemy and hit it with your sword, than to just point-n-click with a gun.

        Plus, range lets you control your position, not just your range. Assume the dozen weak enemies all have ranged weapons. If you use a sword, you have to attack from right next to an enemy, while a gun lets you be anywhere in the room, or even outside, which lets you avoid line-of-sight for at least some of the enemies.

  4. Sleepyfoo says:

    One thing I’ve noticed throughout the talk about the game so far is a desire to “maintain interest” or prolong the game (or the lifespan of game). There also seems to be some conflation of “interest” and “challenge”.

    As not a playtester, it sounds like the original prototype was a fun, disposable, diversion of a game. You flew through it, killing robots with whatever was at hand while kinda zoning out to the pretty lights and electronica music. I don’t know how long it took to get through the ribbon, but it kinda didn’t matter, as it sounds like the kind of game you play for 20 min max at a go. Despite the twitch reflexes and bullet dodging, it was a game to relax with. And there’s nothing wrong with that.

    However, it’s also sounded like the kind of game that people would keep at or leave on a whim. In an attempt to prevent that, the design team attempted to modify features to encourage people to stay with the game. Permadeath, increased danger from robots, Tighter economy restricting choices in a given playthrough, randomness to encourage different playstyles. This resulted in a very different game than the prototype, and a much more tense experience. It also takes a fair amount of time to get through a given playthrough (even with my starting tier 4 weapons it takes an hour or so to play through the whole game). And there’s nothing wrong with that either. But the addition of those features interacted in unexpected ways resulting in the boring Meta of “shoot from offscreen”.

    TL:DR, the attempts to hold interest seem to have unfortunately had an opposite effect based on comments.

    Unrelated to the above, Beams don’t shoot missiles out of the sky and so I never use a beam weapon as my primary.

    1. Fizban says:

      As much as I’d like to come back to it those are indeed the issues. I’m no bullet hell or twitch shooter and on paper it sounds like I should hate roguelikes, but apparently I do like Binding of Issac (and some others I’ve tried). This is only possible because the game includes plenty of ways to mitigate the bullet hell they started shoving in, games tend to turn over pretty quick, and if you do have a game that’s going long it’s because you have a really good chance of finishing things.

      Compare Good Robot, where every mechanic seems to run opposite of that. No health drops, and as someone pointed out above, no I-frames, the least forgiving damage system I think I’ve ever played. You get one hit’s worth of grace per level via the hat, but as the game goes on it becomes basically worthless against the waves of projectiles. Chewing through the first half takes as long as reaching the endgame of Issac, but there’s no chance of building up a pile of awesomeness that can roll to the finish: you’re always one salvo away from death.

      It also kinda bugs me hearing about how the game spent so long with regenerating shields and whatnot, before suddenly losing them at the end. So originally I could have beat the game, but because bullet hell players never get hit anyway they just stripped out the mechanic that let the rest of us play it? And then nerfing the shoot from offscreen tactic that I didn’t like but did at least work.

      Anyway. I could probably beat it if I kept trying, but it’s just too frustrating to want to come back. One bad dodge and it’s another hour grinding back to where you were, with no room for error, probably with worse weapon drops.

      1. Sleepyfoo says:

        Exactly, The equating of “Challenge” with “interest/longevity” for individual players resulted in a vastly reduced potential audience. The design decisions took it from a casual game with broad appeal and maybe no staying power to a “hardcore” game with niche appeal.

        In cynical market speak, it doesn’t matter how long a given person plays your game, so long as they enjoy the bit and spread the word. If they buy it an play 15 min or 30 hours, they still bought it.

        That said, I think Good Robot would have benefited from different game modes a Default mode, with what we have but much less punishing, a Hardcore mode with what we’ve got, and challenge modes with limited weapon choices vs special challenges or just specific enemy types or something.

        I have no idea how to mod that for myself, and I am aware it would be crazy extra work for everyone, so I have no idea how viable an Idea it would have been.

        1. Naota says:

          Well, I won’t make any promises but… today’s patch does add exactly this to the game. If you guys find it still too close to the normal settings in terms of difficulty, just leave some feedback and I’ll try to adjust things.

          What changed in the new difficulty:
          -Player health increased by 50% across the board.
          -Repair stacking costs reduced (50%); scaling by amount of damage taken increased (~50%). How much damage the player has taken now factors into repair costs more.
          -Warranty stacking costs slightly reduced (25%), added cost per upgrade slightly reduced (25%).
          -Price scaling on upgrades reduced by the following amounts:
          Shields: 1.25 -> 1.2
          Damage: 1.1 -> 1.0
          Speed: 1.2 -> 1.12
          Fire Rate: 1.2 -> 1.15
          Vision Radius: 0.6 -> 0.5

          General patch notes:
          -Cash rewards from minibosses increased 40-50%.
          -Increased Thermal Vent damage from 32×22 to 36×22.
          -Increased Ion Pulse damage from 50×32 to 55×32.
          -Increased Radial Heater damage from 20×24 to 24×24.

          1. Ninety-Three says:

            I find it interesting that you acknowledged the thermal vent as a useless pointless trap that genuinely sucks (your words), yet the buff it’s getting is a mere 12% damage boost. If you know the weapon is that bad, and you’re bothering to buff it, why such a small buff?

            1. Naota says:

              No reason other than caution, really. I prefer to work in gradual steps when dealing with multiplicative values and minimal testing, and a 12% buff still adds 88 flat damage to a volley. The real issue with the Thermal Vent is that projectile collisions are a fixed size, so fewer of the “puffs” hit the target than you’d naturally expect. On the other hand, large targets like bosses and worms are quite likely to eat the full +88 damage from a change like this, so I’m looking for a sweet spot between “not useless” and “instantly kills all bosses”.

              On the other, other hand, doing huge chunks of damage to bosses up is probably a worthy tradeoff for getting so close and clearing the level with a gun this dangerous. It’s probably best to test that quality to find a sweet spot rather than its performance against normal bots.

              1. Sleepyfoo says:

                I haven’t tried the beta new mode yet, but I did mod the drops.

                I changed every instance of lx_common to l6_common and changed the weapon drop rate to 60. This resulted in a slew of early upgrades, plenty of extra money for healing and warrenties, and much more fun with the weapons. I imagine that will hit a “wall” around level 4 however, when everything becomes expensive, as you’d’ve already spent about as much money as a normal playthrough to the end.

                Apart from that, with the Tier 4 weapons, I didn’t feel the need to be super cautious because every weapon was reasonably viable, and I could power through ambushes because I only needed to dodge a short period of time, rather than with the level appropriate weapons where I needed to kite the enemies.

                Even better, I tried out a wide variety of weapons in the early levels because none of them were crippling for whatever reason. They also felt flashier, which was also more fun than the early ones. I will say Tier 4 weapons have wildly disparate power levels, though.

                Another change was I didn’t feel the need to go for every scrap of cash because I had enough coming in to get at my feet under me. It felt like I could go for a money level later to make up for a mistake rather than a mistake leading to a death spiral.

                I’m not sure you’d want that for the original game, but maybe something for the new easier mode.

                1. Jonathan says:

                  Ok, where do I go to do this? I am playing through (beta version) and the only reason I didn’t finish New Jaipur with my starting-drop guns (argon blaster & bouncy crystal shards) is because I dropped $5k on the proton blaster 2 levels before the end of NJ.

                  What did the New Jaipur boss (Officer BuddyPal) drop? A proton blaster! I’d like my $5k back please.

                  Seriously, one drop every 6-10 levels is far too low.

                  Factories are still launching robots like missiles directly at the player from time to time.

                  The new “Zone Cleared” message is nice. However, killing a boss kills all the normal spawns in the boss zone as well, but without giving the player any of the $$. That’s $500 to $1500 in lost $$ for killing the boss first.

                  1. Naota says:

                    I could’ve sworn it’s just spawned minions that die along with the boss, and those are worth nothing for this reason (their worth is rolled into the boss’s drops). I’m not sure what causes factories to lob robots.

                    You’ll want to install this mod for increased drop rates (2x-3x depending on the point in the game). Keep in mind it will overwrite if the game is patched, so you may wish to keep the archive around to re-apply after any updates.

                    1. Jonathan says:


                  2. Sleepyfoo says:

                    In Robots.ini do a find/replace on l0_common with l6_common. Do that for l1-l5 as well.

                    In Drop.xml change the l6_common weapon drop chance from 150 or so to 60, and you’ll basically have my set-up.

                    Note that this only affects regular mobs to drop the good weapons, mini-bosses, bosses, and chests all remain level appropriate.

                    It’s worth noting that by default, weapon drops get less common as the game progresses. I assume that it’s based on the assumption of more potential enemies so lower the drop rate to keep things about the same, but it really doesn’t feel like it worked right.

                    I think it would have actually been better to leave the drop rate the same through-out, giving players more chances to find more interesting or more fun higher level weapons, rather than making it harder to find any at all.

          2. Cuthalion says:

            Ooo, interesting! On the one hand, I’ll feel lame for not having quite beat it (got to level 5 or 6) on the previous versions, but on the other, I’ll probably also feel quite good if I am able to beat it now. And it probably won’t take me as many tries.

            Updates remind me I need to get working on my Descent mod again. I doubt it’ll be as replayable as the base game, but it’s an interesting exercise. (I’ve done player primary weapons, but haven’t done secondaries yet or messed with shield/repair prices or enemy weapons.)

  5. Shoeboxjeddy says:

    “or a tooltip in Overwatch advises you to shoot Mei's ice walls.”

    Wait… why do we need to shoot those? They have a health bar that breaks if you do enough damage I’m guessing?

    1. Ninety-Three says:

      Yeah, but half the characters don’t even have enough damage output to knock down a wall before its timer expires, and even the ones who can would usually do better to go around it.

      It’s technically a thing you can do, but the balance works out such that you shouldn’t really. Like aggression in Good Robot.

      1. Shoeboxjeddy says:

        Okay good. I thought doing anything BESIDES shooting a wall would be a better use of time, but that comment surprised me.

        1. Ninety-Three says:

          I believe that was the point of the example, to illustrate:

          this is something that happens to every game designer at every level… how one expects the game to be played does not resemble the way others will play it.

          1. Naota says:

            Case in point: since I wrote this article several weeks ago, I have deliberately shot Mei’s ice wall exactly once in about 70 matches of Overwatch. It was while deployed as Bastion and buffed by an allied Mercy, in front of the final capture point.

            I broke through one of eight (?) segments of the wall about two seconds before the rest of the wall expired, at a cost of 45% of my ammo pool, allowing me to shoot Mei in the toe for about 40 damage before she stepped to the side, out of my view behind another full-health segment.

            1. Trix2000 says:

              It’s occasionally useful to do so if you can get people to focus fire, since each section has 500 health (a lot, but not insurmountable). It’s at least helpful when for the annoying Mei-wall on the spawn door. :)

  6. Matt Downie says:

    GMing tip for that kind of situation: The phrase, “You’re splitting the group?!?” in the right tone of voice, plus a raised eyebrow, usually discourages that type of thing.

    (Splitting the group tends to result either in extreme boredom for the half the group who doesn’t meet anything, or extreme death due to being outnumbered for the half that does.)

    1. Shoeboxjeddy says:

      Splitting the group is bad also because the DM can only run one adventure at once. So one part of the split, by necessity, has to just sit and not play. Then it switches and the other group has to wait. The more splits, the more people waiting at one time. On that basis, the DM should have the authority to refuse splitting up on a conditional basis, just to make the game run better.

      1. Kylroy says:

        Although if you have 2 GMs, it can be a blast to let the party actually split up.

  7. Mephane says:

    For what’s it worth, I loved firing from a safe distance and bouncing projectiles all over the place. As I am not good at this type of game, being able to do all of that felt extremely refreshing, for once a game that does not arbitrarily limit my slow, cautious, strike-from-safety playstyle, meticulously cleaning out every corner of a level before proceeding. I was certainly not playing as intended – instead of diving into the action, I avoided it, stood as far away from danger as possible; but I was having great fun with that.

    Then came the weapon range nerfs and I never beat the game.

    1. Ninety-Three says:

      I have deliberately held onto the pre-nerf game files, if you’d like I can put them up on Dropbox or something so that you can play the old way.

    2. Naota says:

      I also made a point to leave a few weapons in the game with long ranges after the range nerfs, because I too enjoy sniping individual targets from a great distance. It’s just that high-DPS automatics with long range are essentially the perfect gun, and overshadow the others to such an extent that they fill every role.

      It’s an interesting talking point, though: with all of the talk about caution and range being so dominant, technically I could reduce the range of every gun to (more or less) the limits of the screen. Would players want this, though, or would it just make the game even harder and more unforgiving?

      1. Supah Ewok says:

        Y’know, I’m gonna ask a question I’ve been suspecting for a while: are bad robots’ line of sight limited by screen resolution? When I was playing I’d notice that the distance to activate enemies seemed different when I had the game windowed, remaining at the edge of the screen rather than activating from true distance proximity. It actually really helped when entering larger rooms to play on lower resolutions windowed, so I wouldn’t wake everything up at once.

        1. Fizban says:

          I asked something similar on the discussion page and did some testing on my own: I know that if you change the window’s aspect ratio it does not affect the game area, thus if you square things you’ve actually got a margin on both sides that are active game area even though you can’t see them.

          Past that, I’m pretty sure robots don’t fire unless they’re onscreen, but if your window doesn’t match the game’s idea of screen they can fire from outside of your sight. So my guess would be that when in fullscreen you had cutoff portions at the edges that you didn’t have in windowed lower resolutions. I think that’s actually pretty common, I know there’s usually room at the edge of my screen where the mouse can disappear on most full screened programs.

        2. Naota says:

          To my knowledge, bad robots’ activation radius is set manually in a script on a per-robot basis, and they default to “on” when hit by a bullet, regardless of where they are in the level. With that said, it’s entirely possible that lower down in the code there’s an optimization effort in place that corrals this kind of interaction to only entities within a certain distance of the visible area. I know for certain that their sight radius value works as expected within the screen, since I’ve set them by hand, and a number of bots (eg. mines) won’t activate until you wander very close.

        3. Shamus says:

          Screen resolution shouldn’t impact view distance. It shouldn’t matter if you’re running at 1600×900 or 320×180, the mechanics should work the same, you should see the same distance, and the robots should behave the same.


          If you change the ASPECT, then you might get different behavior. If you switch to a 4:3 display mode, then the game will have to adjust for that. Robots are prevented from attacking if they aren’t on-screen, since it didn’t feel good to be shot by stuff you can’t see. A different aspect would mean they would have to get closer to you when attacking horizontally.

      2. Cuthalion says:

        I don’t think so. The problem isn’t that ranged is better; the problem is that for many players (like myself), close combat is suicide. It’s not that I want to be sniper all the time. That’s fun now and then, but I’d like to mix it up occasionally. It’s that if I’m not a sniper, I’m doomed after a couple of expensive, repair-requiring mistakes.

        Alternatively, lowering the escalation rate of repairs might help. While it would make the game easier in general (maybe more than you want to), it could permit more aggressive playstyles. If the boost in gameplay variety is bigger than the drop in difficulty, it could be worth it. Besides, repairs aren’t something you can get mid-fight anyway.

        Edit: I notice the comments are very critical today. I just wanted to add that I like the game, even though I’ve never quite beaten it.

      3. Ninety-Three says:

        I think that’s misdiagnosing the problem. It’s not that sniping is too good, it’s that everything about the game design pushes caution, and caution is boring. That’s a problem nestled deep in the heart of the game, and nerfing specific weapons can’t fix it. You can reduce how cautious we’re able to be by further nerfing range, but the best strategy will still be to hang back as far as possible, advance slowly, and back up while shooting at enemies.

        1. Syal says:

          Specific examples of things I think are bigger problems than range of player weapons (keep in mind I haven’t played since the Electron Repeater was still a victory hose):

          Enemies shoot in inaccurate cones around your location. If you try to dodge outside the cone they’ll just fire another cone. If there’s room to dodge around inside the cone, it takes ridiculous accuracy and they can still fire another cone. The only way to reliably avoid damage is to hide. Making enemies more accurate would work wonders I think. I can dodge around a line.

          Enemies fire instantly. As someone who takes two to five seconds to process new things, that’s a shot I’m never dodging.

          Enemies come in large groups. When they all fire clouds of bullets, rushing forward into a group just vaporizes you.

          Enemies perch around entrances. I don’t know if that’s a pathfinding problem or intentional, but it’s impossible to round a corner without tanking damage if you don’t have a splash weapon or a bounce weapon. Firing delays would help this, or maybe make the enemies stay away from the walls. Would give the spiders something extra as well, if they were the only ones who would ambush like that.

          1. Naota says:

            You could trace the cautious playstyle back to some of these issues for certain. The problem is, they’re also tied to complicated core systems that have been in place since before I even joined the project, and would not be easy to change.

            Enemies, for instance, either come out of machines until they hit a certain limit, or spawn at the level start and wait for you to show up. The game is procedural, so the only control we have over their placement is how many there are in the level, how many are waiting to spawn out of machines, and at what population the machines pause and stop spitting them out.

            The accuracy of enemy bullets is actually spot-on (unless they’re obviously shotgunners, bullet spammers, etc), but many alternate between aiming directly where you are and firing a shot where they anticipate you’ll be every X shots, which leads to them over- and under- shooting with varying amounts of lethality. Generally, the predictive aim makes a robot much more dangerous.

            Likewise, corner-lurking enemies do so because of their pathfinding algorithm. It works well for the majority of the game’s open spaces, but has hangups like these around tight spaces. Anything more accurate would be fractally more complex to write – essentially the same code RTS games use to manage their squads of AI units – and only exist to solve these edge cases.

        2. Mephane says:

          and caution is boring

          That’s where tastes differ. I usually prefer caution regardless whether a game enforce or requires it. I fought from long range in Good Robot right from the start simply because it is my natural play style.

      4. ThaneofFife says:

        I think it would just make the game harder and more unforgiving without other changes. You’d need something like regenerating shields or invulnerability frames after you get hit to balance out the loss.

        I also had a random idea re: ranged weapons. Have you thought about doing something like a scoped sniper rifle. In the context of the game, it could double the area of the visible screen, but it would replace a completely visible environment with a moveable, small-to-medium circle of visibility aimed by the player (maybe with a bulls-eye in the middle) while everything else would be blacked out–meaning that the played could be ambushed while using it the sniper rifle s/he wasn’t careful.

      5. WJS says:

        I think it’s important to distinguish between “sniping” and “spraying bullets off-screen indiscriminately”. Sniping an enemy is satisfying; killing enemies without even seeing them is not, and that’s pretty much all you can do when going up or down, since if you don’t then enemies will pop on screen and immediately shoot you from too close a range to easily dodge from.

  8. Geebs says:

    I played Good Robot again the other day, and I’ve figured out what’s wrong with it:

    Bouncing weapons. They’re boring. They encourage the player to spend all of their time shooting around corners, at enemies they can’t see. The game has been balanced so that shooting around corners is the best strategy.

    I bet if you took out all of the bouncing weapons and rebalanced, it would make the game a heck of a lot more fun.

    (Not to say it isn’t fun already, but in my last playthrough I got instakilled by one of those worm things on level 5, and it’ll be a good long time before I try again >:-E )

    1. Geebs says:

      As a corollary – games in which the player character is this fragile work better for me if they are both harder and shorter than Good Robot – like Devil Daggers, for example.

    2. Supah Ewok says:

      Bouncing weapons are actually terrible with the worms levels, which is where I think a lot of folks get their trouble from: they’ve been relying on one strategy the whole game, when they get to a level where the strategy is actually terrible, they assume the problem is the level, not the strategy.

      Not to say the worms aren’t hard, they are! But the key to getting through them is careful aggro management. You should only fight at most 2 at a time, maybe 3 if you’ve got the right weapons. But if you go in spraying bouncing bullets, they’re gonna travel beyond line of sight, hit half a dozen worms and wake them up, then they’ll surprise you through a wall and make mincemeat outta ya. And if you’ve been hanging back and letting the bouncing shots take care of your problems, odds are you haven’t developed the reflexes to deal with a group of enemies like that, so your screwed on respawn.

      Much better switching over to another weapon by the time of the boss fight immediately preceeding the worms. Something like the minigun. It’s been too long for me to remember the other good high tier weapons.

      1. Izicata says:

        Bouncing weapons are fine against worms, you just need to use them differently. Instead of firing ahead of you, you fill the room you’re in with bullets before you pull, so the worms walk into a bullet wall and die.

      2. I was also fine using my bouncy green balls of death on the first couple of worm zones I did. It was only when a whole bunch of them spawned on my doorstep and then swarmed me that I had a problem…

    3. Ninety-Three says:

      The real problem is that this game is fun when played aggressively, but the punishing repair/death system makes caution optimal, and caution is boring. Bouncing weapons are just the best way to be cautious, nerf them and power players will find the next best way to safely spray bullets at enemies.

      1. If bouncy weapons get nerfed, then pass me that rivet gun! :p

      2. Geebs says:

        That’s easily fixed: pull the camera out a bit further and make sure that no weapon can fire further than the edge of the screen. That gets rid of the second most boring tactic.

        I completely agree that the Good Robot should be made a bit more resilient, to allow more in-your-face play. The Flak Cannon should be one of the best weapons in any game, and not a terrible one that immediately gets the player killed.

        1. Sleepyfoo says:

          I use the Flak Cannon defensively in the final levels. Just fire in the general direction of the enemies and do a little damage but never worry about missiles again (which are much more prevalent in the last levels), My damage output comes primarily from Missile Rain or Cluster Missiles secondary. They are homing, explode to hurt stacks, and each missile does decent damage on it’s own.

          It’s a very cautious playstyle though. Also doesn’t really require Aiming, which allows all my focus to go to dodging.

          1. I don’t think I have ever even seen the Missile Rain weapon. :/

            1. Sleepyfoo says:

              Sadness : (

              I think Cluster Missile is the earliest version you can get (Tier 2 weapon?) and a middle version I can’t remember the name of right now that’s Tier 3. Missile Rain is the Tier 4 version, and I haven’t actually seen it that many times myself, but the Tier 3 version is slightly easier to find and works well enough.

  9. Tizzy says:

    The same way that the 20-th century saw the death of the author, I think that thanks to the internet we’re seeing the death of the game designer. I.e., the same way that the intent of a novelist or a film director is irrelevant to what popular culture will make of the work, for better or for worse, gamers will play the game however they damn want. Which was always the case, of course, but the multitude of YouTube let’s plays and twitch streams makes the fact plain for all to see.

    So the job of the designer becomes even more complicated, it’s to have a set of mechanics that works well together and promotes diverse play styles, at least if you intend to have the broadest possible appeal. You could cite a number of sandbox games as examples, but for my money, one of the most surprisingly successful current example is the Souls series. The games may be difficult, but they do a good job of offering a variety of mechanics and are fairly forgiving of the new player in the sense that you cannot really spec yourself into an impasse. They allow the players to experiment until they find a style that suits them. Which takes patience, of course. But the games feel pretty even handed to different play styles, and encourage exploration and self-expression more than you would expect at first. And it feels like the product of a very deliberate design process.

  10. Tim Keating says:

    That anecdote right there is why the “perception check” mechanic needs to die in a fire.

    1. ZekeCool says:

      That’s why I like Powered by the Apocalypse or Fate much more. They go as follows:
      Player A (Good Roll): I roll to read the situation. (rolls well) I know what’s going on perfectly and see the assassin/trap/whatever.
      Player B (Poor Roll): I roll to read the situation. (rolls poorly) Well there was no danger here but now there is an appropriate issue because of your poor roll.

    2. Fizban says:

      Is it a particular Pathfinder wording, or just being run wrong in general? In 3.5 you get passive spot/listen whenever something provokes it, which is supposed to be rolled by the DM when it happens, as are all spot/listen/search/whatever checks for secret information. And yet so many people can’t seem to wrap their head around the idea.

      You can roll an active spot/listen check if you think you missed something, and if 10 steps=30′ then that equates to one move action and one active spot each round. So two rolls instead of one. . . and the second is still rolled by the DM because duh.

      1. Naota says:

        I run my game by basically the same rules (Pathfinder puts spot/listen into one skill), but my players prefer to roll their own perception after each move in a situation like this. I never actually asked why it turned out this way, but looking back I think it adds an interesting element to the scenario (you need to actually examine things manually or risk blundering into a bad situation).

        That is, so long as nobody gets tired of the overhead. Requiring so many confirming dice rolls back to back is probably why the DM usually handles this. Since my game is run on Roll20, though, we get instant automated dice, making this setup a little less tiresome than it would be in person.

    3. WJS says:

      Aren’t you basically advocating the GM railroading the players into his ambush? I mean, what else is going to be the result of banning the players from being cautious?

  11. Izicata says:

    It all comes down to Good Robot having a roguelike’s heart but a twinstick shooter’s face. As an avid roguelike player myself, I had no problem with the game. Why does the game have a whole ton of useless short-range weapons, and bouncing explosives that can self-kill the player? Having useless weapons and the ability to screw yourself over is a cherished roguelike tradition. Playing very conservatively, as any mistake can potentially lead to a death spiral? That’s how roguelikes are meant to be played. A whole bunch of people thought they were getting a twinstick shooter that plays like a twinstick shooter, but got a twinstick shooter that plays like a roguelike instead, and found that they didn’t like roguelikes.

    1. Mistwraithe says:

      A fair point. But then why didn’t people who like roguelike games rush out to buy the game?

      I think there is an unusual and interesting commercial aspect which arises from Shamus’s blog. He was guaranteed a certain number of sales to blog readers (I don’t know the exact number, I would guess between 200 and 1000), plus a bias towards positive reviews, and a lack of public negativity from these same people. I think this is beyond dispute from the overwhelming positive Steam reviews (many of which mention the blog) which contrasts strongly with the largely negative feedback in these postmortem threads.

      If this community was made up of people who love roguelike games then I suspect this would have given Good Robot a significant initial boost from word of mouth (assuming it is of course a good roguelike game, I guess that is in dispute).

      However evidence suggests that the majority of the community is not that into roguelikes, which isn’t overly surprising given that Shamus himself has criticised permadeath in his writing and we’re on this blog because we find it interesting reading Shamus’s writing. A number of people have stated that they didn’t recommend Good Robot to friends because of that, and many of those positive Steam reviewers only played the game for 1-2 hours.

      Is the blog community factor big enough that it should have featured in the team’s planning? Obviously they needed to make a good game, but should they have been aiming for a good game which also suited the free ambassadors for the game? Or was it actually a good idea to go for something different and try to get two different types of community to both buy the game?

      1. Fizban says:

        I think part of this also depends on how you define “roguelike.” Many people, such as myself, probably define “roguelike” with more modern takes like Binding of Issac, and Issac does in fact give you plenty of room for error.

        I don’t think I actually have any familiarity with twinstick shooters, but my problem with Good Robot is that it’s actually less forgiving than the roguelikes I’ve played. Binding of Issac has tons of ways you can recover/bolster health which don’t affect your other upgrades. FTL has more of a one resource death spiral, but the most dodging required gives you about a full second of delay. Rogue Legacy has health drops, Risk of Rain has health drops, Ziggurat has health drops. Hand of Fate doesn’t really screw you until the later stages and even then you keep all the unlocks you managed before dying.

        So it’s not that the community didn’t like roguelikes: it’s that most of the community can’t handle twinstick/bullet hell shooters, and none of the compensating systems that put the -like in the name to ease the pain are present.

      2. Izicata says:

        Why didn't people who like roguelike games buy the game? Because most roguelikes are free. DCSS, Brogue, Cataclysm:DDA, ToME, Nethack, the millions of Angband forks, etc., are all free. People who like roguelikes are not likely to actually pay for one; that’s part of what keeps the games good and the community insular. There’s no market, so there’s no marketing and thus no mainstream appeal, and thus no diluting of the core game elements in order to appeal to a wider audience.

        “Roguelikes” that make money are usually not actually roguelikes. They may incorporate elements of roguelike gameplay, such as permadeath, but just because you include salsa in a dish doesn’t make it Mexican. Good Robot isn’t a roguelike either, but it has the same sort of punishing difficulty, and just like a roguelike you can win almost every time if you know what you’re doing. I’ve won all 3 of my last GR games and I don’t even have Elite Hat.

  12. WILL says:

    You talk about gameplay a lot but the end product is 2 viable weapons, with half the roster more likely to kill you and the other half the equivalent of wet farts (that flame bomb thing is possibly the most disappointing thing I’ve ever equipped). Experimenting is painful and dangerous, the game is best played investing every point into damage and firing blindly into rooms, abusing bounce mechanics. Defeating bosses yields very low rewards compared to just rushing a normal room.

    Multiple times I’d find myself going downwards and running into projectiles the enemy fired offscreen before I could see them. Traps and ambushes constantly happen with no warning. There isn’t a single iframe either – a spray pattern will cause you to take ALL those pellets as damage, getting knocked around will bounce you into more damage, etc… This is game design 101, all the way from NES – when the player gets hit, he gets a short moment to react.

    Upgrades are so generic too – damage and attack speed? Why would I ever put points into visiion when enemies’ HP constantly goes up and my upgrades are like, 5%? Not to mention the late game will spam small enemies but will have them take quite a lot of damage to kill. The idea behind spamming bots is that they die in one shot, they’re easy to clear, but very quickly you notice your weapon falls off even against the smallest foe and suddenly they take longer to kill.

    And then you add in robots that spawn other robots?

    Some weapons are just upgraded versions of other weapons. Why? What’s the point? Why not just tie that to the upgrade system?

  13. MelTorefas says:

    So, thinking about this myself, I mostly agree with what has been posted above. The game is simply too punishing to be played in the way that is (at least to me) the most fun: diving through crowds of robots and making them all explode in a cascade of machine violence. I tweaked the game files a bit to try and mitigate this, removing some of the things that frustrated me the most.

    Specifically: I removed the shared cost-scaling on upgrades, I re-tuned upgrade prices (actually increasing their base prices and individual scaling), dramatically lowered the price scaling and cost-per-hp on the shield repair, and removed all friendly fire splash from player weapons.

    This was enough to make the game more fun (and enough for me to actually beat it), but honestly the game still feels like it too strongly punishes the aggressive style. I really like the ideas suggested above of regenerating shields with non-regenerating HP, or invincibility frames triggered by being hit (I lean a bit towards the invincibility frames personally). I think one of those, combined with a cash multiplier of some sort for killing robots quickly, could really help incentivize (and make viable) the aggressive style of play.

    No matter what, though, I think I am still going to hate that damn worm level. >.> It is a *little* frustrating to suddenly reach a point where every behavior you learned earlier in the game is not only wrong but actually a liability.

    1. Tizzy says:

      Regenerating shields will incentivize caution, so I’d join my vote to yours for iframes.

    2. Naota says:

      I’ve got no issue with i-frames as a concept, though they do come with a few considerations the way the game is presently balanced. I think any weapon that shoots multiple low-damage bullets in a single shot would have to bypass the immunity for that volley (eg. shotguns), and the spam-happy bots would need some kind of consideration – either competitive damage on every bullet or simply that cyan spam shots don’t trigger i-frames. Similarly, a few other things shouldn’t necessarily trigger them either, like contact with enemy bots or hurt-walls.

      1. Ninety-Three says:

        Similarly, a few other things shouldn't necessarily trigger them either, like contact with enemy bots or hurt-walls.

        If you were to put in iframes, please please please figure out some way to balance them triggering off enemy bots. 100% of my deaths in this game happen when ten worms stack on top of each other and kill me instantly on contact. Through a hat, which just doesn’t seem like it’s working as intended (even though I understand why).

      2. Cuthalion says:

        Maybe only the hat grants i-frames? That makes it more of a “get out of jail free card” that only works once per level. (Or twice if you still had your hat from last time, but then you still have to survive long enough to go back to buy the replacement.) And then you don’t have to make exceptions, because normal shields continue to work normally; hats just make you invincible for a second as they nobly sacrifice themselves to save you.

        Funny thought: enemies target the hat instead of you until it is hit.

  14. Steve C says:

    Lots of people offered advice- like me. When you stalled, you actually took my advice. I suggested you get a group together of like minded indie developers. Form a team and finish the game that way.

  15. Philadelphus says:

    So after reading this I’d be curious to hear the dev team’s thoughts on the concept of early access. (The idea in general, not just Valve’s implementation of it on Steam, though that is part of it.)

    On the face of it it sounds like a promising solution to the problem of not having enough testing during development, since you’ll have a lot more eyes on the game to find these sorts of strategies. Since it’s earlier in development and the game isn’t (hopefully) as set in stone, you can then decide whether to embrace the meta and change the game in that direction or reject it and find some way to better incentivize the preferred playstyle/disincentivize the disliked playstyle. Rinse, release, and repeat for as many cycles as you can afford/until the game reaches your standards for being released, whatever those are.

    Obviously there can be downsides to the concept as well, such as coordinating releasing probably more-buggy code and a thousand voices all trying to get you to change the game to be more like their vision. From my perspective it seems like it has more positives than negatives, but then I’ve never released a game, hence why I’d like some opinions. :)

    1. Syal says:

      I can’t find Vlambeer’s talk on it, but they had at least one discussion about Nuclear Throne’s two years of Early Access and some of the snags to keep in mind with that process.

  16. Naota says:

    Following up on this article’s theme: I’ve seen a lot of comments that say half of the game’s weapons are useless/traps/pointless, but with a few exceptions (thermal vent!) it’s never fully agreed on which. I assume some of this is personal preference, some is a few weapons that genuinely suck (thermal vent!), and the rest is the result of range/safety/shooting offscreen being the most effective choice for the “meta”. It’s only natural to categorize them this way based on the way you play and the factors you value most. There is nothing strange or unexpected about this.

    …but you should know that I use all of the weapons you hate when I run into them in gameplay, and I’ve beaten the game with most of them by now. I know they’re not optimal, but I don’t value optimal that highly when considering my choices. Those shotguns everyone despises are my second-favourite gun, because I still play a pseudo-aggressive game that needs corner-peeking wherever possible. The gravity-based mortars and grenades are right up there on my “grab on sight” list, because I enjoy the challenge of aiming them (especially over cover). I feel like I should make a video to prove that they’re, if nothing else, viable for reaching the end of the game.

    Obviously I’m a weirdo, but I don’t find these weapons useless – just challenging in different ways. I’m the guy who plays post-nerf McCree or Zenyatta as ranged DPS, and has never touched Soldier 76. I played all three Witcher games as full-time alchemist bomber. If I can pick a weird thing that’s hard to do and make it work effectively, I will do it over an expected thing with guaranteed results – even if I can be demonstrably more effective leveraging the safer, more reliable option.

    1. Mistwraithe says:

      OK, but does this make you representative of the general gaming community, or quite an unusual player?

      From what you have written I would say the latter. You are down one end of the spectrum. Which is fine, but needs to recognised.

      If the game is easy then yes, I may well experiment with different guns for fun. If the game is hard though, and making a mistake means restarting, then I am going to use the optimal weapon I can find. I suspect that there are many, many more players like me, than there are with your philosophy.

      Which group of players were you guys designing the game for?

      1. Naota says:

        Not for me, certainly. It would be pretty much impossible to design a game of only sub-optimal choices, and my style of play only exists because more conventional and reliable options exist. I don’t pick weird things because they’re bad, mind – I pick them because they’re weird and I can make them good. The inherent challenge in taking an under-explored strategy and proving it viable appeals to me. If I try it and it doesn’t actually work, I’ll willfully drop it in favour of something else.

        To my mind, the “average” player tries out a bit of everything and sticks with the things that work best based off this initial impression. It’s worth noting that these things may not actually be the optimal strategy – just the one that is most appealing to a beginner. There are long-standing video game tropes like the Fire Emblem Jeigan archetype* that epitomize this distinction.

        The weapons are designed to range on a scale from “situational/powerful” to “universal/average”, with the random nature of which ones you encounter leading to odd choices and best-of-the-worst scenarios like Izicata suggested above. It doesn’t surprise me that the automatics with decent range and damage are the most popular gun, but I do think some of the situational ones get a bit more flak (pun intended?) than they’re due. If anything it’s a case where I don’t feel the oppression of the game’s difficulty nearly so much, and so it doesn’t constrain my choices to the same extent.

        *Jeigan archetypes are characters given at the start of the game which possess extreme power relative to your other units, but have minimal capacity to improve as they level up. Their strength is immediately tantalizing to beginners, but since they start out at a high level, they gain almost nothing for killing enemy troops. Overusing them is essentially starving your weaker units, which will later grow into much stronger characters than the Jeigan, for EXP.

  17. Naota says:

    Another thought: the common sentiment in these comments is that buying weapons is a waste of money. Evidently this is because weapon drops from enemies are reliable enough to be counted on, even though they’re random. I could easily change this and make weapons drop less than they presently do (and compensate with cheaper guns or more cash), but would that be a welcome change?

    I’d imagine the break point on this mechanic is pretty far down there – to make players really consider buying guns, the odds would need to be low enough that finding a new one off an enemy in time to keep up with the power curve each level could be considered an abnormality.

    1. I am tempted to advocate the opposite – that weapon drops should be increased. Why? It gives players much more chance to “try before they buy” – there are some weapons I’ve seen mentioned in comments here that I don’t think I’ve ever seen in either drops or shops, and that suggests that either there are too many weapons or that there’s not enough exposure for the number of weapons available – and bear in mind that I now play in a way that maximises the number of zones per level that I play. I’m pretty sure I’ve been through zones where nothing dropped at all – heck even mini-bosses can’t be relied upon to drop anything. Couple increased drops with differentiating “shop” vs “drop” weapons to make the shop ones have the edge, cf. our discussion last week, and players may have that much more incentive to get on and buy something that 1) they have tried and know they enjoy using and 2) is “better” in some way than the ones they might be dropped by chance.

    2. Geebs says:

      Definitely not (to nerfing weapon drops). The reason buying weapons is a waste of money is that they’re far too expensive and most of them are rubbish, so an upgrade that will definitely help trumps a weapon which will probably get switched back immediately every time.

      Also, would it be possible to have bosses drop weapons that aren’t significantly worse than the one the player already has? Boss weapon drops are a huge let-down.

      1. Naota says:

        Since the boss drops reference from a nice scripted list system, it’s theoretically possible, though what’s considered a “bad” weapon seems to vary quite a lot from person to person. In strict terms of power, we sort weapons into five tiers as it is, and distribute them as follows:

        Tier 0 – Exclusively the two starter guns. They exist solely to kill the first ten enemies, then be replaced with something better (which drops, guaranteed, from the first ten enemies).
        Tier 1 – Guns which drop in the prologue and level 1.
        Tier 2 – Guns which drop in levels 1, 2, and 3.
        Tier 3 – Guns which drop in levels 3, 4, and 5.
        Tier 4 – Guns which drop in levels 5 and 6.

        Bosses drop a weapon of the appropriate tier guaranteed. If there are two tiers of weapons dropping in the level (1, 3, and 5), the higher tier list is used.

        1. Sleepyfoo says:

          Level Appropriate guns cost almost as much as an upgrade. The gun will be “useless” in a level or at most 2 while the upgrade will be with you forever. That coupled with the guns on sale being as random as the drops (only 2 guns/2 secondaries, decent chance none are a kind you like) basically means it’s a waste of money.

          Contrarywise, drops feel rather sparse, such that I only really expect one or 2 a level. To be fair, that’s probably a function of dropping weapons I already have or weapons I actively avoid now (like saws and Thermal Vent).

          I currently have vague plans to mod the drops so that everyone drops Tier 4 weapons, as giving myself 2 buffed Tier 4 weapons as starters doesn’t feel like it broke the game*, but did make it more fun. Low level weapons are just boring visually and mechanically.

          *It did break the level 5 boss though, turns out explosions ignore his armor and hitting him with superheated flak cannon bits that explode really makes him a trivial fight.

    3. Jonathan says:

      Buying a weapon is risky. If I have to spend $6000 on a weapon, and it’s trash, I just lost out on a badly-needed repair or an upgrade. The store lacks any kind of data as to what kind of weapon I’m buying… if I haven’t seen something before, I don’t know if it’s a beam, laser dart, bouncy ball, or shotgun. I don’t know if it’s short-ranged or long ranged. I don’t know if it’s high-ROF or low-ROF.

      The store provides no information except a price, a name, and a logo.

      Unless I have seen a weapon drop in game and liked it, I don’t buy it at the store.

      Also – a check to stop weapons that I currently have from dropping would be great. Duplicate drops are just disappointing…and they seem to happen 10-20% of the time.

      1. Very much agree to all of this, and I think Naota is now very aware of how I feel regarding duplicate drops after the last column. :)

    4. ThaneofFife says:

      Cheaper weapons would be helpful, and decrease buyer’s remorse.

      1. Jonathan says:

        Making weapons free at the stores would not substantially change the gameplay in a negative fashion.

  18. RCN says:

    2 questions here:

    1. Have you ever gave consideration to making a demo? The game seems ripe for a demo, to let people test the waters with a couple of levels and a boss, while armed with a pick of the game’s most fun weapons but without stuff like upgrades to distract them, before putting the money down. I may be talking out of my ass, but it looks like a game where making a demo would be trivial.

    2. Have you ever reconsidered the upgrade system? Or, thought about how it lacks these peaks and valleys you talk about in gameplay? For me, it looks like the most neglected aspect of the game. As proof, damage upgrades were literally useless when the game was launched because there wasn’t enough granularity in the damage for the upgrades to make a difference until you’ve invested a handful of upgrades into it.

    What I think that the upgrade system mostly lacks is something beyond the linear scaling of power. Think about other roguelikes. In FTL, for instance, there’s a huge gap between having secure doors and having unsecure doors, it gives you much more room to deal with intruders. Likewise there’s a huge difference between having no sensors (you don’t even know what’s going on in your own ship), to having basic sensors (you know what’s up in your ship), to having advanced sensors (you now what is going on in the enemy ship) and maxed sensors (you know when the enemy ships are about to fire and where their energy is allocated). There’s also a subtle case for the medbay (you need at least level 2 medbay in order to heal a crew member while he is asfixiating). Not to mention that each advanced system is another chance to solve an encounter positively.

    For instance, you could add little systems to the game with the upgrades to make them more interesting than a flat bonus.

    Attack Power: The first upgrade into it could add penetration to any weapon that doesn’t have it, doing 1/10 of damage to the secondary target. Then, somewhere after (say, four upgrades in), you could add a charge (like Megaman) where you could hold the attack and then release for a double-damage shot (or shoot for double rate of fire for a second or two). And then, in the last upgrade, you could have a special charged attack that’s unique to every weapon. Or even a special combined attack that joined the primary and secondary weapon (I admit this last one would require a lot of work, but would certainly make the investment feel worthwhile).

    Rate of Fire: The first upgrade could give a minor damage bonus for every consecutive hit on targets (first hit is normal, consecutive hits gains a cumulative 3% bonus to damage, capping at 15% damage, only counts for primary targets). Then after 4 upgrades you could have the secondary weapon to be able to be used twice in a row, but with twice the cooldown if you do. And once maxed you could have an “full burst mode” where you double your fire rate for five seconds with a one minute cooldown.

    Movement: The first upgrade you could get a boost of two seconds for 20% extra speed that took 10 seconds to refuel. Then somewhere down the line a dodge function where pressing twice into a direction would make a quick dash into that direction (with a two second cooldown).

    Shields: The first point could make you ignore a small amount of damage received (like, the first point makes you ignore the first 5 points of damage from any source). Then with Four points you could have shield mitigation, where if you’re hit with more than one high-damaging attacks in a row the damage of the following attacks is halved. Then at max you could have regenerating shields (at 1% per second).

    View Radius: Frankly these already give some non-linear advantages. These are fine.

  19. Micamo says:

    I wanted to like Good Robot, but ultimately it fell victim to the same thing every Roguelike falls victim to for me: I get bored of replaying the trivial early sections of the game so I can get back to where my skill level will be challenged way, *way* before I can actually get through to the end of the game.

  20. Droid says:

    So you actually added a new difficulty option, so long after release. How long was it in planning? Did you do it as a response to postmortem comments?

    And while we’re totally not on that topic, Shamus and Rutskarn were already on Spoiler Warning together before you teamed up for Good Robot, right? Was that how everyone else on the team learned about it?

    1. Naota says:

      It wasn’t so much planned as a coincidence of factors that came together this week. Some kind of alternate difficulty or mode has been something I’ve been in support of since we overhauled the game’s economy, but there were more pertinent things going on at the time – polishing, promoting, and bug testing for the imminent release – that kept us from implementing it. While permadeath and resource systems will be a turn-off for certain players regardless, I think there’s a sizeable group that would enjoy them just fine if the systems were merely a bit more forgiving of mistakes – what I’d refer to later as the “simply too hard” feedback group.

      After release we got a lot of feedback both here and on the forums, but actually quite little of it regarding the game’s difficulty, so the other issues (upgrade system, gun damage, consistency in self-damaging weapons) took precedence. In the end I assumed people either enjoyed the game or had issues with the core design that were beyond the scope of a post-release patch, so the idea was shelved.

      Once the first postmortem post dropped, however, I saw enough comments to the tune of “I enjoy/would enjoy the game, but it’s simply too difficult” that it changed my mind on the subject. I approached Arvind with the idea to throw together an official easy-difficulty mod with my articles – something I could make and distribute without taking anyone else’s time. I’m usually the last one in the pipeline of a new game project, you see, so this is a window where I’ve got minimal work to do.

      Naturally, Arvind just cracked his knuckles and dropped the alternate game settings into the code as a new feature over the course of that day, and now Good Robot has an easy mode. I don’t expect it to solve any deep-seated hangups people might have with the core design, but for those who find the difficulty constraining more than anything else, I hope this update will turn things around.

      1. Droid says:

        Thanks for the explanation! You’re Ross, right? You mentioned being in a time window between two games, so did you already start work on your next project?

        1. Naota says:

          I am indeed. There’s a project I’ll be able to start work on coming up soon that should be very interesting in a lot of ways, but I think is going to be very solid and play perfectly to our strengths.

          As for my workload, development is little hard to categorize since it varies a lot by what type of game we’re making, but generally there are a few obvious stages I’ll get to below. My job is also weird: I do scripting and game design, but also record and edit our trailers, run interference now and again on sprites and visual effects, and have always wanted but never found the opportunity to write a thing. I have a diploma in 3D Art and a 3D modeling background that has never made me money or seen use in a commercial product, but noodling around in level editors and game scripts via notepad somehow has!

          And now for a lot of text.

          Pre-Production – We pitch ideas in the form of high-level design documents with a potential game’s concept, core systems, and other info. Basically anything goes and anyone can pitch anything, though only the best ideas survive. This can range from something like Late To The Party’s (“Unrest-style game but fancier and in not-Estonia, here are the plot beats”) to some very uncharted territory. We’ve had everything from a sidescroller about cute witches to a spaceship replicator technician sim to a game about playing a game.

          In choosing these there’s always a balance to be struck between what we’re already good at (narrative, textual content, branching dialogue) and what we want to become good at through experience (solid gameplay loops, game feel, graphics). Also, of course, there’s the factor of what we’re equipped to do from previous projects – another adventure game style RPG like Unrest is very easy to do, given we already have the tools for it, and even if not, any game with branching dialogue can still swipe its dialogue framework wholesale.

          Prototyping – Having decided on a concept, we typically put together a basic scenario from the game to see how it will work. This usually means the pitch becomes some more solid piece of content first – a game about exploring a solar system decides exactly how many planets there are, what can be found on them, and so on, until it’s a proper design document. For Unrest this stage was the first chapter in its original form. In the case of something like Good Robot that doesn’t work by “scenes”, we just create the simplest possible playable version of the game and start from that.

          Generally, artists get to work here finding a style and format that works for the game’s visuals/sound/etc. Programmers do an absurd percentage of their work in this stage just creating the framework to allow for the prototype to exist. After this initial push I tie it all together with some scripting, but until then I generally have little to do.

          Initial Development – The GDD is more or less locked down, and we begin work on the content of the game. If there’s a script, the writer starts working on it in full here. Artists start creating assets, and the programmers delve into more complex, less central features like the UI. If the game has systems complicated enough to require some game designing (like hand-designed levels, or enemy/gun stats), I might get to work on those here!

          Mid-Development – I finally get something big to do. There’s a script covering the first playable bits of the game. The artists have sprites, backgrounds, music, and sounds all lined up. The code is nailed down enough that I have tools to work with and (usually) won’t find the rug pulled out from under me. I go in and put it all together, turning raw images into game objects, placing items and laying out levels, and converting text from the writer into branching dialogue or menu blurbs. We go at it like this until the artists are basically all done and can start polishing whatever they think needs it, while I continue to assemble the rest until the game is finished at last.

          Late Development – The game is “done”, but likely there are elements that we’re still not quite happy with, so we end up playing it rigorously and polishing up the bits that stand out for obvious/simple improvement. Confusing things are untangled, some new features might be added (usually stuff like compasses, exit indicators, etc), and sometimes shaky ones are cut before release. Essentially, we pretend to be gamers playing our game for the first time and try to fix our own biggest complaints.

          Oh, and the PR loop begins, so I end up making countless trailer videos.

          Post-Development – Gamers actually play our game for the first time. We try to fix their biggest complaints quickly. Occasionally a shelved feature creeps back out of cold storage as a response to feedback and becomes real, like you saw with GR’s easy mode. At this point the artists are basically free to do something else and the programmers need only make small changes by request as we deal with bugs and balance tweaks. Only scripting is regularly required – hence my continued ability to tinker with Good Robot while the others work on the company’s next output.

          1. Droid says:

            So since you already know the outline of the game and its core concepts, that new project has probably already reached prototyping stage?

            I actually wondered about workload. Now that Good Robot is finished, you no longer have Shamus as a programmer, right? So you have a programmer, a writer, an artist and an animator with you in the team? Isn’t that a bit unusual?
            If the dev studios I’ve heard of so far are anything to go by, two programmers out of five seem to be the rule. Is it your games’ genres / core concepts that allow for more people to work on things other than the engine? Or is it just that I have no clue how a dev team actually works?

            1. Naota says:

              The type of game and re-usable asset stock matters a lot when it comes to allocating team members for indie games.

              A game like Good Robot that’s largely shooter mechanics needs a lot of programming in general, and others in the genre take it further (unique boss behaviours, for instance). Most assets are incredibly modular, and there’s not so much event-style scripting as there is merely roping content together and tweaking balance variables. Generally the more systemically complex the game (not just mechanics, but visuals, procedural gen, sound systems, whatever), the greater the need for programming.

              On the other hand, Unrest only needs a basic engine to move sprites around, handle objects, and display menus and dialogues, but it also needs miles of proofed writing down on paper, and even more scripting to debug and put it into the game. The overhead for programming is comparatively light, but you need as much original content as there is game to play.

              Then there’s re-use benefits. Imagine we were to make another Unrest game tomorrow: other than new engine features or improvements, we could make a full-length sequel (or another adventure-style RPG entirely) using Unrest’s framework with just an artist, a writer, and a scripter. With an episodic spinoff, we could almost get by without an artist at all.

              As for the new project, I doubt it would cause any waves to say we’ve decided on a concept and are in the early prototyping stages.

              1. Droid says:

                “Pyrodactyl Leaks Fact That They Got An Idea To Work With” makes for an unusual headline, true.

                Thanks again for your time and effort, your answers were insightful!

  21. Zekiel says:

    Really enjoying this portmortem series. Thank you for sharing it.

    And this:

    ‘Lots of people have said, “I'm surprised you made a game with feature X, Shamus, since you've spoken out against it in the past.” But I hope it's clear at this point that I wasn't the only person who contributed to the final product. My original idea had failed to produce a viable game. Sure, I could have put my foot down and refused to let the team implement features that fell outside of my personal preferences.’

    …is a very helpful and gracious explanation.

    1. anaphysik says:


  22. RCN says:

    Oh, have you guys ever discussed a cooperative mode? Or any kind of multiplayer mode? Because I’ve pitched the game to some friends, but since they aren’t familiar with Descent, my pitch was more along the lines of “a bullet hell like R-Type, but with exploration and only one life.”

    To witch they asked “And coop?”

    And when I said “Huh… no.”

    Their universal response was “Meh.”

    1. Naota says:

      The topic did come up a few times (I love co-op in games like these), but ultimately the game was already programmed around the idea of a single player at every level of abstraction. Regardless of netcode and the usual suspects for absent online play in indie games, it would’ve required even more work to go through the game altering everything to account for 2+ player entities that move, shoot, and die independently of one another.

Thanks for joining the discussion. Be nice, don't post angry, and enjoy yourself. This is supposed to be fun. Your email address will not be published. Required fields are marked*

You can enclose spoilers in <strike> tags like so:
<strike>Darth Vader is Luke's father!</strike>

You can make things italics like this:
Can you imagine having Darth Vader as your <i>father</i>?

You can make things bold like this:
I'm <b>very</b> glad Darth Vader isn't my father.

You can make links like this:
I'm reading about <a href="">Darth Vader</a> on Wikipedia!

You can quote someone like this:
Darth Vader said <blockquote>Luke, I am your father.</blockquote>

Leave a Reply

Your email address will not be published.