CGD Podcast Episode 23 – “On Games At the Games”, A Conversation with Frank Lantz


This week I had a great conversation with NYU Game Center director Frank Lantz about randomness and general game design philosophy. We meant to get to three other topics – execution, reading and improvisation, but not all-that-surprisingly, we never got there in the 70+ minutes of this episode.

Mentioned in this episode:

Frank’s “Against Design” article

David Deutsch’s The Beginning of Infinity

Nassim Taleb’s The Black Swan

Nick Bostrom’s Superintelligence


Let me know what you thought of the conversation in the comments below (if you’re lucky, Frank may even be hovering nearby to respond!)

If you enjoyed the show, please consider supporting the show by becoming a patron at!

  • Too much to comment on everything, so here are some slightly disjointed thoughts on a scattering of points…

    Frank wants people to question their models more, have less confidence. I agree! Probably my biggest gripe with your (Keith’s) theories has been that (at least from my perspective) they seem overly confident. Now I know at the end you said that you’re maybe not quite as confident as Frank (or I) think, and I will also point out that for me this is often more true of non-game-design topics, so perhaps more elaboration on this isn’t necessary.

    RE: Wanting structure in games:

    Games with little to no structure are maybe kind of like a song with only verses and it’s literally just a series of verses one after another and they’re all the same with different words and then at some point it just stops and it’s over. On the other hand, games with structure are like songs that have verses, choruses, a bridge, a deceptive cadence, development, surprise, drama, etc.

    RE: Different methodologies/Different values/Purported opposition to having grand theory/etc.:

    This has come up before. I remember when Frank posted “Parley” you expressed that you didn’t really understand what he was trying to say, and I suggested that one thing he was saying was basically that his emphasis on mechanics shouldn’t be taken as indicating any sort of desire to downplay or oppress or deny any or all other possible approaches. And I think this may have been an example of the kind of overcorrection error Frank is talking about here, because you seemed to think that he thought that “The only people who aren’t correct are those who say that not everyone is correct!”

    However, asserting the plurality of values doesn’t mean everything is equal or that there isn’t any such thing as being wrong or right. As far as I can tell, the point here is that there are potentially different kinds of value that can be derived from games, and even if one mountain delivers more value quantitatively than another, that doesn’t mean that a shorter mountain necessarily ought to be rejected wholesale once we discover a taller one. Why not? Because they are different mountains with different flora and fauna and characteristically different nooks and crannies and caverns inside them and so on. The value of one mountain may be quantitatively less on net than the value of the other, but we may get something characteristically different from the one than the other. It is qualitatively different value. We should therefore try to find the peak of every mountain we come across!

    You really seem to think Frank is opposed to having a greater design philosophy, and I don’t understand why. He isn’t. Rather, he seems to me to be in favor of having multiple design philosophies that are useful for different reasons. (And in light of comments made towards the end of the podcast, also maintaining a healthy openmindedness about what can be found by exploring different mountains, because it is very hard to tell which mountain is actually the tallest anyway.)

    RE: Value in art

    Very often what I want (not always, just very often) from art is a trigger for subjective transcendent experiences. I want to be moved, I want to be shaken up, rattled around, forced into a place where I have to reorganize my emotional thinking a bit. I don’t always want a puzzle of art components where I am just trying to wrap my head around the formal relationships between the art’s pieces. I want a piece of art that knocks my socks off emotionally, but specifically in a subjective way. This doesn’t mean I never want a formalist puzzle or whatever. Just not always. Sometimes I want an occult magic spell in the form of blackened thrash metal about Mesopotamian deities. Other times I want “Starless” by King Crimson. There is room for both (and I get very different things from each of them).

  • Forgot to mention this, but in general this was a very good podcast episode! I believe the ones with conversations between two or more people are always the best, so if indeed you guys do more of these I will listen again and am looking forward to it!

  • tinytouchtales

    Man this might be my ultimate fanboy episode so far. Thanks for making it! Next up: Reiner Knizia!

  • Just so people don’t think I’m crazy for leveling this critique at Frank, I quote from the “Against Design” article:

    >>Look at Shadow of the Colossus for example. What do we, as game designers, know about videogames? Well, we know a few things, we know boss battles suck. We know jumping puzzles suck. We know you get great games by focusing on meaningful interaction and you don’t get great games from aping cinema and focusing on graphics.So, how about a videogame that is nothing but boss battles, and each boss battle is a jumping puzzle, and the whole thing is set in a giant empty world with nothing to interact with, and a lot of the main motivation of the game was an attempt to achieve some film-inspired visual effects? Does that sound like a good recipe for creating one of the greatest videogames of all time?


    >>Look at the AWP, the signature 1-hit-kill weapon in Counter-Strike. It’s completely unbalanced. Any sensible game designer would have rejected it. Luckily for us, Counter-Strike wasn’t made by sensible designers, it was made by unreasonable people who kept this unbalanced ingredient and evolved the rest of the game around it.

    In both cases, the message seems *to me* to be a celebration of the wrongness of theory. It’s cherry picking examples where the theory DIDN’T work (nevermind the countless examples in both games where there absolutely is theory that explains much about these games) in order to send the message that theory in general doesn’t work.

    But again, as I said in the episode, I understand that Frank comes off this way more than he intends to because he is trying to pull a pretty small circle of game design theorists in the direction of “be less confident”. Which I think is legitimate. But I think my general interpretation of his work is also pretty fair, given what is written down.

  • hilbert90

    This is, without a doubt, one of my favorite episodes so far. I actually came around more to Keith’s side than I thought I would, almost entirely from Frank’s use of Magic as an example. He said something like: Richard Garfield basically had this random idea and tinkered with it and it turned into this great game.

    In reality, Richard Garfield had an idea outside of some philosophy, but then to make it good he probably used quite a bit of theory, even if he didn’t think of it in that way. If each person has their own deck, then the space of possibilities is cracked wide open. But why did he think this was a good thing? Probably because he already had a philosophy that said having a huge probability space would make for interesting decisions.

    Why did he think having rules on the cards that would interact in interesting ways was a good thing? In the fine tuning stages, why did he make sure that skill would play a large role rather than complete randomness? How did he balance a game where each person would bring their own decks?

    My guess is that these weren’t random stumblings upon something that worked. He had to make actual choices all along the way, and those choices were probably in line with some overarching philosophy about how he wanted the game to look once he had that first idea. And if you just unpack Frank’s own explanation of how Garfield made Magic and why it’s a good game, you end up more on Keith’s side than Frank’s.

    Anyway, I want more of these too!

  • hilbert90

    Sorry to flood the comments section, but now that I’ve thought about this more, I think I’ve achieved more clarity. Frank seems to be conflating two things that should be separate. One is the idea or concept itself. The other is the execution of that concept.

    The Wright brothers didn’t build an airplane from some deep understanding of physics. But our modern airplanes were certainly fine tuned by starting with the physics and tailoring with that understanding and modelling. The core idea of a novel might come in some mysterious insight, but most of the work that goes into writing a novel has to do with crafting the prose according to your aesthetic theory and making sure it conforms to a preordained structure and so on.

    I think similarly, game design can start with some mysterious insight, but in order to make a good game, you have to build off of that idea in a direction dictated by rules that you’ve already deemed as likely to make a good game. In the mountain analogy, maybe there’s 10,000 directions you can go, and only one of them leads up to the peak. Sure, you could tinker with intuition until you randomly find that direction, or you can use a theoretical framework to find that direction more quickly. Can you imagine trying to strike upon the modern airplane while ignoring all theory provided by physics?

  • Knizia is one of the world’s best designers but I’ve talked to him a few times and I’m not sure he’d be a great fit for the podcast. He seems to be not-consciously-philosophical about design; he doesn’t verbally seem to have many strong opinions on game design (I’ve often wondered if he just is protective of his secrets, haha). In either case, yeah, I don’t know if he would work too well.

    You know who I’d love to get would be that Clearclaw guy from Boargamegeek.

  • franklantz

    Actually, I’m not really trying to persuade anyone to change their attitude or behavior, I’m just trying to get at the truth. I just want to have as clear a picture as possible of how things really work.

  • Van

    Let’s say you’ve made a game that’s really enjoyable, but drags on a little too long. I get the feeling you would say that “see, making this game drag on wasn’t so bad”, while Keith Burgun would say “we can probably do something about that”. Or maybe you would say something like “well there’s this other game that works great BECAUSE it drags on for too long, so don’t bother designing around that issue”. Or would you? I don’t know.
    I try and try to make sense out of your reasoning and somehow it always ends up in the “everything is subjective” and “if we don’t know something NOW, we never will” category.

  • franklantz

    Your feeling is wrong. I wouldn’t say any of the things you imagine I would say. Nor have I done so on any of the 40+ games I’ve designed, each one of which involved a painstaking, blood-sweating process of summoning all the effort and ingenuity I could muster to attempt to make it as good as it could possibly be.

    A more reasonable imaginary conversation would go something like this:

    Keith: This game would be better if it was shorter.
    Me: You are probably wrong about that.
    Keith: Why?
    Me: Because your elaborate formula for explaining what makes a game good is terrible.
    Keith: [Raises one finger in triumphant gesture] Aha! So you admit that things can be bad!
    Me: Yeah, and even though you literally have a 50% chance of being right about this particular design decision I’m going to try the opposite thing just to make a point about how terrible your formula is and how borderline obsessive/compulsive you are about declaring it’s ability to predict and explain all of game design. [Playtests slightly *longer* version of game.]
    Keith: But I –
    Me: Hey, this is much better!
    Keith: No fair! You just got lucky.

  • Van

    Well, I’m glad to hear that I was wrong about your perspective and I had a good chuckle from the imaginary conversation. Thanks for your reply, things make a little more sense now.

  • I have to echo the other sentiments – this episode is incredible! The immediate feedback to stated ideas forced them to be explored deeper which was great.

  • All right, I’ve consulted the pros (my wife, my son) and none of us have heard of natty-twen. So you’re being sent to Mulligan Island until you repent. (Further check – just 3 results on Google…that’s like less than zero. So, there’s no way off the island. Sorry!)

    Some excellent talk in this video, too. I had wondered if this would be a Siskel and Ebert kind-of-thing.

  • Jeff S

    Podcast transcription with my own twist:

    Kieth: Hey! Welcome!

    Frank: I hate you subconciously but some of the things you say are in my head.

    Kieth: I don’t really hate you but I subconciously hate how you don’t really seem to stand for anything.

    Frank: I agree.

    Kieth: I agree.

    Frank: I think we can both agree upon agreeing.

    Kieth: Yeah!

    Frank: I don’t like your opinion on randomness.

    Kieth: I don’t like randomness because we don’t learn anything.

    Frank: I can see that.

    Kieth: So, when we implement certain things into our games, it draws us away from “Good Design” because we want the player not to calculate, not to throw things to chance, but to learn, experience and understand that their choices matter.

    (Frank forgetting that he hates Kieth)

    (18:55) Frank: I agree… wait… so why am I here? Why am I still arguing?

    Kieth: Well, I’m trying to highlight that I’ve quantified/labeled/analyzed randomness in a way that we can reduce/eliminate/use it in a way that’s helpful rather than just inject into a game.

    (25:50)(Frank remembers)

    Frank: Oh yeah back to arguing rather than understanding. I have to maintain this false dichotomy for the sake of my sanity because agreeing with you mentally is easy but emotionally it isn’t. I have a bunch of points as to why randomness is cool to combat your randomness is lame statments.

    Kieth: Yeah, I’m not against those. For what they have been used they’ve been used well. I just don’t think/feel that they are being used properly to achieve a system where randomness (rare occasions/conditions) manifests itself rather than injected randomly by the designer.

    (Frank still feeling confined)

    Frank: Look Kieth, I have a problem/challenge/blessing. I live a lifestyle where I don’t see right or wrong or good or bad. I see something or an experience. BUT I don’t want to even call it that. I ask you why something is good or bad because I don’t seem to understand that FEELING even though I can understand the UNDERSTANDING and can/often agree with you. BUT since I don’t FEEL congruent with my thought and emotions I will continue to push back. I don’t really label/categorize/quantify anything because the core of my personality is based upon FREEDOM. Even labeling myself as a person who enjoys freedom and experiences feels limiting to me. I am not right or wrong or good or bad. I am often both or none at the same time. I am Frank Lantz.

    Kieth: Frank… I have a problem/challenge/blessing too. I spend copious amounts of time vetting and undestanding systems, adding or subtracting to theories and analyzing methods to excruciating measures. (And guess what I enjoy it) Frank, you may not know this, but I completely and fully believe that there is a large possibility that I am wrong.

    Frank: Aha! So you can be wrong! Although I want to gloat right now, why do you have to be wrong why can’t you “just be?”


    Kieth: Shut up and let me finish! Right now, my intuition/math/etc is telling me that this method is the best I’ve developed so far. And so far, I think it makes a lot of sense t’wards the feeling/thought I’m going for and so it feels right. I’ve invited you to come along because I think you’re an outside the box thinker.

    Frank: Why can’t I be outside and inside the box? I mean, why does there have to be a box? Who said there had to be a box? Where did this box idea come from? I read this great Richard Dawkins bo..

    Kieth: When a better method comes around I will adopt that one, but right now I am the only one that I know of that’s really searching for a certain something by looking for a good method or theory. I’m trying to explore to find ways to enhance my theory. Too many people are either ignoring design altogether or very possessive/stubborn to their own way.

    Frank: But why do you have to have a method?

    Kieth: Everyone has a method.

    Frank: But why does there have to be one?

    Kieth: Because if there is already one way to make the wheel to move objects then I can use that to work on developing experiences or joining cool things like horses or engines to the wheels.

    Frank: I’m not like you where you have to have proscriptions. Art is relevent.

    Kieth: I’m always going to have a method, whether it’s mine or whether it’s someone else’s.

    Frank: Then I’m probably always going to tell you to stop because it makes me subconsiously hate you.

    Kieth: Are you ever going to have a theory that sticks with you?

    Frank: I don’t really like the idea of “One” theory, and my answer is both.

    Kieth: Both?

    Frank: Yes and no and both and none. Look the more you limit yourself (remember I’m all about freedom) then the less time you’ll spend enjoying the experiences of the “mountains” in front of you.

    Kieth: I can see that. Well, time’s up.

    Frank: But I do want to bring up the algorithmically thinking thing vs hard calculation. That’s fascinating. Maybe we can make this a regular thing?

    Kieth: Yeah, talk to you soon.
    Kieth: Be my patreon bitches!

    Did I capture the essence?

  • This is perhaps the fun-niest and yet very nicest comment ever.

  • Isaac Shalev

    Nat Twenty, otoh, has like 90 million google results. Maybe Keith’s friend is just really close w Nat, and calls him Natty.

  • Isaac Shalev

    An enjoyable episode, for sure, with lots of ideas to chew on.

    Keith, I think you triggered me most when you said “Everything is reason” and “art is the softest science.” To me that’s the art-critic’s fallacy, which is that our ability to dissect elements of greatness in existing art means that there is a formula that we could apply prospectively to create it. That’s just the Bob Ross school of painting. I don’t think any artists or scientists would agree that they’re engaged in the same project, their processes and ideas about their work are radically different, and their work is experienced in incredibly different ways. The claim that ‘art is the softest science’ needs to be clarified and defended, or perhaps taken in the looses sense possible – that principles and process are applicable to both in some degree.

    Finally, looping back to randomness, I think Frank missed a great opportunity when he declined to give an example of a game where the designer injected randmoness to create satisfying challenges. Garfield’s formulation that randomness reduces return to skill is, I think, imprecise. It reduces the return to many skills, but greatly increases the return to the skills of risk management and valuation. I think we all agree that there is a lot of value to input randomness as a design tool, and that output randomness deserves stricter scrutiny.

    I wonder if you would you agree with this proposition: output randomness is not a presumptive problem in games where the more-skilled player wins over the less-skilled player at some substantial rate, perhaps in the neighborhood of 70% of the time.

    I also wonder if your objection to randomness from an input-response perspective is ameliorated if the signal about how good your decision was isn’t clouded by the randomness of the outcome. If your odds and payouts are clear, the game has already done its job as an information system, no?

  • It’s not so much that the “more skilled player” wins some percentage of the time, it’s the “player who made
    better inputs on this match” wins… really, close to 100% of the time. Any time that Player A makes inputs to Player B, and wins, the game has failed to work. Granted, it’s going to happen, even in games that only have input randomness, but output randomness really makes the situation much worse and I just really don’t see any need for it.

    >If your odds and payouts are clear, the game has already done its job as an information system, no?

    No. You’re thinking of a game as “a series of loosely connected decisions”, which I will grant you, describes most existing games. But for games that actually use structure, that have a beginning, middle and end, which have larger strategic arcs and a meaningful win/loss, it doesn’t matter that I know what the possible outcomes are of one move. The one move is just a tiny piece of a much larger picture, and the issue is that with output randomness in there, it becomes impossible quickly to form a coherent, long-enough strategy that will make the final outcome meaningful.