Sunday 27 March 2016

Still working on Imbroglio! This is maybe the most technically complex game I've made? Not sure, Vertex Dispenser's somewhat-broken online multiplayer was a whole lot of mess, but here all the complications are just in the central game itself. There's a whole bunch of cards that can interact in different ways, triggering each other and modifying each other's effects. Just on the programming side it's been heaps of work making all the effects actually work but then on top of that is the design complexity; figuring how all the combinations ought to work, keeping in mind all the different interactions that might conflict with each other. I'll be out somewhere and suddenly realise "hey you can get infinite hit points" and then have to work out whether that's okay.

Its big weakness is still what I was talking about before, a lack of clarity in the risk/reward structure. It doesn't do a great job of generating counterfactuals, letting you reflect back on your failure and see "if I'd done Y instead of X then Z would have happened instead..". Adam calls this "unpacking" in this talk and cites 868-HACK as a positive example. Maybe my problem is trying to not make the exact same thing again so deliberately not reusing the things that worked well before? 868-HACK had super clear risk/reward, directly tying enemy generation to acquisition, if you lost it's probably because you took a deliberate risk. Imbroglio loss-reasons are not obvious at all, enemies just keep coming on their own schedule so maybe it's because you spent some turns inefficiently and now your power level hasn't kept up with the opposition, maybe you made the wrong choice what to level up fifty turns ago, things like that.

But also it has deck construction so even though it can be hard to figure out tactical errors you can also just think about trying a different deck. And honestly this is kind of a problem Magic has too? It's easy to change your deck (maybe buy some more rare cards lol) but it's pretty hard to tell when you lost through subtly inefficient resource management over several turns. So maybe i'm not doing too badly if my game's biggest weakness is also present in the most successful game of its genre?

Anyway yeah I don't think this is a fatal weakness? I will just make the game and let it be what it is, I shouldn't be worrying about whether it's a less good game than something else I made or whatever.

Been making good progress this week. Took me a long time to figure out what to do for sound effects but I've got the right idea now and that's coming together well. Turns out being in a bit of a daze while recovering from having a wisdom tooth out is a good state for just sitting down and churning out lots of samples. Fingers crossed they still sound good when I'm back to normal. Also I keep getting sidetracked trying out weird new card ideas, I'm supposed to be working on essential stuff so I can release the game not making it bigger. I tell myself it is to some extent useful to plan out the ways I might want to expand it later so I can architect things to allow for them. Anyway it has room to grow.


  1. (I realized after writing that resources in Magic aren't just the mana, but essentially every card you play. Maybe you can coax some benefit from my sleepy rambling.)

    I disagree that it's difficult to tell in Magic why you lost or won. I'd reckon the average game of Magic is over (or decided) within 10 turns. The main resource is also very "big"; one additional available mana makes a large difference, and they're always in view.

    Point being, you're always very aware of your resources, and how they were used. Utilizing the mana at your disposal efficiently is the crux of the game. Some of the most effective strategies in the game is to circumvent the use of mana (Quicksilver Amulet), or speed up the gain of the resource, but that's a different topic.

    After a game, unless it was a long one, it's usually possible to identify moves you (or the opponent) made that decided the game. In my experience, there are two main types of Magic games:
    1. Steady buildup of board-presence that overpowers the opponent
    2. A tie that is broken with a bomb

    In the first, you can identify errors in your playing if there were ways to get a bigger presence earlier if you'd used your resources differently.

    In the second, it's about the timing of the bomb-drop. Could I have won sooner if I dropped it earlier?

    Sometimes you just get screwed by the draw, not getting mana or one of the core cards you needed. Hearthstone fixed the randomness of getting mana, and not getting the core cards in your hand is a bit of luck but you can also do a redraw to mitigate it.

    1. yeah i was thinking of things like when you should have kept a spell in your hand rather than playing it at the first opportunity, when you should have attacked on an earlier turn to trade hp rather than holding creatures back to defend. & not saying you can't ever identify errors, just that there are subtle ones that are hard to identify

      it's exacerbated in imbroglio though because it's a high-score game; you will always lose eventually, efficient play just staves that off. whereas in a beginner magic game you just have to be more efficient than your opponent..

  2. > Enemies just keep coming on their own schedule so maybe it's because you spent some turns inefficiently.

    Do you show the schedule? Not necessarily the whole thing, but a "Next enemy in 5 turns" kind of thing, or maybe "Enemies start spawning faster in 5 turns" depending on your system. Just having it on the screen says "Pay attention to resource management, or this will kill you." Especially if your schedule is chunky, so people are anticipating things, so they apply emotional weight to that thing they're supposed to be paying attention to.

    > It's a high-score game; you will always lose eventually, efficient play just staves that off.

    Named benchmarks can help with that. Their existence implies that the benchmark is possible. You could be boring and do that with achievements, but it could also be content. Your enemy schedule, especially. If, when you die, it says "If you lasted another 40 turns you'd have seen a dragon," then it implies that extra 40 turns was possible and that you should care. When you run out of content carrots you could switch to comparing to the leaderboard.