Diecast #304: The Take Two Backstab, Obduction, SpaceX

By Shamus Posted Monday Jun 8, 2020

Filed under: Diecast 120 comments

Hosts: Paul, Shamus. Episode edited by Issac.


Link (YouTube)

Show notes:

00:00 Moving

Two more weeks until moving day. Fingers crossed.

02:34 Take-Two’s nefarious backstab.

During this segment, Paul played devil’s advocate and proposed an alternate theory for Take-Two’s egregious behavior: What if the developers were doing a bad job and TT just wanted to get the project back on track?

This theory is a little wobbly, since TT was trying to poach the team for Kerbal Space Program 2. If they’re doing a bad job, then why are you trying to hire them? But you could fix this by suggesting that, perhaps, the leadership of indie developer Star Theory was terrible / difficult to work with. So maybe TT just wanted the team under their own leadership. Allowing for the fact that this isn’t likelyNone of the anonymous sources at Star Theory were concerned about their leadership., what would be the appropriate course of action for Take-Two?

Let’s imagine that the game is way off-track, has a bunch of bad ideas that fans will probably hateWe added a first-person gunfight minigame where you have to kill four waves of aliens before you can plant a flag on a new world, and at the end The Illusive Kerbal has you build a Crucible and blow up the space center to stop AI from killing all organic life., looks awful, and none of the core physics systems work properly. The Star Theory leadership refuses to accept any changes. Now imagine you’re the president of Take-Two, and let’s also imagine you’re basically a good person that wants to put out a decent game. What’s your move?

  1. Take the project away from the current developers and hand it to a new team. It will take MONTHS to hire a fresh batch of devs, and more time for those people to coalesce into a creative  team, and more months for those new people to decipher this half-constructed game so they can get the project back on track. Result: Star Theory goes out of business, the current devs become unemployed, and the game is pushed back a year.
  2. Cut your losses and start over with a different studio. Result: Star Theory goes out of business, the current devs become unemployed, and the game is pushed back two or three years.
  3. Go to the developer and issue an ultimatum that they need to change course. Result: The devs call up Jason Schreier, who publishes an exposé about how the heavy-handed publisher is “ruining” the game with strong-arm tactics by threatening to shut the project down. Nothing changes, the game continues on its misguided course, you get a bunch of bad press, and now you’re forced to pick one of the other options in this list. Also, when the game finally comes out, you’ll be blamed for the lack of quality.
  4. Start a new studio and cut funding to Star Theory. Attempt to “quietly” offer the current rank-and-file creative people the opportunity to join the new studio. Result: Star Theory goes out of business, the current devs can join the new studio and smooth out the transition, and the game gets pushed back just a few months.

In this scenario, I think that Take-Two’s behavior would represent the least bad of all of the available options.

12:45 Paul plays Obduction

25:12 SpaceX Capsule looked like low-budget movie set.

Correction: In this segment I said the teacher on the Challenger disaster was Terry McAuliffe. I’m not sure where “Terry” came from, but it was Christa McAuliffe. I was 15 at the time.

35:38 Mailbag: AI Cheating

Hello Diecast,

Your last episode’s segment on Civ VI, and some of the comments it inspired, brought to mind a related topic I have felt passionately about since first getting into strategy games as a child: AI cheating, wherein AI players get to do things like ignore fog of war, receive free resources from the ether, what have you. I hate it. It sucks. It’s also a near universal staple of the genre, as well as many others.

There are valid reasons for this. I’m in the software field, and I fully understand why good AI is so rare and challenging to implement, as well as the game designer’s dilemma that AI that was genuinely really good probably wouldn’t be all that fun to play against either. The ubiquity of aggressively stupid AIs that cheat like hell suggests that’s the best tradeoff for designers to take, but designers don’t always make the best choices. Can you think of AIs in games, from any genres, that you found particularly memorable? Challenging, devious, at times possibly quite dastardly, but still clearly fair and playing by the rules?


49:21 Mailbag: Difficulty

Hi Shamus and Paul!

I have been enjoying listening to the Diecast for quite a while and have recently returned to it, listening through old and new ones. Really appreciate it. In any case, question below:

Difficulty is something that comes up with some frequency both on the Diecast and the Blog. Not really surprising considering that games tend to work by different rules, have different mechanics, pacing and so on and so forth. Now, what I had been wondering about is the question what kind of difficulty is particularly tough on you. Or rather: Are there genres that you have an easier time with than with others?

For instance, “the mystical Quicksave key” (as Yathzee once put it) is something that makes me shrug off repeated deaths in FPS’ comparatively easily and I suppose this is likely the same for many other people. With RTS’ on the other hand I tend to break off campaigns early if the difficulty ramps up too quickly and I have no other reasons (story, atmosphere) to continue. And that’s in spite of being fairly competent in that genre. I also tend to have trouble with difficult RPGs that are relatively open. The old Fallouts and especially Temple of Elemental Evil are good examples. Icewind Dale II is difficult but linear. ToEE on the other hand is difficult and is fairly open in a number of ways. Thus it’s often difficult to know what the problem is, when having trouble with an encounter. ToEE is pretty much the threshold for me, with Dark Souls already being on the other side. Excessive Trial-and-Error, Third-Person-timing-based combat, long runs back to encounters, shitty port etc. all contribute to it for me.

What kind of factors such as game mechanics, difficulty design, game length, pacing, weak story or writing, weak atmosphere, uninteresing setting etc. contribute to you personal diffulty threshold? (Especially in regards to some recently mentioned titles such as Dark Souls and Doom Eternal.)

Greetings, WaveofKittens

56:02 Mailbag: Historical Preservation

Hello Shamus & Paul,

Since 2009 I have belonged to a group that has been attempting to create a free private server for the NCSoft MMO Tabula Rasa. It has been a difficult process because NCSoft continues to issue cease and desist orders for a game that was shut down 11 years ago.

Do you think historical preservation is important for online games such as this? Why or why not?

May all your hits be crits,
Leslee Beldotti



[1] None of the anonymous sources at Star Theory were concerned about their leadership.

[2] We added a first-person gunfight minigame where you have to kill four waves of aliens before you can plant a flag on a new world, and at the end The Illusive Kerbal has you build a Crucible and blow up the space center to stop AI from killing all organic life.

From The Archives:

120 thoughts on “Diecast #304: The Take Two Backstab, Obduction, SpaceX

  1. MerryWeathers says:

    I don’t what is it with Take Two Interactive resorting to blatantly evil corporate tactics nowadays but it’s pretty amusing.

  2. Narkis says:

    I remember reading bad things about KSP’s management some years ago. With some quick googling I could only find this, which has less details than remembered but is a good enough indicator. I wouldn’t be surprised at all if T2 came to the conclusion that the leadership is terrible and tried to hire the rest of the studio from under them. They were trying to sell the studio anyway, so the end result would be the same, minus incompetent leadership.

    1. Geoff says:

      From what I understand from the Bloomberg article / Googling, KSP: The First was developed by Squad, bought by Take-Two in 2017, and then contracted Star Theory for development of KSP: The Second. So, KSP1’s dev / management problems are separate from this issue.

      Re: Paul’s question about Take-Two’s ability to cancel the contract: Most work-for-hire contracts include payments per milestone and clauses to allow the publisher to cancel the contract with a certain stipulated notice and payment (ie: some combination of 30 days notice, payment in full through the next milestone, cancellation fee, etc.). If I was guessing, that’s probably a large part of how they were planning to fund their developers pitch development ahead of GDC.

      All that said, I’m really surprised that there apparently wasn’t a “no-hire” clause in the contract. Every contract I have ever done / worked under in my career has included a clause that prevents either company from directly soliciting hires from the other company. Often those clauses can be hard to prove, but in this case it seems pretty blatantly obvious when every member of the company receives a recruitment message specifically saying they are doing that. Seems like Star Theory should have been able to take them to court and win enough money to keep the business going for awhile.

      1. Decius says:

        A simple court case can be done in a year, but most take 20 months before appeals. Star Theory can’t afford to keep going during the lawsuit.

  3. Joe says:

    Hey, Shamus, when you’ve finished moving and have recovered, is there any chance you could write up any kind of advice/pitfalls thing? Turns out I’ll be moving later this year, much to my frustration. It’d be great to know what the process is like these days.

    Apparently corporations have to show increased growth on a yearly basis. Yeah, you made half a billion $ last year. But if you don’t top that this year, you’re in trouble. Apparently cancelling/poaching games is a path to profit? I dunno. And you’re right about a third of the staff. That’s the number I remember from Jim Sterling’s video.

    The Wikipedia page for Tabula Rasa is interesting. In development for 6 years. High staff turnover. Led by Richard Garriott. He says he was fired. The game shut down soon after. With that kind of turmoil, I can’t imagine we’ll ever know the full story. There won’t be a War Stories or a Noclip. I suspect that everyone had to sign an NDA. Which is a pity, because it’d be good to know the pitfalls. What went wrong, and maybe some ways to avoid the same trouble.

    1. Joshua says:

      In the past 8 years, I’ve lived in 4 houses and 3 apartments, so my family has moved quite a bit (long story). Are you moving from an apartment to a house, apartment to apartment, etc.? Are you located in the U.S. or somewhere else?

      General advice:
      1. If you can afford to hire movers, do it. For us, it’s usually just been a few hundred dollars, and avoiding the exhaustion and physical injuries is worth it. I had a kidney stone after one of those moves, and the medical costs greatly exceeded costs of movers, not to mention the discomfort.
      2. Go through all of your stuff first, and decide if you really want to keep everything. It really sucks to move things that you decide you don’t need after arriving at your new place. Obviously, this somewhat depends upon how much knowledge you have of the space and needs of your new abode, but there are still plenty of things that you should consider discarding.
      3. Accept that at least one thing will break after a move. If it doesn’t happen, be happy.

      As far as the growth, it’s very hard for most businesses to just “tread water” for very long, so it’s assumed that you are either shrinking or growing, and staying the same is just shrinking that hasn’t started yet. Certain markets will reach saturation, but it’s expected that a company will use its assets and its accumulated business wisdom to penetrate new markets or expand existing ones. This is especially expected of public corporations. Maybe Jim Bob’s Crab Shack in small town can stay afloat for a number of years doing about the same business (although not forever!), but publicly traded companies are a lot more volatile. Expected growth is also a significant part of stock evaluation. Amazon is a company famous for a high Price/Earnings ratio, meaning that investors are HEAVILY expecting significant future growth.

      1. Joe says:

        I’m in Australia. Thanks for the advice. I suspect much of it will be relevant.

        1. pseudonym says:

          I recently moved as well. I can second Joshua on the hire movers part: do it!
          When we moved we let the movers also do the packing. We both have jobs and this saved a lot of time. They basically packed up all our stuff in one morning, deconstructed the large furniture, drove to our new place, unloaded everything and recostructed the large furniture. They were so productive, and so little involvement from us was needed, that we could start cleaning the old house and were mostly done by the time we left. Also during the unloading we could alreay unpack nearly everything that was supposed to go in the kitchen. It was great! It was worth every cent.

          A second tip for hiring movers: make sure they are insured for breaking things. While this did not happen during our move (as they were very skilled) it is still comforting to know that they are able to refund you if they fatally drop expensive equipment.

          A third tip: let a few moving companies come to your house first and have them make a price estimate. The prices can wildly differ!

          As regards to doing away with stuff before the move. Do this with big and space-taking stuff. The very small stuff doesn’t add up to large volumes. And you’ll know that you haven’t used it anyway after it is still in the moving box for a few months. By that moment it is much easier to make a decision to throw it away.

          1. pseudonym says:

            Throw away should have been: bring to a second-hand shop. There is alway somebody else you can make happy with your stuff.

            1. Syal says:

              If you go to the dump in person you can give it directly to the people who go there to find stuff.

          2. Sleeping Dragon says:

            If you’re the one doing the packing that is a great time to figure throwing some stuff out rather than just put it mindlessly into boxes. Admittedly I’ve always moved solo so did not have to consult with family members or housemates which might have eased the process.

    2. GoStu says:

      Moving Pro Tips Time! I don’t know your age, approximate strength of arms/legs/back, or financial situation, but I can generalize a bit.

      My last couple moves I rented & drove the U-Haul myself. Moving companies are upcharging a bunch for the just-in-case liability on hauling all your stuff. If you’re willing to take that on yourself, you can bring the cost down a bunch. Not sure how it works down in Oz (saw your other comment) but here they had a loose relationship with a few day-labor companies. Rather than paying for the full cost of movers, I hired laborers for ~3 hours on each end; they’re there to pick up the heavy crap (couch, bed, etc.) while I move lighter things around on my own.

      If you’re moving between cities this is an even better deal; you’re not paying for their hours on the truck. Pay for their time loading, pay for the other laborers time unloading.

      It’s not as cheap as “break your back loading it all yourself, then rub salt in the wound unloading, while begging friends and acquaintances to assist for the price of pizza”… but it’s cheaper than hiring a moving company. All-told I think I paid about ~35% of what the movers quoted me; I just had to drive a shitty truck a couple hundred kilometers and do medium amounts of lifting.

  4. Gordon says:

    There’s a GDC article about how the F.E.A.R. AI works http://alumni.media.mit.edu/~jorkin/gdc2006_orkin_jeff_fear.pdf
    It’s fun how much of it is smoke and mirrors.

    1. The Rocketeer says:

      Now, let’s look at an example of how the STRIPS planning process works. Let’s say that Alma is hungry.Alma could call Domino’s and order a pizza, but only if she has the phone number for Domino’s.Pizza is not her only option, however; she could also bake a pie, but only she has a recipe.So, Alma’s goal is to get to a state of the world where she is not hungry. She has two actions she can take to satisfy that goal: calling Domino’s or baking a pie. If she is currently in a state of the world where she has the phone number for Domino’s, then she can formulate the plan of calling Domino’s to satisfy her hunger. Alternatively, if she is in the state of the world where she has a recipe, she can bake a pie.If Alma is in the fortunate situation where she has both a phone number and a recipe, either plan is valid to satisfy the goal. We’ll talk later about ways to force the planner to prefer one plan over another. If Alma has neither a phone number nor a recipe, she is out of luck; there is no possible plan that can be formulated to satisfy her hunger.These examples show trivially simple plans with only one action, but in reality a plan may have an arbitrary number of actions, chained by preconditions and effects. For instance, if ordering pizza has a precondition that Alma has enough money, the satisfying plan may require first driving to the bank.

      As someone that recently replayed FEAR and its expansions, I can’t tell you how much this changes the game for me.

      This was a fascinating read. Thanks for posting it.

      1. Retsam says:

        The Rocketeer is alive!

        I don’t have anything to actually say here about your comment, I just noticed awhile back that you hadn’t commented in months, and it’s always a little scary when internet people disappear. Good to see that you’re not dead (or that the afterlife has wifi). I hope the banal apocalypse is treating you well!

        1. The Rocketeer says:

          I live in a converted freezer trailer with Daemian Lucifer and Andy Kaufman. Daemian sells krokodil, I write a weekly sex advice column, and Andy keeps the power on with his accountancy job.

          For real, though, I work in the ER of a large regional hospital, but COVID hasn’t actually treated me badly. I just haven’t had a lot to comment on as far as blog content goes, and whenever the “whatcha playing” post goes up, I don’t notice it until there’s a couple hundred comments already and the iron has cooled.

  5. Kathryn says:

    The AI cheating that always irritated me was in Mario Kart, when the Princess had an endless supply of shrinking toadstools, Bowser had an endless supply of those rotating things, etc., while I was stuck with whatever I’d gotten from the square I drove over. Seems to me that if I was driving the Princess, I should get the endless toadstools too.

    (Yes, I’m old.)

    1. Platypus says:

      I think you just hit on a brillant idea, instead of having complete random drop in mario kart they could make each character have certain “favourites” that they are more like to get from a drop or maybe even a unique item for each character. Would add abit more diversity to the gameplay and more personality, like Luigi could be this games Gandhi but with bomb ombs instead of nukes.

      1. BlueHorus says:

        I can already hear the arguments, just at the suggestion.

        “Bowser AGAIN? You ALWAYS play Bowser!”
        “That’s ‘cos he’s the best character!”
        “That’s the point. It’s no fun when you’re playing Bowser.”
        “So get better at the game, noob. Or play Bowser as well.”
        “No. I want to have fun, not ruin other people’s.”
        “Bowser’s fine. You’re just shit at Mariokart. It’s in the game, the devs meant it to be like this.”
        “I’m not playing if he plays as Bowser. Bowser’s OP.”
        (etc etc etc)

        1. Geebs says:

          On reflection, I wonder whether the ruthless rubber-banding in Mario Kart is actually not intended to balance gameplay, but instead is an expression of “Derukuihautareru*“.

          * “ the stake that sticks up gets hammered down”

          1. tmtvl says:

            And every other proverb gets proper spacing, sigh. Oh well: “deru kui wa utareru.”

            1. Echo Tango says:

              You know you can edit Wikipedia right? That’s like, its whole deal.

              1. tmtvl says:

                I know, but seeing mediawiki markup gives me a massive headache.

        2. Abnaxis says:


          Am I seriously the only person who remembers Double Dash? Also, did they take special items out for later entries, because individual characters had their own special items in that game. Toad/Toadette were the ones that got the infinite toadstool.

          Also IIRC certain characters had greater likelihood of picking up certain items, but I’m less certain on whether that was actually true or my in confirmation bias.

          Incidentally, I haven’t actually played Mario Cart since Double Dash, because the Coop with me driving while my wife shoots stuff is what sold it for us. We still play it sometimes…

          1. galacticplumber says:

            Yeah they took that out and hard. Dramatically unbalanced things as some unique items were just undeniably superior, and even if you didn’t like the light weight class of the characters who had your likely favorite, you could have a heavier character teammate to get the non-light kart you wanted.

            Probably not though. Light high-acceleration characters tend to be better anyway in addition to having chain chomps, endless mushrooms, triple red shells, and so on.

            1. Asdasd says:

              We played a lot of Double Dash at university. The specials were fun and kept things interesting as you didn’t always get who you wanted (we drafted characters in ABCDDCBA order). But yeah, it was a very double-light-character dominated meta among us, apart from one guy who tried to buck the trend by always going double-heavy. He didn’t usually win, but if he got far enough ahead he was impossible for us to catch with our low top speeds.

  6. GoStu says:

    Memorable AIs

    Good Example: Dark Reign
    This was a two-game RTS franchise that began in 1997 and managed a sequel in 2000 before vanishing into obscurity. I mostly remember that the player could take advantage of the AI and delegate some chores to AI control. First, you could give some generic broad commands like ‘Scout’, ‘Harass’, and ‘Seek and Destroy’ to a unit and the AI would carry them out for you. Scouters would just seek to view the map and find where the enemy was; Harassers would engage right until they were fired upon and then run; Seek & Destroyers would scout the map and engage whatever they found. One could also tweak a few behavioral settings, like “Damage Threshold” – on lower settings, units would automatically retreat to the nearest repair/healing facility.

    Good Example: Half Life
    A lot of the enemy AI could be pretty clever with space to maneuver, particularly once you got the human marine enemies. Flank attacks, cover, etc.

    Bad Example: Mario Kart
    I think this one’s notorious because you’ll NEVER be truly ahead of these guys. You can take a secret shortcut and cut off half the level, and yet they’ll simply teleport to you through egregious rubber-banding. I think the nature of this game is as a party game that it shouldn’t be easy to get in an unwinnable position and that back-of-the-pack players can still influence the game (hence the Blue Shell).

    Bad Example: Civilization series
    The difficulty where the AI has parity with you is Settler. (At least in 5, not 100% certain about that on 6 or 4 or earlier)
    The in-game description claims Prince is even but this is a damn dirty lie.

    1. Echo Tango says:

      I myself was thinking of Half Life during this segment. None of the AI in that game is particularly “smart”, but the pre-programmed behaviours they have feel right for each of the enemy types. Phrasing it like that makes me think that’s a big part of having AI that feels “good” without having to actually be very smart. Like, when I found out that the AIs in Red Alert or Civilization were cheating in their own ways, the games felt cheaper, but when I found out that the soldiers in Half Life only had a few simple behaviours, I gained appreciation for the game. The simple dumb things the AI is doing are all concealed by other enemies or hazards in the scene, or by dark lighting or fog, so the magic-trick (so-to-speak) works. Same with Left 4 Dead 1 and 2 – the AIs are all pretty dumb individually, but they add up to an enjoyable experience. :)

      1. Sleeping Dragon says:

        In truth most players don’t want to play against a strong opponents, most players want to play against opponents who are satisfying to beat.

    2. BlueHorus says:

      That Dark Reign AI sounds great. My main complain with any RTS game is the micromanagement; you as the player often have to function as the basic common sense of YOUR ENTIRE ARMY.
      From having your defenders just sit there getting shot to death because the enemy is out of range to the classic Tank Army Traffic Jam than can (and will) cause assaults to utterly fail, the AI can utterly put me off.

      How reliable was this AI system? I mean, if the game series failed…

      1. evilmrhenry says:

        The Kohan series is also of interest if you don’t like micromanagement, as you build squads that you manage as a group. This means you usually only have a half-dozen units to control at a time.

        1. BlueHorus says:

          Yeah, it’s a similar system that got me to enjoy the Dawn of War games. Being able to reinforce and upgrade squads in the field (even as they fight!) was perfect – get my new resources into the fight quickly and efficiently, as opposed to the standard ‘I need more [X], but my [X]-building facility is on the other side of the map so they won’t arrive in time’ problem you get with RTS games.

          I remember reading a complaint online that ‘DoW was just fighting with blobs that take too long to die’ – which I get, but I’ll definitely take it over ‘you looked away from that unit for 10 seconds and now it’s dead of its own stupidity’…

        2. Mistwraithe says:

          Kohan is the best RTS ever made (or at least it was and I haven’t heard of anything better since).

          I say this mainly because strategy is pretty important in Kohan and there are some major trade-offs in the gameplay. Many other RTS games are heavily dominated by micro such that a slight deficiency in micro ability will outweigh a more substantial advantage in strategic thinking.

          Dawn of War is quite good too, but the economy isn’t as strategic.

      2. GoStu says:

        It was good enough to give a rough direction to and then check up on it occasionally. The “Scout” function could put your units veering off into the obscure corners of the map, and the game’s terrain could get a little troublesome*, but by and large it was pretty trustworthy.

        If I had to blame a single thing for Dark Reign’s lack of success, it’s probably competition from bigger studios. That was the era where RTS games were on top, so to speak – Dark Reign released in 1997, when Starcraft and Command & Conquer were the big names of the genre. Command & Conquer: Red Alert alone dropped two expansion packs that outsold Dark Reign. I don’t think there was anything wrong with Dark Reign… it was just up against bigger competitors.

        If you’re a fan of RTS games of that era and can dig up a copy, it’d be worth a look. Its campaign had some unique ideas and it was fun. It probably had the largest clusterfuck of a final mission I’ve seen anywhere outside Ender’s Game too, to the point that I remember it here 23 years later. I wouldn’t pay too much for it though – I suspect that it’d really show its age.

        1. Hector says:

          To clarify, the terrain system was wonky because its semi 3d. Every spot on the map has an elevation which affects units in a couple ways, but it’s difficult to tell what’s what because it’s a 2D engine.

          I can vouch that I’ve never beaten the final mission, which is Nintendo hard levels of difficulty. It makes Dark Souls look like Candy Crush.

          1. GoStu says:

            For those who’ve never seen the mission: You are playing as a technologically-advanced but largely peaceful people on a planet that the two major galactic powers are fighting over. The Imperium are attacking the Freedom Guard and both sides have massive bases that dwarf yours. You start in one corner of the map, and your technological advantage means you can build units from either side’s arsenal.

            Over the course of the mission(s) prior you’ve been replaying historical battles between the Imperium and the Freedom Guard – you may choose to play either side in the battle, and once you complete one for either side, you’re permitted to advance to the next mission. The game also informs you of how that battle ended historically.

            In the final mission, the Imperium will eventually manage to overwhelm the Freedom Guard and destroy them… which ends the mission in failure, as the Freedom Guard are defending an orbital defense system that is preventing the Imperium from simply scouring the entire surface of the planet. Your entire people will be wiped out as collateral damage if you cannot prevent this. You must defeat both enemy armies; if the Imperium survive, they annihilate the planet; if the Freedom Guard survive, a new Imperium Force will mobilize later and try again.

            At first both sides are focused on each other, but the Imperium will eventually come for you. After some time, the Freedom Guard will come after you as well. Because their forces outnumber yours terribly you need to be incredibly cost-effective to have the slightest hope in hell… and you need to damage the Imperium pretty badly within a fairly short time limit to give the Freedom Guard a chance of defending their orbital defenses.

            The only way I managed to succeed (after a few tries) was to use mass artillery units and order them to continually fire on the approach routes the Imperium (and afterwards, the Freedom Guard) start attacking you buy. With a continuous blizzard of super-long-range AoE damage you can eradicate all their ground forces as they approach. Some judicious anti-air flyers are needed to pick off the airborne attackers but those aren’t as much of a problem. Once that was done I built EVEN MORE artillery and anti-air flying units to use as spotters, and progressively bombarded about half the Imperium base into rubble.

            That done, I went and kneecapped the Freedom Guard down by about a third, then blew up the rest of the Imperium, and then the last of the Freedom Guard. (Taking the entirety of one side out at once caused a frustrating defeat as their entire war machine was free to hit me from all angles, overwhelming even my blizzard of artillery fire).

    3. Geebs says:

      Halo’s enemy AI has always been a cut above most shooters.

      Personally, I hate cheating strategy game AI when it makes the game confusing to learn, as in “how do they have those units already? Am I doing something wrong?”.

      Then again I’m bad enough at most strategy games that I have basically the same experience playing against humans as well.

    4. Syal says:

      Best example: Tetris Attack.

    5. Mikko Lukkarinen says:

      Bad Example: Space Empires V
      I remember this one time I picked the hardest difficulty AI, generated a galaxy made of clusters of 5-10 stars, fewer warp points so not all of the clusters were connected and would require warp point creation (late game tech) to access. Given hundreds of turns to build up unmolested, especially when I started in an area with only two connected clusters while most AI had four or more, you might assume that the AI should be able to put up a fight. Nope! By the time I got warp point tech, 7 of my light cruisers were strong enough to wipe a fleet of ~220 AI light cruisers and cruisers with no losses. On top of the weak-ass fleets, the AI had carriers with no fighters, planets with free building slots, nothing was terraformed, and their economy was fucked. I assume they just kept building ships until they bankrupted themselves, lost a bunch of shit to lack of maintenance, then got stuck unable to build anything new and never tried to fix that.

      1. Sleeping Dragon says:

        At least in this case it seems that the AI was following the same rules since it did manage to bankrupt itself.

  7. Ninety-Three says:

    I had exactly the problem Paul mentioned with Obduction: puzzles take a long time to execute, not because they’re fiendishly challenging but because you pull this lever, spend thirty seconds walking over there to see what it changed, go back to pull a different lever, walk over and no, that didn’t do it, back to the levers… I stopped playing after spending five minutes hiking around the level and swearing to myself “If that didn’t work, I’m not walking back again, I quit.”

    That was the big issue, but I also want to complain about the narrative. Myst had this sense of desolation to it that I really enjoyed. You were out there in this bizarre mystery world and you were alone. Not in a survival or horror way, more of a peaceful wilderness or abandoned city vibe. Obduction botches that very quickly by having you meet its version of Atrus who has a normal conversation with you and tells you what to do. You’re no longer a lone explorer, you’re Knockoff Atrus’ gofer, fixing the pipes because he’s too busy to do it. It’s a very Default Videogame thing to do, fetch quests on behalf of the NPC behind a glass wall, but Myst wasn’t beloved for being the Default Videogame.

    1. Paul Spooner says:

      Yeah, that really bugged me. That jerk in the bunker is a very specific level of “too busy to turn the power back on.”

      1. Higher_Peanut says:

        I started playing obduction and assumed he was some sort of bad guy immediately because he was such a weirdly specific level of busy to help himself. I figured he was tricking the newcomer and that there would be some sort of inevitable betrayal.

        Unfortunately I just sort of stopped playing in the 2nd/3rd “world” because it runs abysmally for some reason. It really didn’t help the loading screens and distant puzzles.

  8. Echo Tango says:

    The menu-transitions in Obduction seem like the most annoying ones, since I was last playing the XCOM remake / reboot games. I think they all took at least a second to play out, and there’s a lot of sub-menus in that game. Everything else in the game is snappy and responsive (placing buildings, ordering your troops to take actions, buying upgrades), but the menus and pseudo-menus, like moving from one “room” to another in your base, are all so slow. :(

  9. Echo Tango says:

    OK so…”Bob and Doug” has even more silliness for me as a Canadian, because it immediately made me think Elon had sent a couple of hosers up into space.

    1. Shamus says:

      THAT’S what I was trying to think of!

      I said “Bob and Doug” and then I hesitated. Wait? Isn’t that already a thing? Where have I heard that before? Is that a cartoon or something?

      Strange Brew was my first exposure to the characters. It wasn’t until 2 decades later that I found out about SCTV.

      Thank you for knocking that memory loose. That’s been bugging me since we recorded this.

      1. Echo Tango says:

        Yeah, I saw the movie in high-school around grade 9 or 10, and was a bit too young to see the TV skits. (At least, without cable TV.) Me and my friends developed the habit of calling each other hosers for like…a year or more. Good stuff. Many laughs. ^^;

  10. tomato says:

    F.E.A.R.’s AI is better on higher difficulty.

  11. Hector says:

    For a very different thing to discuss when talking about AI: Europa Universalis. (I’m discussing 4 here, but that’s just the latest and greatest). This game does a seemingly simple strategy AI, but it does so very very well and the game doesn’t really care much about difficulty modes. Yes, there is a rather silly challenge difficulty that hands out wild bonuses to the AI, but that’s not what almost anyone plays on except a handful of wild game experts.

    In fact, by the nature of the game AI bonuses don’t matter as much as it sounds. This is largely because the AI/s are working with you as much as against you. The game tracks and manages each and every state/kingdom/whatever, on each day, from 1444 to 1820 – which means hundreds of independent actors all trying to do their thing. Every country (or ruling dynasty) is trying to manage their economy, diplomacy, and military through a variety of policies. For the most part, the AI makes really good choices, although often fairly obvious ones. It’s always checking for a tiny advantage. That said, humans are much more organized and systemic than the AI and can beat it despite the AI’s ability to micromanage armies. Additionally, because it’s not “human vs AI” but “every actor working with and against you as their interests dictate” you can make long-term plans and alliances, and sometimes use diplomacy to destroy alliances that get in your way – but the AI is pretty good about playing that game too.

    The one thing the AI never seems to get right is how to use many of the special additions from expansions. They will destroy themselves (literally; some of these can wreck your entire game if misused) because the AI doesn’t understand the drawbacks.

    1. Awetugiw says:

      While I do like the approach Paradox take in EUIV (and most of their other games), it is something that only really works for their games. The AI in the most Paradox games is not good enough to procide a real challenge in a “fair fight”. They solve this by making most of their games very asymmetric. In EUIV you can play as any country, and some of them are much more powerful than others.

      So you typically choose the difficulty not by giving the AI some arbitrary bonuses* (although you can do that as well), but by choosing a harder or easier starting position. Looking for an easy game? Play the Ottomans or France. Looking for a hard game? Play Granada or Byzantium or something.

      But this strategy only works if you accept massivley unfair starting positions. Only one of the Paradox games, Stellaris, has all players start on more or less equal footing. And that game offers relatively little challenge, and is usually best played as more of a “role playing strategy” game.

      So I don’t really see a good way to adapt the Paradox way of having (mostly) non-cheating AI while stil keeping the game challenging to more symmetric games like civilization.

      *The AI in EUIV (and the other Paradox games) does cheat in some small ways, but they take an effort to hide it.

      1. Daimbert says:

        I’m not sure that it wouldn’t work. If you deliberately create civilizations that naturally have advantages or disadvantages and then let players decide between them, including which ones are in the game, then it might be a more natural way to handle difficulty levels. As you noted, play a weaker one if you want a challenge and a more dominant one if you want an easier game. If the differences make sense — either historically or through lore — and the bonuses are clear, it would definitely annoy players less than cheating AIs.

        Then again, for those sorts of games — mostly Hearts of Iron — I’m more interested in it as a sim than as a strategy game, and so the AI I’m most interested is one that can respond to different actions in a sensible way, not necessarily an ideal one. The opponents don’t have to be a challenge, but they at least have to react sensible to a changed “history”.

        1. Awetugiw says:

          I’m not going so far as to say that it’s impossible to make a Civ-like game challenging using asymmetry. But I do think the game you’d end up with is very different from civilization as it is currently.

          Right now there is already a noticable difference in power level between the different playable civs. But that difference is tiny compared to the difference between Settler difficulty and Deity difficult.

          1. Daimbert says:

            The idea would be to eliminate difficulty levels and instead allow difficulty to be set by starting civ, location and a host of other things. It would probably be different — requiring more set-up from the player, for example — but I think you could make a Civ-style game that did difficulty that way. It’ll come down to how different you’re willing to allow such a game to be before it’s too different.

            1. Mistwraithe says:

              Europa Universalis 4 is a bit like this. Yes, it does have difficulty levels which do make a definite difference, but many people would be happy just playing on the standard difficulty level and altering how hard the game is by choosing who they start as. There is a massive difference in difficulty between starting with France or England, vs Ulm or Byzantium, just because of the massively asymmetrical nature of the starting map.

        2. Retsam says:

          I’m not sure that it wouldn’t work. If you deliberately create civilizations that naturally have advantages or disadvantages and then let players decide between them, including which ones are in the game, then it might be a more natural way to handle difficulty levels.

          How is this different than how the game already works, with each Civ having unique abilities, units, and buildings that give it advantages and disadvantages compared to other Civs?

          I don’t feel you could introduce Paradox-style asymmetry in Civ without just making a completely different game.

          1. Daimbert says:

            As noted above, it will depend on what you mean by “completely different game”. The idea would be to drive the difficulty by starting set-up — civs, areas, etc — and then just play the game normally. So the idea would be that once the game was set-up it would play the same. About the only thing you’d lose is the random start, as it would generally be more useful to configure all the parameters instead of having random civs and starting points happen, or even the historically accurate one to some degree.

            1. Hector says:

              I like your ideas and want to subscribe to your newsletter.

      2. John says:

        Master of Orion 2 is an approximately symmetric 4X game. The AI may or may not cheat. I wouldn’t know. I suspect, however, that it doesn’t need to. The player can, at least in part, determine the difficulty of the game by choosing his faction or species. There are several pre-generated choices and they are not, as far as I can tell, particularly balanced. The player also has the option of designing his own custom species. Really hardcore players increase the challenging by designing terrible species for themselves.

        Not me though. Telepathic lithovores forever!

        1. Awetugiw says:

          That is indeed a possibility. I have never played MoO2, so I can’t really comment on how well this worked for the purpose of difficulty setting.

          Stellaris does also allow you to design your starting species, with some options being stronger or weaker than others. But my impression is that people are significantly less likely to design and play as a weak species than they are to play a weak nation in EUIV.

          I think that nerfing yourself by creating a bad species simply feels worse than playing a bad country, although opinions will obviously differ on the matter.

          1. Echo Tango says:

            MoO2 definitely gets easier with some traits and more difficult with others. (You can also adjust the real difficulty setting, if you need to.) Lithovore (don’t need farmers), cybernetics (need half as many farmers), anything touching science (both positive and negative), economy… I think there’s also some “noob-trap” traits that seem like they’d be useful, but actually don’t help because you can just build different ships or use different tactics to bypass them entirely. Ground-combat is pointless if you just bomb places from orbit, especially because the AI builds everything out of order[1]. I think ship-strength might be one too, since the economy, farming, science options all directly (lasers +2) or indirectly (build more ships) let you have a stronger fleet anyways. Spying too, until the late game?

            [1] Farms and factories come first then downstream items like spaceports, you ninny!

          2. John says:

            Master of Orion 2 was released in 1996. The people who are still playing it–apart from latecomers like myself who picked it up from GOG–are by definition or by attrition the hardest of hard core. So, in all fairness, I don’t know how common it was to play as a terrible species back in the game’s heyday.

            1. Decius says:

              High score runs, back when score was a thing, would win on the highest difficulty with a race that had the maximum score of disadvantages and no advantages.

              The spaceship design and combat system had a couple of win conditions; the cloak that made you untargetable if you didn’t attack last turn, the time-space dilator that gives you two turns per turn, and hyper-x capacitors, which let you fire energy weapons twice if you didn’t fire them last turn. The only hard counter to beat that one ship is with one that has the cloak but not the double-turns, since the cloaking field only works for 30 turns, and that ship has a hard counter in the form of the shield-piercing armor-bypassing weapons loadout, which is countered by having unpierceable shields.

              The MoO2 AI uses *one* of those design elements in ship production: It will sometimes use shield piercing weapons, but not with the armor-piercing Achilles Targeting Unit.

              While those designs are all fairly late-game, once you get them the conquest victory is assured.

          3. Sleeping Dragon says:

            Because Stellaris starts everybody on a roughly equal playing field (random circumstance, suitability for your species, advanced start and fallen empires aside) it is mostly devoid of the idea of “playing as the underdog” in the way EU games have it*. However, a lot of players tend to do randomized species** or roleplay to the point where they make suboptimal, sometimes suicidal, decisions for the sake of staying in character. Also, sometimes the mechanics of your species will be particularly incompatible with the situation you find yourself in (pacifist xenophile megacorp surrounded by advanced start fanatic purifiers for example) but there is little option of setting that up in advance and on purpose.

            *The new origins feature does have one that starts you as a vassal to a Fallen Empire which is the closest I can think of.

            **Not sure if there’s a mechanism preventing that but I have yet to see a completely broken species come out of the random generator, some are worse than others but I haven’t seen something outright unplayable.

        2. Echo Tango says:

          Cybernetic scientists or GTFO.

      3. Steve C says:

        Only one of the Paradox games, Stellaris, has all players start on more or less equal footing

        I find it surprising you use Stellaris as an example. The very reason why I ditched Stellaris was because that was not true at all. Stellaris was going to be my example of a bullshit difficulty.

        In my first game, my goal wasn’t to play the game but only to learn the mechanics. I deliberately put all my resources to accomplish a single thing. (I forget what.) It was an unbalanced and unsustainable strategy though. My plan was to do that one thing then quit and load back at the start. I discovered when I quit that the other players were very different than I expected. I expected I would be good (at least competitive) at the one thing I was doing, and the AI players would be better in every other way. Except I sucked compared to every other player in every way. Even the thing I was deliberately maximizing. There was no possible way I could do what the AI did.

        Either the AI was cheating or I did not understand the game mechanics in a fundamental way. Neither were acceptable to me. My only goal was to learn the game mechanics. If I couldn’t do that in the hours I had already spent then there was no point to playing the game.

        1. Decius says:

          No matter what you’re trying to do, the best way to do it is to improve your overall capacity, which means expanding.

          The way to get more science, production, metal, political favor, expansion, or any other thing is to have more star systems making it.

          Once you expand to the point that you need to attack other empires to expand more, then you can start to transition from expanding your capacity for expansion to your goals.

        2. Philadelphus says:

          Either the AI was cheating or I did not understand the game mechanics in a fundamental way.

          It depends; on the two lowest difficulty levels the AI gets no bonuses relative to humans at all (in fact human players get bonuses on the lowest difficulty level), though Advanced AI Starts is an option, where you can choose to have some fraction of the AI players start out more advanced than you are. I don’t know if you used this or what difficulty, obviously.

          For the second thing, it also depends on when you started playing Stellaris; in the Major Economy Overhaul of 2.2 in December 2018 the economy became more complicated, and a lot more interdependent, such that focusing exclusively on a single aspect is almost certainly not going to work compared to a more balanced approach, and in exactly the way you mentioned: that you’ll be worse off even in the thing you were trying to maximize compared to someone who is growing all parts of their economy more evenly.

          (Briefly, you have two primary raw resources, food and minerals, plus energy [i.e., money], but you also have three secondary resources [motes, crystals, and gasses] which are critical for upgrading buildings to get more jobs, and you also need alloys for building anything space-related. Minerals can be turned into those last four using a variety of refinery buildings, so you’ll need to watch your net flow of all of them and make sure you’re increasing your industrial refinery capacity to keep up with your growing economy. This also isn’t covering consumer goods, which are also created from minerals via various jobs and are necessary for keeping your people happy, or other, rarer resources which don’t tie directly into these.)

          So, yeah; leaving out Advanced AI Starts (which I always do), Stellaris empires do tend to have pretty even starts overall. (Although I’m not 100% sure that randomly-generated AI races always make maximally beneficial use of their trait points and some combinations of civics will necessarily turn out to be slightly better depending on circumstances.) I’ll say that the economy overhaul definitely made it more complicated, even for long-time players like myself at the time; I took a game or two after the update to really come to grips with the new system.

          1. Steve C says:

            Ok? Not sure why you responded with details like that. I was using it only as an example on the topic of difficulty in the Diecast. I hated the game and uninstalled it. An explanation about the mechanics is beside the point. I felt either the game was cheating, making it impossible to gauge my own progress properly, and therefore not worth playing. OR it was incapable of teaching its mechanics in a reasonable amount of time and effort, and therefore not worth playing. The key part was that it wasn’t for me.

            The point was a counter example to Awetugiw. I disagreed that Stellaris starts on more or less equal footing. And that Stellaris had “factors such as game mechanics, difficulty design, game length, pacing, weak story or writing, weak atmosphere, uninteresting setting etc.” (all of the above) that were beyond my personal difficulty threshold. Paradox publishes my favorite genre of games and I do not like the games they make at all. IE WaveofKittens’ question.

            1. Retsam says:

              Ok? Not sure why you responded with details like that.

              I think that was their polite way of saying “you fundamentally misunderstood the mechanics”. Stellaris does start its players on an equal footing.

              I’m not going to try to convince you to like a game that you dislike, but your bad experience with the game doesn’t change factual statements like “Stellaris starts players on a more-or-less equal footing”.

              1. Steve C says:

                Except there was also: “leaving out Advanced AI Starts” . Which is the very definition of an uneven footing. It is the very point. Those give all sorts of major bonus starting resources.

                But even putting that aside, every possible game difficulty gives bonuses. It is literally impossible for players to start on an equal footing. Either the human gets a bonus on the lowest difficulty, or the AI gets a bonus on every other difficulty. Your statement is factually incorrect. Stallaris only allows you to pick how unbalanced it is and in which ways.

                More importantly, it doesn’t matter. My goal (learning from AI players) was fundamentally invalidated by the game. I deleted Stellaris faster than Shamus trashed Dark Souls because that is not something I’m willing to put up. Which was WaveofKittens’ topic. I will not play games that both 1)cheat (either for or against) and 2) fail to teach mechanics. It can do one, the other, or neither. Never both. That’s my line in the sand. It makes me rage.

                1. Monteizo says:

                  Please read the link you provided. It clearly states that on ensign difficulty the player and the AI receives no bonuses. After all the Spaceborne Alien bonuses are completely and utterly irrelevant in this as that would be things like Void Clouds and Marauders and the other stuff like that, and not the other AI players.

                2. Retsam says:

                  Yes, the game starts on an even-footing: unless you specifically use the game configuration option that gives random enemy AIs an advantage. I kinda assumed that latter part went without saying.

                  And, as Monteizo pointed out, Ensign only gives bonuses to the “Civ-barbarian-like” aliens, it’s no bonuses to player or their AI equivalents, but even if it did give small bonuses to AI, I think that still qualifies as “more or less even footing”.

                  Especially given the original statement was in comparison to games like CKII or EUIV.

                  I don’t really care about your arbitrary lines in the sand or why you hate the game. Like I said in the last reply, I’m not trying to convince you to like the game.

                  1. Steve C says:

                    Retsam, we are talking at cross purposes here and it is more than a little frustrating.

                    You (and others) are talking about Stallaris. The original statement was not about that game nor was it about games like CKII or EUIV. The original statement was the very question WaveofKittens asked; A question which is about game factors that are beyond when a game isn’t fun. Problems in difficulty etc. And I keep bringing it back to that in every reply.

                    I don’t really care about your arbitrary lines in the sand or why you hate the game.

                    You might not care. But you’ve gone on a tangent and refuse to be budged off it. I feel like I’ve just said I don’t like Dark Souls because it makes me angry. Then a bunch of people gave me in depth information about Dark Souls. Please stop. The key part was “angry” not the name of the game. I don’t care at all about the specific game mechanics as it is beside the point. (I’m repeating that yet again.) The topic is literally about the line in the sand and where it lies:

                    WaveofKittens said: “ToEE is pretty much the threshold for me, with Dark Souls already being on the other side.”

                    WaveofKittens is defining a line. The topic is NOT about Stallaris. It is about where the line is and why. My line has to do with how well a game (any game!) explains the fundamental mechanics. If I’m saying I hate “Shoot Guy” because head shots do nothing, nothing is explained, and the bots are cheating, I do NOT want to be told that I should have known I had to shoot everyone in the dick and a detailed explanation of the AI of the bots. Because that just falls under “nothing is explained.” I’m completely indifferent where weak points are. The problem to me is that 1)It’s not intuitive and 2)It’s not explained.

                    People repeatedly telling me that I should have already known what esoteric settings to tweak in the menu before playing the game to overcome a fundamental teaching problem with a game… well that is going to fall on deaf ears. It is the very thing I do not tolerate in a game.

                    1. Hector says:

                      Paradox really does have a terrible track record for explaining mechanics. Unless you read years of forum posts many of the game systems are maddeningly opaque. I put up with that, but even to get my brother into the games I had to sit down and give him a 1-on-1 tutorial for hours. Sox games are very good, but it’s deeply unfair to ask this of new players.

            2. Sleeping Dragon says:

              Oh yeah, Paradox grand strategies are notorious for not teaching you how to play, to the point of not explaining at least some of both basic and crucial mechanics. I’d stay away from Europa Universalis and Crusader Kings series if I were you.

              1. Hector says:

                It’s true that you kind of need either a lot of time or to walk-through with another player, but they are fun (well, for some of us madmen). It’s complicated, but that’s not necessarily a bad thing.

                Plus, there’s something so very enjoyable about warping the entire fabric of history.

            3. Decius says:

              I think the general takeaway from the people who do understand the mechanics is that Stellaris failed to teach you those mechanics.

              However, many of us find that the basic mechanics are pretty simple and intuitive, in that everything tells you exactly what it does.

      4. Ninety-Three says:

        Only one of the Paradox games, Stellaris, has all players start on more or less equal footing. And that game offers relatively little challenge, and is usually best played as more of a “role playing strategy” game.

        Stellaris has difficulty settings. They pull the standard 4X move of giving the braindead AI a 100% boost to everything in order to make it threatening. It’s like Civilization which is also pretty easy if you play on the setting where the bots are on close to even mechanical footing with the player (weirdly there’s no difficulty that makes them exactly equivalent to a human).

  12. Hector says:

    I have a separate comment on the last Mailbag, when the subject of APB: All Points Bulletin came up. And weirdly enough, it shows Gamestop in a positive light*

    *Disclaimer: I don’t have quite the negative view towards Gamestop that many other people do. I find it’s just not a very well-run business instead of being heinously evil or whatever. That’s a long conversation and if anyone wants to have it we can discuss.

    Basically, right after the news broke that APB, which had just launched to a ton of marketing, was immediately going to be killed off at the source, I went down to buy a copy. Yes, I went to buy a copy AFTER the game was confirmed to be dead. Why? Well, thought it would be really cool to have an unopened boxed copy, like a kind of collector’s edition of something unusual and kind of cool. Heck, maybe I would buy two just to have one to open.

    However, Gamestop would not sell the game. They told me it was corporate-wide and even though they did have the copies in the back, everything had been pulled off the shelves and would not be sold anymore. Gamestop, the cruel corporate juggernaut of gaming retail, had too much ethics (or just didn’t want to get in trouble, your pick) to sell the game anymore as soon as they heard. I was a bit bummed I couldn’t get a copy though. The concept behind the game was a very good one, but even once they finally went FtP, it just wasn’t that well done.

    1. Sleeping Dragon says:

      I’m not saying this to diss on Gamestop, I don’t live in the part of the world where they have a physical presence and so my knowledge of them is largely second hand but I do need to point out this likely has nothing to do with ethics and more with legal obligations.

      1. Hector says:

        it was not clear to me then or now. As far as I know, however, a retailer has no obligations regarding manufacturer support of a product. My guess is, however, that some executive up the line decided that selling it would be a customer service nightmare and it just wasn’t worth it.

        1. Thomas says:

          Either way, it would be a good example of how in certain situations big businesses can be reliably ‘moral’ just because they have a lot to lose and everyone knows who they are.

  13. Sleeping Dragon says:

    With strategy games I’m usually not good enough for the AI difficulty cheats to matter but I do get annoyed when they blatantly ignore mechanics. Like, if I fight an empire that is big in terms of economy and I destroy their trade routes (or some equivalent) I’m doing this because they shouldn’t be able to maintain an army ten times of mine and will either need to disband at least some of it or collapse under the upkeep… but nope, nothing short of direct assaults against cities/bases/planets/whatever matters. I dislike it because it essentially forces a brute force approach and disallows smart combat.

    1. Joshua says:

      I think that this can be summed up in does the AI cheating make reasonable strategies useless? If you have to not only learn how the game affects you, but how it works differently for the AI, it can be rather frustrating. You spend all of this time learning how all of these different systems interact to improve your own performance, but then have to learn a somewhat different game just to use AI exploits.

      For example, maybe your issue of the computer getting to ignore maintenance costs invalidates a reasonable strategy, but then you have someone else using the famous Civilization 3 exploit of deliberately leaving a city defenseless, and when the AI sends all of its units that direction, popping a unit in and then leaving a different city defenseless on the opposite side of the empire.

      1. Hector says:

        This actually brings me to a point about Civ. The AI even on supposedly “honest”difficulty, cheats wildly. You’ll capture a capital city and discover it’s got practically no infrastructure built – certainly not enough to support its armies and research or be a threat. Even after having played countless hours of every civ game, I still am not sure how AI production works but it isn’t under the players’ rules.

        1. Joshua says:

          Which could affect decisions like, “Do I pillage all of the terrain improvements in their territory to sabotage their unit production?” If they don’t play by the same rules, aside from whatever amounts of gold or healing the particular Civ game you’re playing grants you, you’re just wasting your time. Should you use a precious World Congress resolution to suggest a boycott for a particular resource or Civilization if it’s irrelevant to the AI? Would the AI consider a 1 for 1 trade on a Luxury (say giving them gold for their only copy of ivory) if it means getting credit for a City-State quest?

          Having completely different rules for how the AI works makes understanding the game that much harder, especially as Sean says below.

          1. Sleeping Dragon says:

            Oh don’t get me started on diplomatic gameplay, I understand this is where things become really challenging to program because diplomacy can be anything but straightforward but AI tends to flounder so badly here. In my experience the AI is either a diplomatic pushover, happily willing to take scraps from your table for being your best friend and forgetting about the genocidal war between the two of your a few turns ago, or the opposite it’s been programmed as if to always assume the player is having some winnning strategy dependent on this exact diplomatic action (to be fair that sometimes is the case) and will resist anything less than trading it five star systems for a single wrench. I also wish the concept of trying to hobble the strongest guy wasn’t so alien to AI players though again I understand this is something that’s easily exploitable by players.

            1. Joshua says:

              Diplomacy is one of those things that it seems like it would be impossible to achieve parity on between human and AI players. For example, you’ll almost never see an AI trying to bribe the humans with gifts to make them like the AI better. Still, there are some things the Devs could do to make diplomacy more equitable.

              My largest complaint with diplomacy is the lack of information that would make it slightly more tolerable to the humans. What I’m specifically referring to are when the AI constantly makes requests or demands and the player isn’t able to pause and look up information before making a decision, as if the AI diplomat stormed into their office, is looming over their desk, and saying “I need an answer, yes or no, RIGHT NOW!”. In Civilization, before making a choice for your next research or cultural selection/civic, whatever, you can take time to review pertinent information before making your choice, but you don’t have as much information at your fingertips when an AI wants to make a deal with you. I’ve said previously that was what drove me from IV to V after getting repeated demands from AI to go to war, stop trading with someone, etc. without being able to look at relevant factors before making a decision, even something as basic as “What am I currently trading with X that would be lost if I agree to these demands?” Even V has its problems that could be remedied*. Not sure whether VI improved things or not.

              *One thing I would like is for deals with opponents (whether AI or Human) to automatically have a “Right of First Refusal” clause in them for when they expire. Renewing trade deals shouldn’t require micromanaging as much as it does.

        2. GoStu says:

          Short answer: they get *massive* discounts on everything. Production costs to build things, gold costs to buy things, science costs to unlock tech…

          They’re probably generating some fraction of what you do of all these resources, but they keep up on pure discount power alone. It doesn’t matter if they’re getting 60% of the production that you do if they only pay 40% of what it costs a player to build anything.

          What makes them possible to defeat is their ridiculously-poor decision making. They’ll attack into fortified choke points or just shuffle units around under fire; they’ll do a godawful job improving their lands at all (200+ turns into the game and Cattle by their capital city isn’t Improved), and they make silly research decisions.

          If I recall correctly most of the AIs in 5 are trying for a “cultural” victory… which gets incredibly difficult when multiple similar civs are going for it. Winning culturally involves producing a lot of tourism (and a byproduct of that is culture), and overwhelming the other civ’s own culture. If everyone’s trying to do this at once then all civs are going balls-out producing Culture of their own, and the key Wonders needed to generate lots & lots of Tourism end up divided between them, largely cancelling each other out.

          Meanwhile the War-Mongers have probably devoured their neighbor and paid heavily in technological progression to do so (hard to build libraries and universities when you spend all your production on crossbows and muskets) and then “burn out” into technological irrelevance. Annoying to spawn next to but utterly harmless if they’re across the map – they just munch down another competitor and then wallow in primitive tech and ludicrous Unhappiness penalties.

      2. GoStu says:

        I like the way you think.

        Having seen the massive difference between vs-AI Civilization and vs-Humans, it’s eye-opening. Strategies and advantages that are pivotal in multiplayer are wholly irrelevant in single-player and vice versa.

        There’s early strategies for multiplayer that begin with aggressive military production, and you use that strength to extract Tribute from city-states: that’s totally impossible against AIs on middling-high difficulties. There’s plans to build key early Wonders that just can’t happen in single-player because the AI will build the damn wonder several turns before you even begin researching the requisite tech. On the flip side, the AI pursues terrible strategies and will never seriously try to contest a lot of player strategies directly.

    2. Sean says:

      I was the (excessively verbose) author of the AI mailbag entry, and this is perhaps the greatest reason for my grievance.

      I’ve always hated cheating AI in strategy games, but I don’t think I’ve ever quite hated it as much as I have in the last year or so, when a group of my friends started playing Civ VI multiplayer. I’d played the game quite extensively single player, but I’d never really tried it with other humans (I very rarely do MP in general). I was quickly just kind of staggered how much more *rich* the game systems seemed, as if all of them seemed to suddenly start working as intended. Clever plays and use of previously inconsequential mechanics suddenly became a thing. Since that experience, I haven’t been able to enjoy games against the AI anymore, and my contempt for just how much they warp the game has tripled down.

      1. Sleeping Dragon says:

        Which leads to an interesting question, how extensive the testing of the game was in multiplayer VS testing against AI? I mean, I do not expect the AI to be able to play at the level players do, especially not to be able to act out some emergent or metamechanical strategies, and I understand the concept of giving it certain numerical bonuses to compensate. But if for the AI can’t be fully interacting with the game mechanics at least stretch the illusion so it gives an accurate impression of doing so rather than laughing in your face.

  14. King Marth says:

    The AI in Fire Emblem is interesting. It’s laughably predictable (enemy units will always attack if there’s an option, preferring encounters where they aren’t counterattacked but with no regard for whether they’ll win or even inflict any damage), which is an essential part of the challenge as the difficulty comes from correctly using your tiny customized force to wipe out a massive army of faceless mooks. All information required to assess combats is freely available, though you only get a nice combat prediction view when directly proposing a battle; if you’re parking a unit in enemy attack range to shred through them with counterattacks, then it’s up to you to make sure that every one of those fights is winnable. Failed predictions are usually a matter of “I had everything I needed to avoid that and messed up, oh well” until the RNG decides to kill you anyway (often by your unit performing too well, wiping out every attacker and thus leaving tile space for more attackers to whittle them down).

    Special mention to the mobile game Fire Emblem Heroes, in which the AI is surprisingly smart at using the various alternate movement options. When units move 2 tiles a turn (3 for cavalry, 1 for armor) then any extra movement can radically change how the battlefield evolves; also, since there’s no randomness in combat, any mistakes truly are due to analysis rather than RNG and leave you with an appreciation for how you got outplayed. Fortunately the AI hasn’t figured out how to use Draw Away and Reposition offensively; the primary use of these skills by players is to move your units closer to the enemy before they activate in order to move further than expected, while the designer-intent-as-expressed-by-AI is to use these skills on units that already attacked in order to remove them from the front line.

    1. Thomas says:

      There’s a lot to be said for AI which is predictable. If the AI operates according to a few simple, easy to grok principles, then out-manouvering the AI can be part of the game and balanced for.

      1. Echo Tango says:

        That’s pretty much how “AI” works in every stealth game. :)

      2. Joshua says:

        In Civilization, I sometimes use Workers as bait because the AI can’t resist going after them as opposed to more strategically important targets. :)

    2. ivan says:

      At least for the FE games on the GBA, you’re wrong about what damage they’ll deal not being a factor. The AI will attack if they can, yes, but if they have multiple options, one factor that is high in priority for who to attack, is how much damage they can deal to those targets. As in, the raw damage number a hit would deal, notably with no significant weight going to the likelihood of said hit. So, a target that would take 15 damage per hit, with a 1% chance to hit, will be attacked in favour of a target that would take 5 damage, with 100% chance to hit.

    3. Fizban says:

      In Fire Emblem (Fates): Conquest, IIRC the AI just straight refused to take fights that would result in death on Normal, which made a heck of a change to battle plans and difficulty coming from previous games. Conquest is the hardcore route, so that was a nice touch. But the other two Fates games didn’t do that, at least on normal, making for a similarly jarring return to easy mode. Dunno if playing on Hard makes it apply to all of them or what the newest game does.

  15. Hal says:

    Terry McAuliffe was chair of the DNC and then governor of Virginia. It’s not surprising that the names would co-mingle.

  16. Alex Alda says:

    Podcasts are not my thing (I can’t concentrate on one unless I pay attention to ONLY the podcast, so I can’t listen to them while driving or working, also I’m totally crap at telling apart voices so it’s always a headache to keep track of who’s talking), so it always bums me out there is no full transcript of Diecast.
    I’d love to know what you answered to the question about memorable game AI.

    1. tmtvl says:

      Paul believes a way to solve it is to have an AI on your team, like in Planetary Annihilation, where in multiplayer you can have an AI commander on your team, and in singleplayer you can have subcommanders controlled by AI on your team. That allows your units to do stuff without you having to micromanaging them all the time.

      Shamus considers the F.E.A.R. (First Encounter Assault Recon) AI memorable, but mentions that most of the intelligence people ascribe to it is smoke and mirrors where the reason it seems smart is due to the level designers and animators allowing the AI (which simply tries to flank or get around you) to look smart while just following simple rules.

      They then talk a bit about how AI difficulty generally works, Shamus mentions how an AI in a shooter is unbeatable until the programmer tells it to miss 80% of the time. Paul mentions how in PUBG for a human it can be difficult to see another player hiding in the terrain, but an AI usually has X-ray vision, and the range can be limited, but then it becomes a question of where to set the limit. Next Shamus talks about how the AI in Civ doesn’t play the same game as the human, and Paul chimes in that it depends on the difficulty (he thinks it’s Prince where it’s fair).

      Finally they talk about Noita and how it would be neat if instead of shooting straight at the player, an enemy would be intelligent enough to shoot at environmental hazards near the player to dump, say, a poison reservoir on them. Paul mentions how that would be neat the first three times or so, but then the player would start draining the poison reservoirs and stuff like that, so it would be tedious.

      And that’s about it.

      1. Paul Spooner says:

        Fair summary, though I think it was Shamus that brought up the tedious downside of AI that takes advantage of the physics environment (48:21).

  17. tmtvl says:

    How are we talking about AI without mentioning Galactic Civilizations?

    Like, PC Gamer had a series on GalCiv2 where it turned out one of the AI was keeping the player weak but not destroyed because otherwise another AI player would have won.

    But it gets better, the warmongering AI was planning on backstabbing the diplomatic AI, and the diplomatic AI was preemptively researching defensive technologies that countered the warmongers’ weapons.

    Now that is some good AI.

    1. Hector says:

      However, IIRC, the developer themselves said that was a bug or other issue. The human player was the one ascribing brilliant planning and human motives to what was, in reality, a mistake.

      1. Sleeping Dragon says:

        reads tmtvl’s comment: Oooh, this sounds really cool, this is what I was talking about when I said I wish AI was better at diplomatic gameplay!

        reads Hector’s comment: Oh, nevermind.

        1. Liam says:

          It sounds a bit like this article from Shawn Hargraves (Ex MotoGP, Microsoft XNA)


  18. Grimwear says:

    Company sabotage is pretty common and it always just feels bad. I’m not sure of the specifics behind the Take 2 one but I remember growing up the parents of one of the kids my sister went to school with owned a small chocolate making business. They got a huge contract from…I believe it was Costco though I may be wrong on that. Unfortunately they couldn’t meet the deadline, lost the contract and lost their entire business. I also recently was reading a story about Nick Frost (of Shawn of the Dead fame) growing up and his parents had their own furniture assembly business. Company came in with a huge order, order couldn’t be fulfilled, contract was lost and they lost their business. It seems like it’s a common tactic for large businesses to put in large orders that can’t be filled in order to bankrupt small businesses. Maybe they see them as potential competition down the line? Not sure but when you’re running a family business with 2-4 employees I guess you don’t have the time or resources to evaluate a contract to make sure you’re protected. Even though they should know better, they see a large contract and don’t see any nefarious meaning behind it.

    1. Steve C says:

      The strategy is not to bankrupt the small company. The idea is for the large company to put themselves into a superior bargaining position.

      A small business that expands in order to fill a contract now has to keep that contract. They have to due to the extra costs associated with the previous expansion. That’s the problem when one company is a disproportionately large amount of your revenue stream. The same thing applies to municipalities. Once a company knows that if they walk, the town or business collapses then they start dictating terms. Bankrupting a business or destroying a town is just a side effect.

      1. Bubble181 says:

        Depending too much on one customer can be an absolute death knell for even large companies, and some will go out of their way to avoid it.
        I work for a certain logistics provider/delivery company. I’m probably better off not naming names, but it’s got a purple-and-orange logo with an arrow between the E and the x. Anyway, we handled all shipping for Amazon…Until they grew to be so big that even though we were already losing money on every parcel delivered, they wanted to push prices even lower. Yes, volume/bulk allows you to keep lanes open that would otherwise be unused or unsustainable, bu at a certain point it’s just too much of a drain. So the contract was cancelled, and now our main competitor is delivering for Amazon…And while that certainly means they’ve grown more than we have in the short term, they’re slowly killing themselves from the inside for a company that more-than-likely will become a competitor with their own delivery systems in the near future.

        1. Hector says:

          I work in logistics and I have an interest in economics. As part of that I’m often interested in cost vs. price. Amazon fascinates me, especially due to thinking about whether they can run their shopping that cheap forever.

          1. Bubble181 says:

            As long as they could be considered a loss leader to attract other business and keep lines open, sure. With volumes growing too large and necessitating separate lines, no.
            Unless their warehouses go even more automated, there *is* a lower limit to how much costs you can cut for storage/order picking/shipping/delivery.

  19. evilmrhenry says:

    I agree on that rotating puzzle as being “that one puzzle”. I’ll note that the area you teleport to is a separate area from the rest of the starting world, (like the under-tree area) so they were aware that loading screens were a problem, they just needed better technology. (Also on a SSD here.)

    I did end up beating it; once you’re past the rotating puzzle, you’re past the worst of it.

  20. Redrock says:

    We had an interesting experience with Obduction at home: I’m not really into puzzle-heavy adventure games, but my girlfriend is, so we gave it a go. I think it suffers greatly from having too much story for a game in this genre, and that’s not something I say lightly – usually I prefer my games to have as much story as possible. To give an example, I’d take Observation over Obduction any time. But what really made us give up on the game were the bugs. In particular, there was some sort of garage/autoshop where I somehow just managed to reliably glitch through the garage door. And that’s when I was done. Obscure bactracking puzzles try my patience enough as it is, but not being able to tell whether or not I’m even supposed to be at a given location at a certain time elevates the whole thing from “tedious” to “unbearable”. Which is a shame, because the world did seem intriguing.

  21. Chris says:

    If i recall right companies spam C&Ds because if they dont there is legal grounds for any company to copy their IP. So if you dont C&D some fan project, another company can point at them and say “well company X didnt sue timmy and friends, so why are we being targeted”. Although I have no legal experience and this is just something i heard.

    1. Chad Miller says:

      If i recall right companies spam C&Ds because if they dont there is legal grounds for any company to copy their IP.

      This is a trademark concept that doesn’t apply to copyright. If this were true, virtually every fanfic site would have been DMCA’d out of existence by now; it’s common practice for big content companies to look the other way on small benign violations and save the legal action for enterprises that actually threaten their business model.

      1. Joshua says:

        Yep, the classic cases being Aspirin, Band-Aid, and Kleenex. (Just did a quick Google search, and interestingly enough “App Store” was another case I hadn’t previously heard of.

        1. Philadelphus says:

          How much you wanna bet that Zoom becomes a generic brand-name term for “video-call”?

          1. Joshua says:

            Good point. I’m organizing a D&D 5E game on Roll20 now, and one of the players asked if we were Zooming it.

  22. Ninety-Three says:

    This is a random test of comment formatting, don't mind me.

    Or is it?

    Oh look, undocumented features!

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="http://en.wikipedia.org/wiki/Darth_Vader">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.