Project Good Robot 16: Level Up

By Shamus
on Sep 20, 2013
Filed under:
Good Robot

I cited some of my influences way back at the start of this series. I left one out. One other thing that’s really shaping my vision for this gameplay is this:


Mass Effect 2 (and to be fair, a LOT of other games) are an influence on how I want leveling to work. But not in an imitative sort of way. Like, I play those games and my goal is to do the exact opposite of everything they did. What do you call that? A negative influence? But that makes it sound like I’m just imitating the bad parts of the game. An inverse influence? Anti-influence?

I don’t know. Whatever. My pitch for the leveling in this game is to show you that screenshot above with the subtitle of, “NOT THIS.”

Here’s what I find unsatisfying about it:

  1. Increasing costs. To unlock level 4 face-shooting costs 1+2+3+4=10 points, which is enough points to take TWO levels of singing, baking, and schmoozing, and still have a point left over to put in hula-hooping. It’s basically a system that punishes any attempt to min-max. This system of point allocation tells the player, “Spread your points around. Don’t specialize. Make sure your character is bland. Like everyone else’s.”
  2. There aren’t very many levels. It’s a ladder with 4 rungs. You’re not going to get vertigo when you reach the top and look down.
  3. It’s possible to get everything. It’s not that hard to max your character out, which means that those choices you made about where to spend your skill points weren’t all that important. In the end, your character is just like everyone else’s.
  4. The levels don’t FEEL very significant. Well, it depends on the game. In Mass Effect 3, the tier 4 unlocks did feel like they did something.) But it’s rare to take a level and find yourself thinking, “Wow! That’s a lot nicer!”
  5. Foes level with you. Or not. You can’t tell. The game doesn’t say. But fights at the start of the game take a lot less time than the later fights, despite the player being max level and having great gear. Maybe you’re ten times more powerful than those Cerberus mooks at the start of the game. You can’t tell, because later on you’re fighting mooks with twelve times as many hitpoints. You never get the sensation of crushing a previously daunting foe.

Now, this doesn’t mean that all games that work like this are bad. I’m sure this kind of system is a lot easier to balance. But it’s not the kind of game I want to make.

I want a game where you can’t get all the levels, and you have to make choices about what powers mean the most to you. I want levels to be important. I want the player to feel like they’re getting more powerful, and I want them to be able to see the results of the power gains when they pull the trigger.

So leveling works something like this:

This is an “artist’s” mock-up, and should give you a pretty good idea of why they don’t let programmers design interfaces. At least I’m only using one font face. And it’s not Comic Sans.

So what I’m going for is a system where every attribute has ten levels, and the player should be able to “feel” all of them. That is, the player should not have to squint at the screen thinking, “I just put points into movement speed, but I can’t tell if it did anything.” By the end of the game, they should have about 66% to 75% of the available skill-ups.

Monsters will not have levels or any level-scaling. If you fight one of these green things in the first minutes of the game and it takes 6 shots to kill it:


It will have the exact same number of hitpoints when you’re effortlessly tearing through groups of them on your way to the final boss.

As the game goes on, formidable foes become common, common foes become mooks, and mooks become cannon fodder. New robot types will appear at the top of the hierarchy, but you’ll still see those lowly bots and always have them around as a subconscious baseline measurement of your relative power.

How you spend points has a huge impact on how the game feels. Let’s consider the first three abilities in isolation: If you dump everything into shot speed and energy you’ll be able to pump out huge volumes of ineffectual bullets. If you put everything into shot power and energy you’ll be pounding foes with uber-strong shots that blow them apart with explosive force. There will be a long delay between shots, so you kind of have the feeling of the DOOM 2 shotgun, where you have a huge burst of damage and then a nail-biting wait. If you put everything into shot power AND shot speed, then you’ll be able to pump out MASSIVE damage for about two seconds, and which point you will be completely out of energy and need to go hide and suck your thumb until you recover.

So maybe you’ll dogfight. Maybe you’ll snipe from a distance. Maybe you’ll play hit-and-run. Maybe you’ll put a lot into shields and be kind of tank-y. Maybe you’ll put points into movement speed and outrun missiles.

This game is as much about skill as it is leveling. I mean, that awesome level 8 movement speed doesn’t do you any good if you fight like a turret. And those devastating maxed-out lasers won’t help you if you can’t hit anything. This means that the “best” way to allocate your points probably varies from person to person, depending on what sorts of things they’re good at.

Hm. I’m pretty sure that text is supposed to be in FRONT of the characters.

I’ve done some playtesting with the Spoiler Warning cast, and so far the game seems to be doing what I want it to do. These leveling decisions are spaced out, so the player is probably going to be kind of reactionary on their first play-through. If they’re running out of murder juice, they’ll upgrade energy. If they’ feel like they’re not hitting hard enough, they’ll buy more shot power. If they’re dying a lot, they can buy more shields.

Now, there ARE drawbacks to a system like this. Without level scaling, there’s no safety valve for struggling players. Sure, grinding for a couple of extra levels would help, but it will never negate the challenge. In World of Warcraft or Borderlands, there’s eventually a point where foes can’t overcome the level gap and you can allow a crowd of them to gnaw on your ankles without suffering any harm. That doesn’t happen here.

It’s also possible to “break” a game like this. And I’m okay with that. I enjoy experimenting with leveling mechanics, looking for sweet spots and finding things that feel fun. I don’t feel the need to “harden” the game against Josh-type players. For some people, breaking the game is the game.

So that’s the idea. The gameplay won’t please everyone, but I’m pretty sure it will please someone.

And just so we have some screenshots to look at, I made this:


I turned the buzz-saw foes into these drilling… robo… worms? I guess? Anyway, it fixed that swastika problem I had and also made these guys really creepy.

Enjoyed this post? Please share!

A Hundred!202017We've got 157 comments. But one more probably won't hurt.

From the Archives:

  1. Brandon says:

    I think you’re on the right track with the leveling system. Leveling up should feel significant and the impact should be noticeable immediately. A lot of games now seem to tack on the RPG system last minute now (because every game needs levels) and the impact of levels is almost impossible to notice.

    I also like that you’ve decided to make it so people will not max out their robot, that makes choices so much more meaningful.

    Well, this is cool and exciting! I can’t wait to play it, Shamus, it looks great!

  2. Volfram says:

    “There aren’t very many levels. It’s a ladder with 4 rungs. You’re not going to get vertigo when you reach the top and look down.”

    While I can’t say I enjoy it, because I have almost no practical experience with it, this is one of the things I really respect about the Disgaea series. The power curve for individual characters starts at sea level and goes intergalactic. I saw one of my friends once take a standard swordsman mook and train her until she would terrify demon lords.

  3. Nano Proksee says:

    Finding that sweet spot between making the upgrades meaningful and noticeable and not breaking the game must be really hard, specially the movement speed.

    • ET says:

      I think he’s got a good start by only having ten levels.
      Makes it easier for him, since he’s got less things to manage/balance, and also more noticeable since they’re bigger slices.
      But yeah, the speed one might be hard.
      Maybe call it “movement” and have even levels do acceleration, and odd do speed improvements?

  4. Bropocalypse says:

    I notice the guy has white lights instead of red, now. Any particular reason for that?

    • krellen says:

      Because he’s either really really upgraded or has no upgrades at all (not sure where the spectrum begins). Shamus has previously mentioned that Good Robot’s lights change based on its power level.

  5. Eldiran says:

    I also have a LOT of anti-influences. It seems like some of the best influences is a good game that gets some systems horribly wrong. It makes you feel like you need to (and can easily) do better than they did.

    For example, Elder Scrolls may be one of my favorite series, but darn if it isn’t one the most “inspiring” too.

  6. Syal says:

    That’s definitely my preferred kind of leveling system (for a shooter anyway; passive stat bonuses with caps that indirectly exclude each other), though I can’t say I’m a fan of weapon energy. I’d rather have computer-controlled weapon cooldown that lets me fire slowly but infinitely than have to throttle it myself.

    (Obviously the correct build is Full Missile.)

    • Zeta Kai says:

      From what I could tell, Shamus McFamous implied that the energy would regenerate with time. Leveling it up would probably just make it recharge faster.

      • Trix2000 says:

        I like the idea because it leaves room for a burst-fire mode of play, where you blast out a ton of shots in a second, then dodge around while you ‘reload’.

      • Syal says:

        I would certainly hope it would recharge, otherwise it’d be like a mage running out of spells, but I still prefer infinite firing with set cooldown times over burst firing with variable cooldown times.

        Maybe he could set it up so if you hold the button it’ll only fire in such a way that the energy gauge remains at the same level, and then you can override it by tapping the button for bursts.

        …or have a whole different button for energy-saver fire.

    • Paul Spooner says:

      yeah, full missiles seems a bit OP. One stat to boost all missile related attributes? Amazing!
      My suspicion is that this is balanced against missile scarcity.

  7. The RIght Trousers says:

    I approve of your robot tentacles.

  8. Grenaid says:

    Does shot speed refer to rate of fire or velocity of bullets? It might be clearer just to refer to it as shot rate.

    • Jeff says:

      Yeah, “rate” would be a better word.

      Rate of fire, or recharge speed, or delay reduction. “Shot speed” sounds more like “the speed of the shot”, or “the velocity of the bullet”.

    • Scampi says:

      On a note related to bullet velocity: I’d enjoy some kind of range attribute. Maybe one could start with a mediocre range and increase it to sniper range through an additional attribute. I believe in some games increased range was created by manipulating a kind of “bullet life x bullet speed” equation, as any bullet lived a predefined length of time and speed would influence how far it would be able to move in its given time. It might be another idea for an attribute.
      Also: I wonder if every attribute will have appropriate feedback to it.
      If, e.g., Shamus says a high damage output with low shot rate will have a shotgun feel: will the shots (or the impact) look different if they deal large damage than if they were peas fired from a slingshot?

  9. Artur CalDazar says:

    I think you hit one of the most important parts noting levels need to somehow feel significant.

    Because it isn’t actually possible to get everything in ME2, but you can get close enough that it feels exactly the same. How something feels is usually more important than how it actually works.

    I don’t really see the point about few levels though, to me its more about feeling a lack of progress onward than seeing a lot of progress looking back. If the same XP is needed either way then looking back its the same, but moving forward its vastly different.

    • ET says:

      If you have too many levels, the difference between them won’t be immediately apparent.
      Sure, the player will notice eventually that they’re mowing down mooks who used to be tough, but I assume Shamus wants the levels to be obvious both short AND long term.

  10. TMTVL says:

    So, you thinking of porting it to Linux?

    • Shamus says:

      It’s possible. I’m trying not to preclude Linux. So far the tools I’m using (SDL, OpenGL, and OpenAL) are all available on Linux. But I’ve never ported anything like this so I don’t know how hard it would be.

      So I plan on trying, but I have no idea if it will result in a useful port or a time hole of frustration and bugs.

      • Neko says:

        Not precluding it by adding some weird .NET or WPF or XNA or COM dependency or whatever is all I ask =) At least it should be fairly Wine-friendly.

        Also, I like your Energy / Rate of fire design, it reminds me a lot of one of my favourite games – Tyrian 2000.

        • Svick says:

          Developing in XNA is probably more friendly to Linux than Windowns right now, considering that MonoGame (Mono’s port of XNA) is still under active development, but XNA itself isn’t.

          But obviously Shamus isn’t going to go that way.

        • Blake says:

          Dat game.

          Tyrian 2000 was a masterpiece. I played heaps of it, used heaps of the secret ships, learned how to do all the different twiddle key things that cost you armour but had special effects, and I’m quite certain there was still heaps of secret levels I never found out about.

          Man, I should really play that again. Anyone know if it’s purchasable on GOG or something?

      • Df458 says:

        If you release it but decide not to try porting it to Linux, I(and probably many others) would be happy to do that for you.

          • Bryan says:

            Thirded, but only if source code is released, because otherwise I can’t help. :-) But just like I’ve ported all three of Shamus’s previous programs to Linux (though I don’t think I ever posted about the heightmap-based terrain generator), I could do this one pretty easily I suspect.

            Both Frontier and Pixel City have forks (on bitbucket and github, respectively, since that’s where the originals were) allowing them to build and run on Linux. I just found out a couple days ago that Frontier had a few bugs still (which were somewhat covered by *other* bugs, whee!), and am hoping to get some time this weekend to figure out how to backport the parts of the patches from the fork of my fork that are required to fix them.

            But I already have a lot of helper code written for stuff like Ini.cpp (which I wouldn’t be at all surprised to see show up in this project too), which uses Windows APIs but needs a Linux equivalent. Also the GL extension stuff. So I expect it’ll be less work, unless Shamus has thrown away *all* of his previous projects’ code.

            (Which is possible, though I think unlikely. We’ll see, if source is released.)

            If source is *not* released, though, beware: releasing binaries means you’re tying yourself to at least a minimum glibc version, which not everyone may have, and probably also a C++ compiler version, which not everyone may have either. (Also a runtime linker path, which if your distro follows the SysV ABI shouldn’t be a problem, but not all definitively do.) Source is buildable pretty close to anywhere, and definitely anywhere by someone who knows how to code for the target.

  11. Vegedus says:

    Please include a respec option. It doesn’t matter if it cost xp, a level or can only be done once per game. Please have it. 80 % of RPGs should have an respec option. Two major reasons:

    Having a do-over if you mess up your build. I get the sense you’re deliberately not balancing the system too much. Cool, whatever. This also means a player can inadvertently screw up their build, make a character that’s completely ineffective. Maybe they make a good, reactive build that then turns out to be useless against a specific enemy. This is even more likely to happen when you want every increment to be meaningful. You can argue that there’s mitigating factors, like you can just grind more levels. But this can be serious amount of time you need to do that, all because you failed at effectively guessing which build is best in the long run. I’m playing Castle Crashers at the moment, which is hardly even an RPG. There’s a speed stat that’s only really useful for one kind of character, but since it’s vaguely described and not sure of my build in the beginning, I put some points into it anyway. Plus, I had lots of them. Except at one point the game suddenly effectively halves the rate at which you level. Now those wasted points laugh in my face every time I watch my stats. Had I put any more points in it, I might as well have started over. And starting over because didn’t guess which stat wasn’t good is super not fun.

    The second is variety, experimentation. You make a comparison with a weapon in shooter, but the big difference is that these are weapons, which you can switch between. And it’s fun to do so. If you can’t respec, you’re effectively saddled with the same weapon for the entire game, that just gets incrementally better. And what if I put all my points into a stat and then decide I don’t like, even if it’s effective. That I THINK I want the shotgun, but I afterwards realise the fast firing would be more fun. Again, it can only be fixed by started over or intense grinding.

    So please, put in a respec option.

    • Mike says:

      I’ve quit multiple games when I realized after 20 hours of playing that my character was completely worthless going forward and the only way to fix it was to redo those 20 hours differently.

      So I’m saying I agree, respeccing is awesome. Unless your sure there is no possible combination that could make you unable to complete the game, which based on your description I think is not the case.

    • Zeta Kai says:

      I don’t think that this game would really need it, based on what I’ve read so far, but I agree that most RPGs (& other games with RPG elements) need to have some reallocation options.

    • Humanoid says:

      It’d be interesting to compare a design where such a thing is done because the player *wants* to tweak/overhaul their playstyle, to a situation where a player is encouraged to switch around their abilties to meet an upcoming challenge. They may notice for example a prevalence of fast-moving enemies at a particular point in the game, or conversely, some lumbering tanks of foes.

      Now the latter approach could be interesting in that it introduces some tactical planning aspects to the game, but on the other hand, it could get pretty tedious if overused, and if the challenges are too narrow (such that a clear-cut optimal setup emerges), may result in sameish experiences between playthroughs and between players.

      • Nathon says:

        This was a problem I had with Guild Wars 2’s PvE content. You find a play style that’s fun for you, then the designers say “No, this boss is immune to debuffs. That’s not how we want you to have your fun.” Requiring people to change the way they play the game (and a specialization is just that) is like having a game that claims to be an open world sandbox but really just lets you turn your head and look at the landscape as you follow the rails.

    • MichaelGC says:

      The respec option will be unlockable from the Good Robot Online Store…

    • Mephane says:

      Yepp, I think respecialization is a must for all games where you spend points on skills and there is a cap to the maximum number of skill points you can get. More often than not, what works and feels fine at lower levels, turns out to be not so effective in later levels, or the player might just change their mind about some decisions etc. I also disagree with it being limited, it should just cost a reasonable amount of whatever ingame currency (or XP) is appropriate. One of the major fun aspects with skill systems is to experiment, to find out by trial and error what works for oneself and what does not.

    • Daemian Lucifer says:

      Indeed,respec is the best thing since sliced bread,and should be available in every single game that has leveling.

    • Scampi says:

      Once again I find myself on the empty side of a cleavage…I’ve never been a friend of respeccing. I used it in Titan Quest, but I also realized it felt more like a cheat mode than a welcome feature (to me). I understand how people come to like respeccing, but it’s just another reason why I will never regret not playing Diablo 3: I’ve been playing D2 for a long time and enjoyed the time it took me to learn how to create my effective builds, starting a character and knowing from the start: “And THIS (mental predesign) is how I want this character to look on this runthrough.” Even when modding, I preferred deactivating respeccing.
      Also: it made it all the more awesome trying to continue even with a messed up build, finding some unusual solution that helped overcome challenges.
      But since I’d just not have to use it: I sure wouldn’t veto this, even if I could;)

      • Vegedus says:

        Yeah, see, personally I thought the way you could experiment with builds in Diablo III was the funnest part of the game. I’ve noticed this is almost a controversial opinion, but mostly around people who liked D2 better, so eh, different strokes. Anyway, in D3, you’re constantly optimising your build. On the first playthrough, something like every fifth enemy way, you’re tinkering with your build, optimising. Learning which spells have synergy and which do not. Not because you have to really, to beat the game on normal, but because it’s fun to constantly discover new combinations. Bosses also generally have different effective set-ups, which may take a death or two to learn. And it’s just super fun to constantly experiment with builds, and an experience you just can’t get in D2 in the same way. It’s also a wholly different game from planning your character though, so I really don’t blame anyone for preferring something else.

        It is also a matter of time. I get sick of most games gameplay before my playthrough, and I have 200 games in my steam library, so even if I like planning my next character ahead, I’m never actually going to make it. I’d rather play something else. Didn’t even beat D2 all the way through more than once.

        Myself, I have some conceptual ideas about a puzzle game based entirely on respeccing stats before every enemy, planning what combination you need to defeat them.

        • Blake says:

          I agree with all of this.

          Respeccing in D3 was awesome. So many different combinations of everything to try out.
          With D2 I didn’t have 50 hours to spend trying out a new skill in the off chance I found it more fun.

          D3 also meant changing your skills based on your drops, have awesome cold spell as wizard, find some item that deals cold damage, change to awesome fire spell.

          Most important thing to make the respeccing work though was having the skills work based off weapon damage, it meant early skills were still completely viable for the entire game, so each skill choice was a playstyle choice (or to deal with a particular challenge you were facing).

      • Daemian Lucifer says:

        As with any such option(quick save,save anywhere),I have only this to say:
        If it feels like a cheat to you,you can always not use it.But for plenty of us who dont want to start a new character simply because we made a mistake somewhere during the last 10 play hours,there is no way to simply undue that mistake if the option is removed because “You will play it hard,and you will like it!”

        Im all for making it a hardcore option that will remove such a feature and give you an achievement in the end if you play like that,but not for making it mandatory for all.

        • WillRiker says:

          So much this. Not allowing respec’s as a way to make a game harder is really just an incredibly frustrating form of Do It Again Stupid.

        • kdansky says:

          The game designer’s job is to make a good game, and that includes balance. The player’s “job” is to play it. It’s wrong to ask of the player to play with a self-inflicted handicap to fix broken design, like turning back time with what should be a pause function: Rogue-likes do it right, they save when you quit the game so you can continue playing later on, but they don’t allow you to fix all your mistakes for free. I could hit my face with a hammer until Farmville got difficult to understand, but that’s no excuse for Farmville!

          As for respeccing, it’s not that easy. If respecs are free, then choosing which stat to pump is meaningless (and having meaningless decisions is bad, right?). If they don’t exist, we run into the 10-hour replay issue if the game is padded for length to begin with, which is just as bad. If the game is short enough for maximum enjoyment per time, I don’t see how replaying it would be a punishment to begin with.

          Stuff that could work if the game must be padded: You can respec at any time, but you lose a few levels in the process. You could find a limited number of respec tokens. There are quite a few respec tokens, but they only respec one point: you can fix mistakes, but you can’t rebuild your character completely for every level.

          • Daemian Lucifer says:

            And who says that the hardest difficulty level has to be the default and only difficulty level?Why must “save anywhere” and “free respecs” be the odd thing out?

            I say that roguelikes do it wrong,because I am yet to play one that wont end with me messing with it so that I can save whenever I like because I already saw everything the early start has to offer,and have died on a later level because of some stupid mistake.

            When I want to play a difficult game,I bump up the difficulty and/or impose penalties on myself.But thats not my mood all the time,and thats surely not the default mood of every player all the time.And any game that caters just to that single narrow mood is a bad game.

            • kdansky says:

              Then how about we let it be a proper game instead of a movie (which is becomes if you remove the game-elements by making them trivially easy), and add cheat codes for you? Then you can play a version that always lets you win, while we get something interesting, where our decisions and strategies matter, because we can lose.

              It’s much more sensible to make a hard game and have an easy mode, than it is to have an easy game and slap a hard mode on top of it. Making a game harder needs play-testing, or you could make it impossible to beat. Making it trivially easy does not, because players like you can just handicap themselves if the designer went too far, right?

              A rogue-like with a low difficulty curve and no perma-death is just like Diablo 3. Quite boring to all but the most casual of players.

          • Daemian Lucifer says:

            Actually,now that I reread this,I find it odd that you complain about padding,when no respec and no save do exactly that:Pad the game unnecessarily.Sure,it is fun going through FTL the first few times when every event is new and you dont have the feel for all the weapons.But after that,having to go through all the same stuff again just to reach that single event that doomed you last time because of poor luck is just padding.Sure,its fun staking the first few levels of hitman,and starting from scratch once or twice,but having to go to the start of a mission and wait out 3 same pointless conversations for the umpteenth time just so you could learn that fourth guards patrol route is padding.

            But with respec,you dont have to spend extra resources to make the intro mission exciting for the 50th time,but rather on making those last few missions just as fun on the first few tries.With quick saving,you dont have to focus on making that one conversation/cutscene not boring on the umpteenth go,when you can make the next 3 conversations/cutscenes be as exciting to the newcomers.

            Heck,even I wanna be the guy knew that,and gave you a save on every screen,because practicing that first screen to perfection stops being fun once you master it.

            But go ahead and argue how challenge comes from not knowing the game instead of from well designed levels.I will point you to a walkthrough of I wanna see you suffer,and then challenge you to pass it on your first go.

            • kdansky says:

              I Wanna Be The Guy doesn’t have quick-save! It has an abundance of well-placed checkpoints. That’s the complete opposite of quicksave-scumming! You’re advocating the same thing that I am: Maximum convenience for the player, but not trivializing the content by allowing what is essentially a tool-assisted run.

              Super Meat Boy has really short levels, and that means the levels are much more interesting than Mario, because Mario can’t throw curve-balls at you by the dozen without being impossibly hard.

              As for FTL: The problem with FTL isn’t that the early game is boring. It’s that the early game is far too easy compared to the boss fight, and after getting good at it, you are only going through the motions while collecting stuff until the last boss, where you actually need to spend resources. Adding a respec option before the boss would not make the game better at all.

              • Daemian Lucifer says:

                “That’s the complete opposite of quicksave-scumming”

                No,its not.Unless you are arguing that baldurs gate doesnt have quick save because you cannot save in combat.

                As for FTL,I dont know if respec would improve the game,but having manual saves definitely does improve it.It was so worth it restarting the game every time I wanted a manual save.And no,the boss fight is not the only goal or challenge of that game,nor the only allure.

      • StashAugustine says:

        Simple: implement a steeply scaling cost for respeccing- that way you can get out of a “Oh God I put all my points in Swimming and Enviornmental Resistance” hole but switching around builds for different areas isn’t going to be feasible.

    • Nathon says:

      I don’t necessarily disagree, and in some games I wholeheartedly agree that the ability to change all your skill point allocations is essential, in others it can damage the fun.

      It’s really all a matter of how much time investment it takes to beat the game. If you can play the whole game through in an hour or two, it can be fun to try different builds and have them be constrained by the need to make it through the whole game. If it takes 20 hours to get to the end, and there’s a chance that you’ll get through 19 of those hours only to find that your build is untenable, then it won’t be fun to have to start over.

      • Syal says:

        I’m not entirely sure how I feel about respecs. I keep thinking of Final Fantasy 5, where you change your character’s powers to fit the particular threat and then go back to whatever you wanted.

        I think the best option option I like most would be to allow full respeccing (at no cost) whenever you get another stat point, but only then.

        And maybe an option to replay a level with a different build, just to see what works.

      • Blake says:

        Agreed, if you have to spend ages building up a character then you don’t want to throw that time away to try something new.
        But if the game only takes a few hours, then it’s perfectly fine to try a different path from the start.

    • Muspel says:

      I’ll throw my weight behind this, too.

      One thing that really irritates me about a lot of games is that they force you make choices about your character build long before you understand the consequences of those choices.

    • droid says:

      Diablo 3’s respecs are well designed for its purpose, what a shame that the auction house undermines its core gameplay.

      But the pattern of unlimited free respecs isn’t universally applicable. It works for Diablo because it isn’t a game as much as it is a lifestyle focused on optimizing drop rate.

      A shorter game would have a lower switching cost. A more flexible game save system would let you backtrack without starting from scratch. A game with fewer skills than D3 can be better balanced so that every build using N points feels powerful.

      Personally I lean toward infinite free respecs, but I recognize it isn’t the only way to make a good game.

  12. krellen says:

    I’m really feeling the Descent influence. A lot of these robots seem like they’re re-purposed mining droids or something.

  13. Humanoid says:

    What’s the current feeling on non-quantitative aspects of levelling? Y’know, things like feats/perks, or perhaps more interestingly, one-off mutually exclusive decisions like (new) XCOM’s. I mean, the proposed system does appear on the surface to satisfy the “interesting decisions” test – unlike DnD’s often-empty level-ups where you’re told “here’s +1 hit and 5hp, have fun with that”.

    I guess the conflict here is that there at 60 skill ‘slots’ available and if it’s one point per level you’re expecting the game to end at level 40-50, and with such a degree of granularity, it goes back to “oh, I’ve levelled up and increased my damage by 5%, cool I guess”. Then again it could be 25 levels of 2 each, or 15 levels x 3 points, which would help in that regard. However with no knowledge of how the pace of levelling ‘feels’, it’s hard to say whether that’d be a sufficient amount of sugar hits over the course of the game to keep the motivation up.

    Now obviously by this point I’m sort of groping around blind, but obviously it’d be too much to make up to sixty ‘this-or-that’ decisions, but at the same time, it introduces the (debatable) issue of some levels being worth more than others. Ditto the Mass Effect 1 or Alpha Protocol system of some arbitrary points in the skill line granting something extra over the usual linear increase. But trying to provide 15 decision points in a playthrough is probably already pushing it, let alone 25-50 of them.

    So uh, I’ve typed a lot of words and just ended up where I started via some musing without providing any solutions. But throwing up an idea – I’d nonetheless like to see a bit more specialisation such that player A’s robot isn’t only differentiated from player B’s by having four points in one skill and six in the other, and vice versa: So how about a mutually exclusive decision, say halfway up each line, that can cater for diverging playstyles?

    Say for shot power, after the initial few points in which you get a grasp of the mechanics, you choose whether further points will give you bonus splash damage on shot impact, versus, say,piercing shots that can pass through foes? Rate-of-fire versus projectile speed? Shield capacity versus recharge rate? (Actually that last one doesn’t sound too interesting since it doesn’t change gameplay from “dodging is good”)

    Okay, I’ll stop armchair designing now.

    • Thomas says:

      I prefer this system a lot ^

      There’s a chance to get really Dishonoured if you do it correctly too (although that’s harder). Say one upgrade makes missiles homing and a different upgrade makes you missiles fragment into smaller missiles… etc. It always seems more fun to develop a new strategy than to make an existing one more effective.

      In Mass Effect 3 the most fun upgrades were the ones that changed how your abilities work, getting perks in F:NV was way more interesting than getting skill points etc. (Also getting guns that fired differently in ME3 was more fun than upgrading or the normal game thing of steadily more effective guns)But I like the mixing it with more normal levelling to because then you still get the same feeling of growth.

      Note: I’m not making suggestions for your game Shamus, just commenting in general =D, I’m outside the target audience and it’s probably annoying if you want to go one way and people keep saying ‘it’d be cool if you went off in this completely separate direction)

      • swenson says:

        And that’s one thing I liked about ME3’s leveling system–you actually had the chance to make mutually exclusive, interesting choices. (ME2 tried to do this, but because for many builds you’re not putting 4 points into all skills, it only occasionally even mattered.) Do you want Area Charge or Heavy Charge? There’s an actual, visible difference in gameplay. And once you get one, you can never get the other. The hours I spent agonizing over which to choose for all my multiplayer characters! (well, minutes, anyway)

    • Jordan says:

      The whole ‘5% better’ problem is also why I pretty much never put points into the weapon skills in the Fallout games over stuff like Speech or Lockpick where meeting thresholds for skill checks was always a far better bet in my eyes than an arbitrary and unclear increase in some damage variable.

    • I don’t think Shamus ever explicitly said the levels were linear. So the assumption of like 5% increment every level may not be accurate.

      And of course, the other part probably goes back a couple of lessons: Feedback! OK, speed and hitpoints and even energy level not so much, but if you increment your weapon power I suspect your bullets are going to look cooler and sound more impressive and make snazzier explosions and maybe one or two feedback methods I haven’t thought of (cause more robot knockback), and with all that you’d probably feel like you were making serious progress even if it didn’t actually do any more damage at all. Same goes for missiles.
      I suspect you could even incorporate some nice effects for increases in rate of fire.

      • Humanoid says:

        I would expect them to be linear though, as otherwise it creates a problem where you’re either encouraged to spread out your points due to diminishing returns, or to stack your points because of increasing returns.

        But yeah, the 5% figure was just an example, since it’s what typically happens in your d20 based games. I’d certainly expect different values depending on which skill is being increased. It’s likely already pushing it if movement speed increased by 10% per point for example, I can see difficulties balancing controls if you have to account for one player moving twice as fast as another at the same point in the game. So we may see the 5% hold for movement speed, but a significantly higher proportional increase per point for the various weapon skills for example (where 10% might feel too low if anything).

    • postinternetsyndrome says:

      I like this, and was going to post something similar.

      I enjoy it when games give me unique options instead of just making numbers go up. As was mentioned, Dishonored does this really well with the powers having just two levels, that always changes the effects significantly with level two. Then again, that system was pretty badly balanced as well.

      The idea of putting in divergence points in the upgrade trees seems to me like a nice middle ground, with the potential for lots of experimentation with different builds.

  14. Risven says:

    Your system of leveling looks great. It reminds me of the system in Iji, one of the best side-scrolling platformers I’ve ever played (and it’s free, so check it out!).

    Is this a game where it will be possible to just avoid enemies and try to run to the exit or boss without killing anything? Just wondering.

    • Paul Spooner says:

      Yeah, Iji was really neat. I played through the whole thing and it had a very solid feel to it. I kind of resented how dark the story became, but I guess that’s just part of the setting. It went a long way towards “humanizing” the foes, even though they were aliens.
      The particle effects at the end caused massive lag though, too bad they couldn’t balance it for performance any better.

      • Cerapa says:

        The story only goes dark when you don’t do a certain thing and kill people.

        A full pacifist run with the beforementioned thing to do isn’t all that dark. (not pure good, but…hopeful?)

        Still it’s a really good idea to do a killy run before you do a pacifist run. Can’t do a run with killing without feeling like crap after a pacifist run.

        • Trix2000 says:

          That and the pacifist run is a bit tricky, even moreso if you don’t know what you are doing.

          • Risven says:

            Yeah, I really loved how you came to understand the enemies. I never had problems with end game lag, but I can see how that would be annoying. But yeah, like Cerapa said, you can make the end a bit brighter depending on how to handle events in the game.

            Pacifist run is fun and tricky, but it should come after a shooty-killy game. And with Iji’s level up system, there are so many play styles to try out, I rarely get bored.

  15. Warrax says:

    Nice fake-out; for a split-second there, I thought you were going to say something nice about Mass Effect 2 :P And I was all liek “whoa, what?”

    • Aldowyn says:

      I knew it was coming from the beginning. I literally tweeted ‘I feel snark incoming’.

      Of course, it sometimes feels like the entire world is constantly trolling me about mass effect, so there’s that…

  16. Humanoid says:

    Separated from my unfortunate wall of text because they’re just straight questions:

    Are the skill levels the only modifier to your base baby-robot abilities, or is it in addition to some sort of static improvement per level?

    Presumably the impact of each skill point will be linear, but for example you say “you’ll be able to pump out huge volumes of ineffectual bullets” if you have shot power as the dump stat. But in effect, will these just be balanced counterpoints to each other – e.g. twice as many shots for half damage each for a player maxing speed over power? Or is there a certain expectation to at least try a balance, lest your shots be literally ineffectual (say like in Fallout when you can’t beat the DT)?

    What kind of progression are we expecting here over the course of a typical game? Say, up to +100% faster movement (10% per point)? +400% damage? (Or, like a JRPG, more like +4900%?) Or, in reference to the first question, like an additive 10% per player level, then a multiplicative 10% per skill point (so like 50x damage at level 50 compared to level 1, rising to 100x at 10/10 shot power?)

    • ET says:

      I’m thinking Shamus needs to add a damage threshold/armor stat of some kind to the enemies.
      Otherwise, pumping up damage wouldn’t have any benefit over pumping up shot rate:
      You’d be able to kill the same bigger number of enemies per second with shot rate, but it also gives an indirect boost to aiming, since you can lead your shot easier.
      Unless, the DPS increases per level of the damage stat are bigger than those of the shot rate stat, which I guess would work.

      • Humanoid says:

        On the other hand, it’s probably trickier to keep the damage going on an enemy while dodging as you’re trying to keep a fire hose pointed at them constantly, rather than being able to focus completely on moving in between instantaneous shots.

        Another aspect that may need to be balanced is to minimise wasted damage due to overkill, which would compromise the shot power skill, though granted unless enemies get taken out in one or two shots without it, it’s not so big a deal.

        • ET says:

          He could have piercing/semi-piercing shots!
          Maybe I’m imagining this, but I thought I played a SHMUP one time where overkill shots went through enemies, but had reduced damage afterwards.
          Kind of like Trample from Magic: The Gathering.

        • Richard says:

          The firehose/sniper choice is a good thing to have.

          While the absolute DPS might be exactly the same, some players are really great at making a small number of heavy shots count, while others are much better at spraying a large number on a moving target.

          For example, in Freespace 2 (and mods) there are several primary cannon which are pretty devastating if you hit, but the firing rate is such that you have to snap-shot pretty carefully to hit something, and others which spray so fast that you are definitely going to hit a few times, but you’re going to have to hold it on-target for a while to destroy anything.

          By making the DPS almost the same you reward both types of player – the ones who can snap-shot and the ones who can hold a firehose steady.

      • Thomas says:

        Having Damage increase at a faster rate than speed DPS wouldn’t be imbalancing either because of the overkill problem. I think it should be tweakable enough to work

    • Shamus says:

      No static improvements per level. If you don’t spend your skill points, you never change.

      The increase in player speed currently goes from 3.0 (starting level) to 5.25 (max level). That doesn’t sound like a lot, but it actually… is? It feels like a big deal, anyway. The top level speed is about as fast as you can usefully go without just bouncing off the walls, completely out of control. And the low speed is as slow as I dared make the player for fear of driving them crazy.

      Of course, the high/low ends might change based on other stuff I fiddle with. Low speeds are more tolerable if the camera is closer. (It’s like the feeling of walking down a narrow hallway versus walking across an open gymnasium. The latter feels slower, even if you’re walking at the same speed.) High speeds are more tolerable if the level geometry is more spacious.

      So, w’ll see. But the gap isn’t massive.

      EDIT: In fact, this is pretty much my goal: To make the low end of everything as feeble as possible without driving the player away in the first few minutes, and then ramping up to as high as it can usefully go.

      • Alrenous says:

        Make the top speed slightly higher. Partly for players with excellent reflexes, partly for a safety margin, and partly so the player has the option to make a chaotic bouncy ball of a ship if they want.

        Which reminds me: what happens if your player makes a mistake levelling? How does it deal with ‘oops’?

      • Tizzy says:

        Does this mean that the camera will zoom out as a function of the top speed? current speed? neither?

      • MrGuy says:

        Curious about this. I know you’re thinking about playing this with a controller on the PC. Is there a keyboard option? One thing I’d worry about with “fast as you can get away with” speed with the fine control of an analog stick is “way faster than you can control” with the binary “go/don’t go” you get with a keyboard.

        Or maybe it’s an argument that I shouldn’t focus on speed as a keyboard player…

        • Shamus says:

          I do most of my testing with the keyboard and mouse, but most of my playtesters use a controller. When it comes time to nail things down and tighten up the mechanics I’ll probably need to do some serious side-by-side comparisons.

          And yeah, most likely some skills will change in value depending on control scheme. You probably don’t need as many bullets when you’re aiming with a mouse, but you probably can’t use as much speed with a controller.

      • Steve C says:

        Suggestion: Have the very start (tutorial) have lots of power. Then have the bad robots steal that power. That way new players get to see what they are in for to hook them in. The feeble power settings boring them away in that initial 10-15mins would be ashame.

        Note I don’t know if this is a good idea or not so I’m not advocating it. Just a suggestion.

  17. Alan says:

    I don’t know if the picture of the skill system is supposed to be representative, or is actually how it appears in game, but a heads up: the green/amber combination you’ve got for filled/empty is tough for me to distinguish, thanks to my red-green color blindness. I suspect others will have similar problems. The easiest way to make it easier to see for me is probably to make one of the two significantly darker than the other.

  18. Paul Spooner says:

    “As the game goes on, formidable foes become common, common foes become mooks, and mooks become cannon fodder.”

    So, are you saying that there are foes that literally shoot robots at you? A Bad Robot Cannon?

    Coincidentally, I showed my daughter a drawing I made, and told her it was called Good Robot. She thought for a moment, and then said “Where do the bad robots are?”
    Indeed little one. Indeed.

    • Zeta Kai says:

      For centuries, & indeed since the dawn of time, man has asked himself “Where DO the bad robots are?”

      Soon, thanks to our dear friend Shamus, we may finally know.

  19. Mephane says:

    Seriously, when do you accept preorders? Also, if you plan to release this in multiple languages, consider this my offer to do the German translation, for free.

  20. Daemian Lucifer says:

    On your first point:

    I kind of prefer a mixed system.For example,the first three levels of damage all cost 1 point of xp,and give you 1 point in damage,but fourth level costs 2 points of xp,but gives you double shot,or some other unique thing.

  21. LazerBlade says:

    “This game is as much about skill as it is leveling.”

    Yes! It’s hard to find a game where skill and level are just as important, and then either they don’t usually have a decent leveling system like this, or their action mechanics aren’t gloriously fun the way this looks. Maybe this won’t please everyone, but the people it does please will be hard pressed to find something comparable.

  22. swenson says:

    “It’s not that hard to max your character out, which means that those choices you made about where to spend your skill points weren’t all that important.”

    Unless, of course, you’re going for an Insanity vanguard run, in which case you had better be sure you put them in the right places or you’ll spend even more time dying than a usual Insanity vanguard run. Seriously–only one point in Charge until you max your passive. It makes more difference than you’d think.

    Oh, we’re talking about Good Robot, not Mass Effect?

    In that case, I like this. I prefer when you can choose an upgrade path within an attribute (aka Mass Effect 3, only more so please), but this is also preferred to anything approaching ME2’s leveling of pointless pointlessness.

    I have to admit 10 divisions seems like a lot to me. Like you say, if you’re not seeing/feeling a difference each upgrade, then what’s the point? Not very satisfying. But that’s just my gut feeling without actually playing it–so maybe it truly is noticeable in-game

    The other potential concern is having too many attributes–a la the original Mass Effect. Then you just get lost because you don’t know what you want to level, how useful/pointless each is going to be, you’re kind of overwhelmed, especially when you’re getting a very small number of points each time. If you keep it at these six, I doubt that’s “too many”, but I don’t know if you’d want to go much beyond that. Not sure how others feel on that subject?

    • Khizan says:

      I would totally disagree with this; maxing Charge is basically the most important thing you can do, because cycling charge/nova is so effective. Both abilities have invulnerability frames, and combined with Charge’s barrier restore you’re basically an unkillable juggernaut.

  23. ooli says:

    May be “Adverse influence” is the term your looking for.

  24. Attercap says:

    “If you dump everything into shot speed and energy you’ll be able to pump out huge volumes of ineffectual bullets.”

    I hope you mean extremely minor damage rather than completely ineffectual (so 1 point of damage rather than 0)? One of my gaming pet peeves is being given the option to min-max to suit a particular style that I like for a game only to find that I can’t proceed further because my min-max choices weren’t considered. I don’t mind spending a long time attacking a boss with hit-and-run papercuts, so long as I know those papercuts are eventually going defeat my foe (unless it defeats me first).

  25. Cerapa says:

    I really like the addition of energy to the standard ROF and DAM. Without it, leveling for the most DPS is trivially solvable, and could be replaced with a single stat.

    • Syal says:

      Except when there’s a mix of enemies that one type is more effective against. Weak flyers get mowed down by high-speed low-power, but then you include a slower enemy with invincibility frames (or counterattacks) after being hit and low-speed high-power is the way to go against them.

  26. Darren says:

    I love discussing RPG leveling mechanics! One that I think gets next-to-no love is Skyrim’s. Obviously it’s not a hugely complex system, but it creates an interesting three-tiered effect:

    1. Skills in which I have heavily invested skill points and which are my best skills

    2. Skills which I use frequently but which are not so critical that I have invested many, if any, skill points

    3. Skills that I’ve ignored

    Now, we could obviously debate how much impact you get from leveling up and allocating points, but it’s an interesting way to approach the issue.

    • Attercap says:

      I think a variation on your #2 (on the list, not bodily functions) is why I liked Daggerfall’s skill set up better as an RPG. One built skills based on use. Of course, that led to the easy abuse of creating a macro that made my character jump around their room, swinging a sword and casting spells in order to level up while I slept. But I didn’t get trapped in the path of “that sounds cool, I’ll invest points” and turned out not using as much as I hoped.

    • Aldowyn says:

      random note on Skyrim: Since the feats aren’t balanced at all, I tend to ignore some of the trees even if my skill there is fairly high. Speech, for example, tends to get to 50 or so, but I never put points there because there’s more important things.

      So I thought of an idea to fix that: You get to choose a feat for a skill every 10-15 levels of that skill. Also, leveling up, instead of choosing where to put that 10 points of hp/magicka/stamina, you get a point in the corresponding stat every time you level up a skill.

      Wait, now we don’t have character levels, just skill levels. Success!

      • Octapode says:

        The other thing that the Skyrim system resulted in, at least for me, is I wanted to use alchemy, but because the skill points were better spent on, say, smithing, I was not able to make anything very useful with alchemy, so I didn’t use alchemy very much, so I couldn’t get the useful alchemy skills even if I had wanted to.

        That makes me think, would a leveling system where abilities (such as being able to make the next level of armour, or perform a knockback attack, or fire several missiles at once, etc.) were treed, like the skill system in skyrim, but were independant from the actual skill level, which provided a pure numerical bonus, be interesting to use?

  27. Henson says:

    I think I’ll contest your first point on ME2’s system. While increasing costs can be quite bad in some games, in ME2’s case, it actually works out fine, due to other elements:

    1) The fact that all powers tied to a single cooldown. Diversifying is less beneficial than min-maxing because you can only use one power at a time anyway. Because of this, you probably don’t even want all of the upgrades (ex: both push and pull).

    2) Leveling points are so abundant that, even with increasing costs, it is not overwhelmingly punishing to min-max.

    3) Upgrades for weapons & armor, separate from character upgrades, give the player enough power that, like above, it is not overwhelmingly punishing to min-max.

    Of course, these choices create new problems as well, problems that kill the system for me. I just don’t think increasing costs are an inherently bad decision; I know other games whose leveling systems I loved (i.e. NOLF 2) that did the same thing.

    As for point 2, I wonder if the 4-rung ladder would have worked had they given the player 20 skills to choose from rather than just 6; going for breadth over depth. Even though it wouldn’t take long to complete a tree, there would still be 19 other trees from which to choose. I mean, take Final Fantasy Tactics: that leveling system is basically a bunch of ladders with only one rung each, but there are so many different ladders to choose from with so many different effects, we don’t really mind.

    At least we can all agree that ME2’s system was not very good. Use the expletive of your choosing.

    • Eldiran says:

      I would say Final Fantasy Tactics is more like a dozen ladders where you can choose which rung is next. And you have each foot on a separate ladder. And certain ladders you can’t get at until you’ve climbed earlier ladders.

  28. Crystalgate says:

    Increasing costs. While I agree that Mass Effects 2’s system wasn’t very good, I think you got the wrong reasons.

    The increasing costs doesn’t alone encourage people to spread the points around. If skill cool down worked like in the first Mass Effect, that may have been the case. However, in ME 2, using a skill will force all skills into cool down. So, you can not hit an enemy with singing, baking, and schmoozing one after another, you have to choose one of them. By spreading the points out, you are now limited to hitting them with one level 2 skill instead of one level 4 skill. Granted, you’d have three choices of what level 2 skill to use, but I have seen few to no games where the flexibility is worth the loss in raw power.

    You also praised a similar system in Depths of Peril.

    Anyway, on to your stats.

    Shoot Power of Shoot Speed are very similar, they increase your DPS. It seems to me there will be an optimal way of increasing your DPS and if you don’t use that way, you don’t have a different build so much as you just gimped your character. Let’s say Shoot Power increases your damage per shot with 20% per point and Shoot Speed increases your rate of fire also by 20% per point. If you spend 10 points in those stats, you get the best DPS by spending 5 in each. By spending 10 points in one of the stats, you lose 25% DPS.

    Energy can fall into the same trap too, there may be one optimal way to spend points into the first three stats.

    The remaining three stats don’t interact with other stats so much, so I see less of a risk that they turn into a math equation instead of a choice of play-style.

    What I said doesn’t have to be true, it’s however the dangers I spot.

    • Eldiran says:

      That’s a good point, but it’s not quite that simple. Even without knowing details I can say Shot Power is extra beneficial for:
      – people with good aim
      – taking out a single large/strong foe
      – frontloading a large amount of damage (a foe that dies to 1 strong, slow shot but takes 2 fast shots would be better taken on with the slow shot)

      Whereas Shot Speed would be good for:
      – people with bad aim
      – taking out many weak foes
      – distributing damage evenly (a foe that dies to 2 strong, slow shots but takes 3 fast shots would be better taken on with the 3 fast shots.)

      There are a few relations that need to hold for it all to be balanced though… mainly, shot power and shot speed both need to increase energy costs by roughly the same amount. This way, if they follow your suggestion of splitting their skills, firing many strong shots quickly will be quite a drain on energy.

      I also think shot power ought to either scale DPS slightly higher than shot speed, (like 23% for speed’s 20%) or it needs a secondary advantage. For example, it might be cool if shot power also slightly increased shot size?

    • Ysen says:

      “While I agree that Mass Effects 2′s system wasn’t very good, I think you got the wrong reasons.”

      I would agree with this. “Increasing cost” skill systems exist in response to a problem present in many games, where leveling up your character was not a series of meaningful decisions, but one decision which then takes fifty hours to implement because dumping points into one thing was clearly superior to spreading them around. In that case you might as well not have leveling and just pick a skill to use at the start of the game. This is even worse when, as in a lot of games, you don’t even get the chance to try a skill before you wed yourself to it.

      Giving diminishing returns or increasing costs can make it more viable to have a varied build, and also means that one misspent point because you wanted to try something out doesn’t instantly make your build unviable. The problem is when it goes too far the other way and every character ends up with equal points in everything.

  29. Vagrant says:

    anyone notice there wasn’t technically a bad robot #15?

  30. Aldowyn says:

    oookay. Might as well do this, did most of it on twitter anyway.

    Point 1: other people have mentioned that the increasing costs are mitigated by several factors, including the single cooldown, so that there are reasons to specialize and reasons to diversify. For the most part, this is just a different philosophy.

    Point 2: Eh, I haven’t played ME2 recently enough, and I rarely use damaging powers anyway. I think they’re more significant than you’re giving them credit for, though.

    Point 3: Technically not true, as others have mentioned, but I’ll give you a pass since it really is quite close.

    Point 4: Pretty much the same as Point 2, really. YMMV.

    Point 5: This is the big sticker for me. No, no they don’t level with you. Not in any of them. Somewhere in the body you said ‘It will have the exact same number of hitpoints when you’re effortlessly tearing through groups of them on your way to the final boss.’ Well, that’s true for mass effect too :/ I seem to remember a lot of dead Geth on the side of the Citadel and a lot of dead Collectors on the collector base. Even Harbinger’s just a roadblock by that time, they have to give you a timer to make it even a little hard. ME3 is the only exception because there’s bigger harder enemies all over the place. :/

    Okay, now my obligatory ME-defending is done I’ll leave now. Bye!

    • StashAugustine says:

      “Point 2: Eh, I haven’t played ME2 recently enough, and I rarely use damaging powers anyway. I think they’re more significant than you’re giving them credit for, though.”

      I played through the Arrival and Shadow Broker DLCs with a broken mouse, entirely by sitting in cover and spamming Incinerate with the touchpad.

    • Aldowyn says:


      TL;DR except for #5, mostly just different styles or too subjective to argue. with #5, I’m pretty sure the base statement (foes level with you, in reference to ME) is factually false.

      For the record, my initial reaction (I think this was before I got to the 5th point) was “‘meh, fair enough’. Different styles for different games.”

  31. Aerik says:

    A humble UI suggestion from a professional UI developer:

    Have you considered having the low end of the Energy and Shields bars be the same point at the middle of the screen? They would “fill up” towards the left and right edge of the screen, and “drain” towards the middle.

    The way it is now, when you’re in dire straits, trying to fight your way out of a bad situation, and your bars are low, you’ll have to look up to the far top-left and up to the far top-right to get an idea of your status, taking your eyes a long way from the combat at mid-screen.

    If both bars drain towards the center of the screen, then as you get into oh-crap-I’m-gonna-die sitautions, you don’t have to look as far (distance from mid-screen to top-middle) and you’re also more likely to only have one place to look (both bars would be low, and therefore very close to each other).

    The game looks fantastic, I can’t wait to see it in action!

    • Humanoid says:

      I’d also be interested in how it’d feel with a HUD-like translucent meters not anchored to the edge of the screen. Especially true of larger screens, UI elements anchored to the extremes of the play area can become all-too-easy to ignore. (Thinking default MMO layouts)

      Actually, a simple toggle that switches the UI elements from the top to the bottom might be a good thing too. Or if practical, freely movable elements so the player can drag or even rotate them to whereever they wish.

  32. Klay F. says:

    Man, Good Robot is turning into something I REALLY want to play.

    Also, I can’t wait till I can do a no upgrade challenge run!

  33. Goggalor says:

    One thing that might be nice to see is a “bling” bar you can put points into which doesn’t improve any combat stats but just makes your ship look cooler.

    It might appeal to the challenge-and-bragging-rights types, and might also increase the likelihood of people replaying. Plus it gives you a chance to add in some more particle effects, in case you feel that 4000 isn’t enough. :-)

  34. Kelmomas says:

    Shamus, I have to partially disagree with the first two points you highlight. I believe those are aspects that ME2/ME3 handles correctly, although in a different manner from Good Robot.

    By far the most common model of skill tree I see is the Diablo-like one where the starter ability is a minor bonus while the top-level ability is awesome. The problem is that this removes your choice out of most of the level-ups: once you’ve dipped your foot into one tree, it becomes more and more obviously better to keep specializing in it, because the rewards keep improving. And, slightly more subtly, it’s also typically just easier to focus your playstyle around exploiting one tree to the max, rather than combining several lesser bonuses (i.e. a fighter/wizard is typically more complicated to play than a pure fighter or a pure wizard).

    The “increasing costs as you level up” is a very effective way to address this problem (and way older than ME2: it’s been in tabletop RPGs for at least 20 years, probably more). You can offer tantalizing endgame superpowers to the player, but you can set them at a higher price so that dipping into some jack-of-all-trading remains an alternative worth of serious consideration. This is tricky to balance, but if done right it creates an environment of possible builds that is much broader than just “how many trees can I max out per playthrough?”.

    The same purpose, build diversity, is also served by your #2 point: having lots of short skill trees rather than a few long ones (note that ME2 bungled this by only having like 4-5 trees). It prevents you from putting literally everything in one tree and trapping yourself into a repetitive one-trick-pony gameplay, and it also makes it difficult or even impossible to try out all trees in one playthrough, leaving something new for replays.

    Now, if I’m reading the picture right, you went with the approach of making skill trees linear (i.e. Shot Power 1->2 is about as big as 5->6). This is another viable way of offer a balanced choice between specializing and spreading out, though it loses the “wow” factor of escalating rewards, and there’s the risk of players pigeonholing themselves into doing the same thing every time. On the plus side, it’s considerably easier to both code (you’re not creating new features for each upgrade) and to balance (everything has the same cost); also Project Good Robot is more of an action game so it doesn’t rely on skill tree appeal as much as other games.

    Anyway, I just wanted to defend the shallow-trees approach, which I don’t believe to be the mistake it is presented to be.

    • Humanoid says:

      Increasing costs is a balanced approach on a macro scale, but is problematic under scrutiny. Maybe I abuse this Meierism too much, but a game should be “a series of interesting decisions”. I would think that a key consideration here is to make each and every level-up an ‘interesting decision’.

      The problem the design is specifically trying to eradicate then, is that of “empty levels” where you make no interesting decisions because you’re forced to collect multiple levels worth of points to get your payoff. This is especially true in this game because, as stated, there are no inherent gains tied to levelling aside from the skill point granted, so increases that require pooling of points from levelling up mean you’re not gaining any power whatsoever during that time.

      There is a compromise that ME2/3 did not take, and that is to ensure that you are given sufficient points per level to always take at least one increase in a skill, i.e. if the top level rung of a skill costs 4 points, then you get four points per level, with the option of instead spreading the points around in multiple other skills. In the context of Bad Robot though, there will likely be too many levels for this approach to be viable.

    • Shamus says:

      To you, and everyone else who is defending Mass Effect.

      Now, this doesn’t mean that all games that work like this are bad. I’m sure this kind of system is a lot easier to balance. But it’s not the kind of game I want to make.

      I anticipated your objections and put that paragraph in there for you, specifically.

      • Heaven Smile says:

        After Mass Effect 3 in general (even before the ending), we can more or less guarantee that there will be a lot of aspiring writers and designers that will take “Anti Spiration” from ME and say to generations to come (in unison from all parts of the world):


        In the other hand, if that happens, its probable that Mass Effect will become a work of “High Art” for this. I mean, isn’t High Art supposed to inspire people to do better for the betterment of humanity? if the works of Art that change the world manage to exist in the first place BECAUSE of ME3 “awe inspiring” stupidity, then ME3 is High Art.

        Casey Hudson will become the Western equivalent of Akira Kurosawa, whose works became the inspiration of thousands of film directors to the point of being a milestone in the human race:

        But we all know that is not going to happen. We ALL know that such privilege belongs to George Lucas for his “awe inspiring” prequel trilogy. He is the one who disappointing hundred of millions to make others do Art before Hudson did.

        Yes, ladies and gentlemen, The Phantom Menace is High Art.


  35. Now, there ARE drawbacks to a system like this. Without level scaling, there’s no safety valve for struggling players. Sure, grinding for a couple of extra levels would help, but it will never negate the challenge.

    Here’s a thought.

    Let’s say the game map has 20 levels (you move from left to right, or is it right to left? then you go “down” one level/area).

    Each time the player reaches the end of the current level, level up the player.

    This ensures that you can always give the player a minimum amount of skill points to spend.

    While players that are XP vacuums will get a few more extra skill points to spend on top of that.

    How many skill points the player should get per level they travel I have no idea. But some minimum amount needed to increase skills to have a chance against the new enemies.

    As to those who gather XP by killing stuff or doing tasks (or is this just a shooter?)
    Well that’s the trick. Maybe you could reward XP for “sneaking” through a level, or XP for minimal combat (0 damage, or no body but shield damage only?).
    Maybe give XP for being able to shoot all rockets fired by the enemy at you?
    Some XP might be weighted by some percentage based on the shill level, rewarding someone for fully using a skill they are “good” at (or put points into).

    Either the points put into skills will act as a modifier to a level based skill base.
    Or it’s the other way and the level is a modifier for the skills and the points put into them.

    I never understood why modern games do not level you up when you enter a new level (there is a correlation in the words themselves there).
    I guess the correct Way to describe what I mean is: Story driven leveling with “free” skill points + xp rewarded skill points through gameplay.

    I hope you understand what I mean here!

  36. Tse says:

    I think there is room for more stats, both offensive and deffensive:
    Health regen- regen PART of your hp, more levels increase how much and how fast you regen
    Evasion- passive chance for a shot to do less or no dmg to the player
    Bullet spread- shoot more than one bullet at a time, for that shotgun feel
    Bullet range- for sniper types
    Explosive bullets- start with something like 10% aoe dmg, get to around 30-40
    Crits- for the gamblers out there
    The all into rockets also seems kinda op.

    Man, I’ve played a LOT of flash games.

  37. Jarenth says:

    I can’t believe nobody’s pointed out yet that the formerly-swastika robot is now basically Trogdor.

  38. MrGuy says:

    So, another random question. Can Bad Robots damage each other? Or only the player?

    I’m wondering if can bait an opponent into launching a missile, and then ducking out of the way as the missle flies harmlessly by me and hits his robo-brother behind me in the face.

    There’s an interesting return to maneuverability if I can use it not just to avoid damage but as a weapon…

    • I don’t know Shamus, nor do I know what he thinks. But If I where to speculate, he is a guy who considers logical thinking very highly.
      Thus one can assume that in the finished game any rockets or other projectiles can damage both the player character and non-player characters regardless of whom fired it.

      So if the player is able to be smart (by luring the enemy into the path of rockets, possibly even their own), then the player should also be able to be punished by playing dumb (by taking damage if the player manages to fly into the player’s own rockets).

  39. silver Harloe says:

    My experience with leveling systems in games suggests that if you don’t have a “60 possible level ups” (10 levels of 6 skills) and only offer 50 upgrades over the game, you need to either (a) explicitly mention there will be only 50 points early on, and also say “max level” when the player gets to 50 or (b) make the player explicitly choose how to limit their maximums (like in Alpha Protocol where you had to pick 3 skills you could max out)
    If you don’t, then you’ll get people complaining either or both of two ways: (1) “OMG, I had to grind for hours to get to max level” or (2) “the leveling system wasn’t balanced with the length because I got nowhere near max by the end of the game”.

    In other words, if a max level of everything isn’t explicitly ruled out early on, many people will assume a max level of everything is an eventual possibility, and many of those people will make it a goal and be disappointed if they can’t achieve it.

  40. Anachronist says:

    I have to say this. I don’t like the idea of “leveling” at all, particularly in a game like this. I’ll try to explain. Maybe we’re in agreement.

    I think of this “leveling” as two separate progressions: the player’s abilities, and the difficulty of foes. Those progressions can be independent, and judging by what Shamus wrote above, they should be.

    Player: When I hear “leveling”, I think of Dungeons & Dragons, in the sense of suddenly gaining new capabilities just because you reached some arbitrary point. I prefer the concept of designing your initial character, and sticking with it, gaining only small incremental improvements, but never really changing much, with the bulk of capability improvements coming from earning enough to upgrade your equipment.

    I had an epiphany when I played Call of Cthulu. It made me realize what bugged me about Dungeons & Dragons: the constant race for XP. I didn’t mind the arms race to get better equipment. But this step-function business of “leveling up” and suddenly being granted new, godlike powers just for earning XP, bugged me. In contrast, I found Call of Cthulu a refreshing, fun, challenging, well-designed RPG – with no player character leveling up! You start out with your unique set of skills and abilities. You can tweak them as the game progresses. Other than your character’s sanity, you’re pretty much the same character at the end of the game. Here and there you can improve a few skills, or buy better equipment.

    That’s what I like about Shamus’s concept. Earning points gives you options to make minor adjustments to capabilities, but you can’t end up dramatically different than you started.

    Foes: This is a game where you’re going through a tunnel. In RPG terms, you’re on rails. That’s fine. This arrangement suggests that the tunnel is populated with progressively more difficult and numerous foes as the player goes through. Here’s what I’d like to see:

    1. The foes don’t get harder just because you’ve played the game longer. They get harder because you have moved further on in the tunnel. So if you charge blindly ahead without proper preparation, you’ll be in trouble. If you hang back trying to earn points to upgrade yourself, you might be bored at the grind. But it’s your choice.

    2. The programmer has a choice whether to have a randomly-generated but fixed population of robots distributed in a progressively more challenging way, or have the game generate new robots according to a probability formula that accounts for the player’s position in the tunnel.

    There are advantages to both. A fixed number of foes gives you a feeling that you’re accomplishing something by clearing out an area, it means you can’t “grind” in the same room waiting for more robots to appear to earn more points; once you clear out a room, you have to move on, but you can still charge through without fighting if you want. A continuously generated stream of foes, on the other hand, gives a different feel to the game, keeps you on your toes, maybe suggests that there’s a robot factory somewhere that you must hunt and destroy.

    I just realized that this might be an interesting option to offer the player, and easy to implement: Do you want a fixed population of robots, or let them be manufactured while playing the game?

    In summary, it’s possible to have an enjoyable game that completely does away with leveling. Instead you have a continuum, for both the player and the foes. The player’s basic stats remain pretty much the same but can earn ways to purchase incremental improvements. The foes’ abilities don’t depend on the player’s abilities; instead, the foes get more numerous and challenging depending on your position relative to the big boss.

  41. aaaaaaa says:

    Hey man,

    I am not sure if you already know this but back in 1980 there was a game about robots where you had to shot-through an auto-generated maze with walls you could not touch. You were a “robot detective” (whoever that is) wielding some kind of disruptor and you shot robots who were auto-generated, chasing you in individual screens/rooms – the concept of scrolling would be only hardly doable for this kind of game on 8bit machines.

    And get this: the name of the game was Shamus. Check it out on youtube (shamus+atari), I am sure you can find some of your enemies there as well. It also spawn a sequel, it was also a very good game which however diverted from the original functionality. BTW, their original programmer later underwent a sex-related surgery and is now officially a woman.

    I also wanted to write you about a concept you’ve been pondering-on before. I call it “disruptive gameplay”: when you introduce a feature to the game which can disrupt the original gameplay to the point of completely changing the feel of it. If you introduce a several of such items/features then you get disruptive gameplay which changes the game ocassionally and the player is essentially not bored by the same style of gameplay in each level. If you actually plan this beforehand and playtest it well then it’s NOT having 2 or 3 games in one, that’s just a misunderstanding of how it works. Please check Binding of Isaac for disruptive gameplay examples. That amazing game is just full of it. Disruption in the gameplay is NOT something to be feared, instead, it should be embraced at every opportunity. And you have had several of them, the flashlight thingy comes to my mind right now as most prominent example.

    I will refrain from more game-design related suggestions although there are several other things I could share if you’re interested.

    • Alan says:

      In most situations it’s best to just refer to someone using the gender they request. Most transgender people want to go about their lives as any other person of the gender they identify as; calling out that they transitioned makes that harder. And generally speaking, one’s medical history is not an appropriate conversation topic unless they have broached it.

      • silver Harloe says:

        Name changes (for whatever reason – marriage, change of religious, change of sex, adoption, whimsy, or anything else) by someone who is published are relevant if they have works published under both names. (Though since aaaaaa didn’t mention a name, it does seem egregious here)

      • aaaaaaaaaa says:

        yeah, right, like that’s an important thing here in these comments. Get real LGBT.

        • Alan says:

          Basic politeness is always important. Unless, of course, one’s intent is to be offensive. If that’s the case, I find it unlikely Ms Mataga visits here, which make it a bit weird…

          • aaaaaaa says:

            /* offtopic

            You’re not right. Politeness for the sake of being polite is just hypocrisy. Ask Linus Torvalds.

            There is even a management style based on NOT being polite:

            So first of all, I refuse to be “polite” just because someone asks me to do it, without my own inherent need to be polite on any matter where I do not think it is relevant or applicable from my own standpoint. Next, I refuse to be turned into a black sheep by being called impolite while I was merely stating a by-the-way information in a matter-of-fact way, no intentions present nor implied. By calling me impolite or implying so from the basis of some weird LGBT political correctness you’re actually the offender and impolite person towards me.


            • I restrained myself from calling you out on your earlier comments, but now you’re officially being a jerk (or possibly troll).

              Thanks ever so much for contributing to the overall TwentySided impression than Linux fans are invariably idiots.

              [edited to remove a vulgarity, which implied anger, rather than my intended contempt.]

            • Shamus says:

              Linus is wrong, of course. Hostility breeds defensiveness and further hostility, which is exactly what you don’t want among engineers who are peer-reviewing each other’s work. Manners exist to smooth over human relationships when there’s ALREADY a risk of hurt feelings.

              More importantly: You can be as rude as you like if you’re Linus and you run the show, but when you’re a guest on someone else’s site there’s no reason to show up looking for a fight.

              On the LGBT issue. You said something that SOME people would find offensive. Even if you think they shouldn’t be offended, their offense is a reality. I get that you were just pointing out that this was an interesting and unusual person, like mentioning that someone once held a world record or is related to the royals. Unusual people make for interesting stories.

              You are free to heed or ignore Alan’s advice as you like. In fact, this is a great illustration of exactly the problem I have with the Linus method of interacting. You could have just ignored Alan, but instead you took offense and pushed back, and now people are angry and I’m moderating an argument that didn’t need to happen.

              Saying it’s hypocrisy to be polite when you want to be rude is like saying it’s hypocritical to not punch someone when you want to punch them. It’s a destructive behavior no matter how you package it.

              On the game Shamus:

              Man, I heard about that game often in the 80’s, but I had no idea what it was about. Kind of funny to realize it’s probably something I really, really would have liked.

              • aaaaaa says:

                Well I think that preaching LGBT political correctness in comments here is misstep and not productive, moreover calling anyone impolite or rude, which I myself find offensive. Me having a B fiancee is only an irony of this situation, I have no reason to be impolite or rude to anyone from the LGBT community.

                Moreover I will politely not agree with the politeness arguments above and that’s just about the end of discussion on this topic from me.

                I will refrain on commenting anymore on the above comments since I find them irrelevant to what I actually wanted say to Shamus. I am starting to think that not coming from english language environment may make some of my sentences above sound like angry or intentionally hurtful while it’s not the case at all.

  42. John Lopez says:

    Late to the party, but there are several good games on Kongregate that use the leveling system you are describing.

    I for one approve of adding harder foes while leaving the “baseline” reference points of the weaker foes around. There is something awesome about mowing through a cloud of targets that earlier would have crushed you under the weight of the assault. Even better is playing a game one time where your lasers cut through targets like butter while another play-through you have fire and forget missile barrages.

    If the first play-through with lasers and the second with missiles upgraded also make different targets easier/harder to overcome, that is the best “secret sauce” for replayability.

  43. Veylon says:

    On the subject of grinding, one method I’ve seen is to allow the player to keep whatever XP they gained thus far through a level when they are killed and sent back. So a level, upon repetition, eventually becomes easier as the player not only becomes familiar with it, but is more powerful.

  44. WJS says:

    I so agree on the use of enemies in the mook/cannon-fodder role that you first encountered in the elite/miniboss/full-boss role. I still remember fighting my way down towards the final boss in Skyward Sword, and going through literally dozens of moblins (or whatever they’re called these days). Not that they were the toughest enemies before, but that many at once really gave that “Holy shit I’m a badass” feeling.
    Of course, you could also go the Oblivion route, and not even have the common courtesy to palette-swap the goblins when they follow you up the level-ladder to godhood…
    It’s sad how few games seem to do such things – in fact, I can’t seem to think of a single one that I’ve played that ever demoted a monster that got the full-blown “this is a boss” presentation down to the rabble of the common enemies later on.

Leave a Reply

Comments are moderated and may not be posted immediately. Required fields are marked *


Thanks for joining the discussion. Be nice, don't post angry, and enjoy yourself. This is supposed to be fun.

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>