Clockwork Game Design: Episode 8 – Building on Structure


This episode is about how we can and should build on the structure that we have, rather than introducing non-sequiturs or removing elements from play. Interestingly, this podcast itself ended up having a bit of a non-sequitur of its own, as I talk a bit about parting ways with my entire huge video game collection and intellectual flexibility, which isn’t actually too related to the rest of the episode. Just so I’m clear: that’s not ideal podcast design!

I also discuss Matt Stone and Trey Parker’s talk on “therefore” in storytelling.

Also relevant is episode 3 of 3 Minute Game Design.

I’ve been a bit sick the last week or so; hopefully I was good enough with the editing to hide that, but that’s also my explanation for the sidetrack.

As always, please visit the show by becoming a patron at

  • Jeff S

    You lost me on this one. I think I was with you until about the 15-20 minute mark.

    I agree with what you’re saying about how other creative arts have structures (or fundamentals) and how video games tend not to. I like your ideas and how you approach things and now I am an easier way of saying why I don’t like certain games at least and sometimes what toys, puzzles, contests, and games are missing or adding. I also agree with your post-modernism rant.

    However, I was definitely lost in the Smash bros section. You talk about how the stock system (which there is a time battle in n64 there is also a coin battle in Wii which for some reason I never tried) and how it’s not necessarily problematic but can be improved upon. Although, my hopes are anything could be improved upon, I don’t know why you’d choose to pick on this element when, from what I understand, it agrees with your theory. I think later you go on to say that Smash Bros is contest-y but on your Youtube game design video you have Smash Bros under “Game” form.

    If I understand you correctly, you are seeking a continuous line from beginning to end that choices in the beginning have meaning all the way to the end. Yet, you say that there is a restart in the match. I would point out that most times the person who was a stock ahead manages to get a few percentage points (if not a life) on the person who is stock behind. That still yields your continuous line goal. The +stock person should change his strategy to a more defensive one, but it’s still up to them on how they want to play it or see an opening etc.

    So the +stock -stock difference actually supports Game theory. The goal of me pushing someone off the platform is obfuscated by me not know how the player will play. By me pushing him off, I have gained some understanding of the scenario (he only likes grabbing) and it is represented by the +stock -stock differential (my strategy was better). If the -stock player manages to break even, he was able to beat my defensive (or whatever) strategy. Now it’s at this “restart” point players have to have the awareness to change strategies, exploit the previous faults in the person tactic, previous bad habits, previous patterns, etc. If someone were very analytical they could chart the evolution of battle strategy (what I am assuming is your continuous line). I think this is what chess masters do, after their games, with the press. Surely a chess match piece wise could go down, up and back down again. Why is that not the same in smash bros?

    I don’t see how any of these examples are deleting or getting rid of information and reseting to the zero point when there is a lead change. Are you saying because you go back to the “beginning” a strategy that previously did work should not work again? Don’t the numerous amount of strategies you have available to you obfuscate 100% solve-ability? Don’t you want to repeat certain moves because you know they work?

    The percentage you accumulated or give to an opponent is as “causal, as reason based, as decision based” (your words) as your asking for. Most people don’t play Samus’ Metriod level in the n64 version because the lava steals kills or saves people, hides items at inopportune times and isn’t player caused. I see adding something to the map as still being execution based as opposed to decision based.

    I don’t know what you’re suggesting. I am curious if you’re just not happy with stock idea anymore and you just want a different one. Or are you saying “I want my non-random decisions I make at the beginning of the game to have the same potential for success as the traditional decisions?”

  • Technically, you’re right that “a few percentage points” is a SLIGHT, TINY piece of information getting passed to that new match, but it’s not nearly enough. It’s *mostly* a reset. In general the concept of “actor removal” tends to have this problem; an actor is moving around and has inertia and damage and position and then woop, they are removed from play and “respawned”. I really think that model needs to be seriously looked at, the “death/respawn” concept.

    To answer your last question, yeah, the stock idea is lazy and weak. Ideal Smash Bros would have way more large-arc things that fill out the entire match and it wouldn’t just be this contest-like linear measurement thing like how stocks are.

  • Jeff S

    In chess, “actor removal” is still there. If you don’t remove the actor, he has the potential to kill you (same as Smash). In Chess, if you remove the actor, you could be at a better or worse position. In Smash bros, if you take a costly route to removing another player (walking through projectiles), it’s the same.

    The large arc things are there. Just one is on a board and the other is in a player’s head. If you are saying you want all the decisions in a game to be on the game and not in someones head, then I am struggling to see how you’re going to make that while not making it:

    1. turn-based: If you already have execution then you’ll need to have speed (execution) to insert your bumper on the map like you were suggesting.
    2. have so many variations that it’s solve-able but not by any computer today (like chess)- because if you’re making it decision based you have to hide the solution somehow by making it unsolvable. — Maybe the reason why checkers was solved was because you can go backwards in chess (ex :placing your knights back) but can’t in checkers.
    3. not completely reliant on the one character on the screen. Chess and RTS’s have people dying all the time but those character removals have less of an impact on the endgame yet still not meaningless.

    The reason to make Smash execution based is to solve 1 and 2. Now, it’s non-turn based and the variations are dependent player creativity/ability.

    Either way. If you make it turn-based, the game will be limited by it’s solvability. If you make it execution based, it will be determined by player limitations and solvability.

    Maybe there aren’t any “Games” there are only easily measured Contests and Contests where we will never know or only know for the moment.

    Am I missing something? If I have misunderstood you, please clarify or if there is an example you can give me, I’d really appreciate it. This is an interesting topic.

  • The equivalent to “stock” for chess would be that if I kill your King, we put all the pieces back into their starting positions (except the killing player can choose 1 unit to keep where it is, maybe) and continue playing. Smash Bros stocks are basically “complete games” and if we play to 100 stock, we’re playing “100 matches of SB” (roughly).

    Also if you’re saying that a 10-stock match is a “single match of smash bros” and so is a 1 stock and so is a 100 stock – the fact that it can scale up and down like that endlessly in both directions should be a sign to you that there are NOT in fact large arcs happening.

    I’m afraid I do not understand your 1, 2, 3 points. My guess is that you should read this article as my response, though:

  • Jeff S

    I see what you’re getting at with Smash. I align with the 5 stock match is a match. And I believe emotions and strategies along with percentage carries over from life to life. I guess under your definition, you’d say basketball, football, tennis are also weak forms of games because score changes can happen and strategies, emotions, and stamina have nothing to do with games.

    On top of that:

    After reading that article, I think, it’s evident that there isn’t such thing as “Game” form.

    “What we should be doing with our deterministic games is giving players just enough deterministic space in front of them so that they both cannot solve it and also hopefully have numerous attractive-looking options. There should be just enough complexity here to allow for real, creative problem solving.”

    Un-Deterministic=Toy. Deterministic = Puzzle or Contest.

    -After you play enough and understand, the attractive looking options you’ll logically pick the best looking one fairly consistently.

    -If all the attractive looking options are equal, you might as well say it’s random or doesn’t matter.

    -If you say picking the best option is contextual (and the game is capped), then the game is a. solvable and/or b. the matches are random and without continuous learning.


    1. Making a game turn-based makes it either a. Solveable or B. (with fog of war or whatever) Random. So it’s either a Puzzle or random. A 5×5 hexagonal grid in one snap shot proposes a problem. If you insert monsters randomly (or so it seems with fog of war). It’s not necessarily a game but a random puzzle. Your strategy in the previous turn won’t work anymore, but now you have a new snap shot.

    2. Adding execution in a game makes it a Contest. If you add fog of war it’s random again until you find the regular best answer. Then it’s somewhat a Contest but a lesser version of itself.

    3. Adding an exponentially large decision tree is about memorizing or even execution. So it’s back to a puzzle or a contest.

    4. Or you put a goal on it (making it deterministic as you say)… but you can never reach the goal (undeterministic). Then it’s a Toy right? Because you don’t have a goal but you do… but not really.

    5. With multiple ways to reach a goal contest gets closer and closer to being solved until it’s more or less random. Ultimately it’s like Team Fortress (with only 2 guns being played) or Rock, Paper, Scissors (where it’s pretty much more or less random and contextual) or Smash (Falco, Fox, JigglyPuff, Peach). At that point it’s a contest… but a lesser version of what it once was.

    —-And if it’s competitive and it’s one player with multiple goals. You solve the Puzzle. And if for some weird reason you want to solve it again then cool you can solve it again but it’s not a Game it’s a Puzzle.

    So, I guess I should tie this back into Structure and you mainly saying you want to make better Games. Adding a longer arc to the game, might make the game more interesting for the moment, but it will suffer the same effect of point number 5 or should we call it the League of Legends Or Magic the Gathering effect where you’re continuously adding stuff to force the game to be different.

    No matter which way you slice it under your metrics, you’re going to hate something in games and always want to make them better… But they’ll never be Games just Toys, Puzzles or Contests.