Game Design (or related) Articles

Discussion in 'Game Design' started by keithburgun, Mar 20, 2014.

  1. Batlad

    Batlad Well-Known Member

  2. Disquisitor Sam

    Disquisitor Sam Well-Known Member

    Man, I can just hear past-me screaming at him for suggesting that roguelikes, the sole remaining hold-out in a sea of casual games, might be BAD because they make him FEEL bad for having to do things OVER AND OVER. Filthy casual.

    Phew, good thing that asshole isn't around anymore.

    Seriously though, it's like this guy wrote his article just begging to read mine. My article has both that same problem statement, and a ready-made solution. :auro:
     
    Nachtfischer and Nomorebirds like this.
  3. ALavaVatChild

    ALavaVatChild Well-Known Member

    Classical level design always strove to give you feelings of having mastered previous 'lessons', with Nintendo setting the gold standard decades ago. It's still a lot of work to do this well across a large amount of unique content though, and the procedural permadeath model gets close enough for most players (at least in the short term).
     
  4. richy

    richy Well-Known Member

    I thought at the time your article didn't really make its case, and still don't. A lot of time is spent praising the Auro system where the game determines appropriate difficulty with its black box ranking algorithm, but then you finish off with "...enjoy games on your own terms, and you’ll have a much better time playing them."

    Playing on their own terms would mean players choosing difficulty themselves. What you're actually advocating is playing the game on the designer's terms. I do remember in some discussion on here around the time of your article, Keith agreed that was the idea, i.e. players cannot be trusted to enjoy the game on their own terms. It's a valid point of view but the article should be honest that's what it's saying.
     
  5. Nachtfischer

    Nachtfischer Well-Known Member

    I don't think so. It means you should just play as yourself. Not as the "hardcore" or "casual" version of yourself because there's really no such thing. Don't even think about these two modes of existence. Just play the game and it'll adjust to you to give you the best possible experience. Don't adjust yourself to the game though.
     
  6. richy

    richy Well-Known Member

    Re. debunking the hardcore/casual war I guess what Sam said is probably right, I have no comment really as I never knew there was such a war. What I'm disagreeing with is that it's definitely a good thing if a game adjusts the experience according to what it, or the designer, thinks is best for the player. I don't think the case was made for that at all. Actually in Auro's case I disagree on two counts, one that I don't think the game's decision of the "best experience" can ever suit all, or even many players, and two that I don't like the way Auro doesn't tell precisely how it adjusts the handicap.
     
  7. Disquisitor Sam

    Disquisitor Sam Well-Known Member

    Well no, in neither case of rogue-like mega games and Auro ranking system is the player choosing their difficulty. The primary point I wanted to make is that one wastes time in a way the other doesn't.

    The whole "play games on your own terms" bit is a separate thought - it's an appeal to people to ditch the idea that wasting time makes you a more worthy gamer. Don't play to posture an unhealthy, made-up identity that isn't even your own. That's really the only thing keeping long-form roguelikes alive: a narrative that pouring time into games indiscriminately is proof of your gamerness.

    I didn't really touch on choose-your-own-difficulty in the article other than to say doing so by score STILL elongates the game unnecessarily. If you want to choose your rank in Auro, I'd still call that an improvement over long-form roguelikes that MAKE you play EVERY difficulty every game.
     
    valavir, cpudreams, richy and 2 others like this.
  8. Jon Perry

    Jon Perry Well-Known Member

    The view that "output randomness is just input randomness for the next turn" (a view I hold) is not at all about looking at a single move. In fact it's the exact opposite. It's a holistic view. Which is why it specifically takes into account not just this turn but future turns as well. I understand the larger point you're making, but you could not have picked a more ill-fitting example.

    "Too-focused-on-the-current-move" is exactly the criticism I would level at the opposing view in this case.
     
  9. keithburgun

    keithburgun Administrator, Lead Designer Staff Member

    I can see how you might be arguing "no no, I'm looking at two moves - the move I just made, and the move I'm making right now." And that is indeed larger than looking at just one move. But my position is that we should be looking at all the moves and how they come together to deliver a final outcome that is meaningful and attached as much as possible to your inputs. Output randomess - even if next turn you can say 'it's just input randomness for me now!' - is severing that causal chain of events and making the final outcome (win/loss) less tied to your agency and more tied to random events.
     
  10. richy

    richy Well-Known Member

    You both seem to be arguing the same thing! Jon's recent thread was about finding ways to look at random events in the context of the entire structure of a match, because the binary input/output distinction seemed too stark and/or didn't seem to tell the whole story.
     
  11. keithburgun

    keithburgun Administrator, Lead Designer Staff Member

    I think this low level / high level distinction you laid out is along a different axis than what I'm talking about. I'm just talking about, from the player's perspective, at the end of a match, how easy is it for them to connect the Win/Loss value to various choices they made throughout the game? If that linkage is strong, then they are getting good feedback - learning a lot per each win. If that linkage is weak, they're learning less. The closer the information horizon (the more output/output-ish), the worse that linkage will be.
     
  12. Bucky

    Bucky Well-Known Member

    Is the "at the end of the match" important?

    Say I'm playing Puzzle Strike. I opt to buy a combine instead of money on turn 1. I win on turn 8 with a 50-50 3 crash that probably loses the game if it gets countercrashed. I think at the end of the game that buying the combine was absolutely necessary to win, since I wouldn't have the 3-gem to crash without it, and that the decision to crash was a good one.

    Inspection of the replay shows that if I'd bought money instead, I would have been able to safely rush a Double Crash Gem and probably wouldn't need to bet the game on a single crash.

    Did the game give good feedback?
     
  13. keithburgun

    keithburgun Administrator, Lead Designer Staff Member

    Yes, because the end of the match dictates the completion of the match structure.

    Assuming the outcome wasn't based on random events, YES. Because you employed a strategy, it worked, and so that informs something about the value of your moves that game. If you're sort of implying that there was some randomness there, then it may be giving false feedback, as random games often do.
     
  14. Bucky

    Bucky Well-Known Member

    There was also the point that I didn't get the correct feedback at the end of the game, I got it when I looked through the replay.
     
  15. Redless

    Redless Well-Known Member

    I understand that basing the validity of moves on their effects on chance of winning is the only sensical way to judge moves, but I don't think that only factoring in whether you won or lost games where you used the move makes much sense. In practice people seem to think a lot about what the move means in the context of the game, to come up with a better estimate of how good the move is, rather than just looking at if it won or lost.
     
  16. Nachtfischer

    Nachtfischer Well-Known Member

    Guys, nobody is suggesting that whether you won or lost is the only thing that matters and that winning means "You couldn't have done anything better and made no mistakes". That's not even a position worth arguing against.
     
    keithburgun likes this.
  17. Redless

    Redless Well-Known Member

    The thing I think people are suggesting is that we should use only empirical results in the game to determine how good a strategy is. If there were an Auro spell that was "you win 99/100 times you use this and lose 1/100 times" and I got the spell for the first time and lost by using it, I would still think it was a good spell, but those who think we should only use empirical results to determine how good a strategy is seem to not.
     
  18. keithburgun

    keithburgun Administrator, Lead Designer Staff Member

    @Redless the key detail your example is that there isn't anything to learn. Obviously if the spell is "99/100" times, you don't even need to use it once to know that it's a kick-ass strategy to use it. It's an absurd example that has no resemblance to strategies in un-solved games. There is nothing like a "win almost all the time" move in unsolved strategy games. By their definition, the win-rate of various strategies or moves is unknown.

    Take an unsolved game. A player employs some strategy. In order to get feedback for their strategy, it needs to get to the end, because the purpose of the strategy was to try to win. So yeah, you start with this tiny piece of information - "I won" - but then you sort of use that bit of information, thinking back to the entire match and all of the things you did in it with that positive "win" charge in mind. It doesn't mean that any individual move has some ridiculous thing like 99% chance to win. It's a very subtle nudging around of the viability of hundreds of positions and possible game states, a slight tweaking of your mental model of how the game works.
     
  19. Redless

    Redless Well-Known Member

    How would you explain how a player who has never played a game before can determine if a strategy is good or not?
     
  20. keithburgun

    keithburgun Administrator, Lead Designer Staff Member

    In an unsolved game, I don't think they can, other than by some kind of really really unlikely lucky guess maybe?
     

Share This Page