May 042009
 

Malcolm Gladwell’s latest article, “How David Beats Goliath,” is a must-read for anyone interested in game design. Or business strategy.

It’s all about how underdog outsiders can come to a “game” (meaning, a formal structure of rules with win conditions) and because they are free of social preconceptions of how it “should” be played, can use unorthodox tactics to win. The article purports to be about game-playing strategy, but I think it has just as much to say about how you set up your rule systems as anything else.

“Eurisko was exposing the fact that any finite set of rules is going to be a very incomplete approximation of reality,” Lenat explained. “What the other entrants were doing was filling in the holes in the rules with real-world, realistic answers. But Eurisko didn’t have that kind of preconception, partly because it didn’t know enough about the world.” So it found solutions that were, as Lenat freely admits, “socially horrifying”: send a thousand defenseless and immobile ships into battle; sink your own ships the moment they get damaged.

This is the second half of the insurgent’s creed. Insurgents work harder than Goliath. But their other advantage is that they will do what is “socially horrifying”—they will challenge the conventions about how battles are supposed to be fought.

  25 Responses to “How David Beats Goliath — a lesson in game design”

  1. Good article. Much of Iron Realms’ early success fits that pattern.

  2. Lots of good stuff here.

    I think this can also be a good explanation for why PvP is often very different from PvE in online games. The developers think like Goliaths when they build the PvE content. Then the players come in as a swarm of Davids, pushing the game systems in ways that were never intended.

    Another thought on game design… the point of war is to win. A game is, or should be, just as much about producing an enjoyable experience for everybody as it is about winning. You could say that the Basketball team in the article was employing a “creative use of game mechanics.” When people made “creative use of game mechanics” in UO to steal things from people’s houses, it created an environment that wasn’t fun; the game was patched, and the mechanics were changed to produce a more desirable set of circumstances. Basketball itself has been patched for much the same reason. The 5 and 10 second rules mentioned in the article (as well as the shot clock) were added in response to another “creative use of game mechanics”: teams were winning by getting an early lead and then just holding onto the ball so the other team never had the chance to score. Perfectly legal, but not a lot of fun for anybody.

    Right, so who has time to actually write a “Designing for David” article?

  3. I think this can also be a good explanation for why PvP is often very different from PvE in online games. The developers think like Goliaths when they build the PvE content. Then the players come in as a swarm of Davids, pushing the game systems in ways that were never intended.

    The difference between Davids and Goliaths is often summed up in game circles as the difference between “playing by the rules” “exploiting” and “cheating.” There’s a subtle common parlance by which “exploit” means to use the letter of the rules rather than the spirit, and it may or may not be considered “cheating” based on the culture around the game.

    I do think that a game does not always have to be fun for everybody. Sometimes what a game teaches you are painful lessons.

  4. Funny you should mention PvP: dedicated PvP players dominate utterly when you fight them on their own terms, because they know the turf better than you do. But when you change the rules of engagement, they often panic, run away, and die fleeing… and then complain to all and sundry that you must have cheated in some way to overcome their innate superiority.

    The best of the best can’t be taken quite so easily, because they know their crew inside and out, adapt fluidly to changing conditions, and retreat in good order when they’re put off their stride. But in a surprising number of cases, all it takes is a ‘victim’ or two who stages a berserk suicide charge when they were expected to run, and the attackers can be thrown into utter disarray. Sometimes all it takes is a character spec or gear that’s “wrong”… something that nobody else uses because it’s not considered optimal by conventional wisdom, and consequently is outside the experience of the attacker.

    And if that doesn’t work… some will also fall for a fake key and a guard-zone rune marked “The Castle”. That one wasn’t very nice, but it’s one of my fondest wins 🙂

  5. I often find myself thinking about the edge cases in non-programmatic systems. Examining how a system reacts when the inputs are within stated limits, but not expected limits, is a good way to discover and prevent “exploits”. It’s easy to assume that people will follow conventions, but all you need is one person who doesn’t to show you how weak the system really is.

    When I play games, I usually poke and prod to see if I can find the weak points in a system. Once I figure out the underlying system, the game usually becomes less interesting.

    I think that although a game with “exploits” might be less fun, it can be very useful as an educational tool. Once you figure out that weight doesn’t affect your customized car’s class in Sega GT 2002, you probably won’t forget the importance of power-weight ratios, and in the process, you have learned important lessons about efficiency.

  6. “Moneyball” said all of this, and said it better and with more interesting statistics, back in 2003:

    http://www.worldcat.org/oclc/51817522&referer=brief_results

    Nobody in either the case of the basketball players, the battleship computer game, or the baseball team in “Moneyball” was even, really, “exploiting” the rules. They were following the rules differently, approaching them with different insights. “Exploiting” to me sounds, well… exploitative. The idea of observing the letter of the law but not the spirit.

    Also, in terms of war, there aren’t (really) rules, per se. Yes, Geneva Conventions, physics, etc. But killing people is often much more serious business than playing fair.

    I also don’t believe that Goliath can’t use interesting, novel approaches to games, wars, etc. Apple is only really a “David” when compared to Microsoft; not a small, indie company by any stretch. And yet they’ve had a decent track record of out-of-the-box thinking. Same for some of the Japanese car makers (and, to be fair, even some US ones, thought they fall down on many other criteria).

    I think the lesson is less about size and more about attitude.

  7. EURISKO, wow, I remember that from back when I was doing my PhD in the 1980s… It represented both domain and control knowledge as frames, and had meta-level heuristics that would work on both, I seem to recall. It did very impressive things, but there was some suspicion that there was something not quite right about it.

    (Picks up PhD and looks it up)

    “The system is immensely powerful and has been applied to many varied domains, although descriptions of it are curiously scant on details of what it produced and how it did it”.

    Yeah, that’s about the size of it.

    Richard

  8. I do think that a game does not always have to be fun for everybody. Sometimes what a game teaches you are painful lessons.

    Good point. The addition of the shot clock to Basketball was more about putting on a good show for the spectators than it was about the players. If you substitute “fun” for “any goal(s) which the game is intended to accomplish” then I think my point still stands though.

  9. Another way to approach this is to even ask yourself if you need to fight Goliath at all.

  10. I do think that a game does not always have to be fun for everybody. Sometimes what a game teaches you are painful lessons.

    I think the key from a player’s perspective is whether the painful lessons pay off in a reasonable timeframe. Otherwise you stop playing the game — or at a bare minimum, stop playing the painful parts.

    If the only lesson is “danger is everywhere, death is unavoidable, and war is eternal”… well, I can dig that if I’m able to jump right into the action and get my licks in (Battlefield, Planetside, et al), but if I’m the designated target for six months before I have the chops to fight back, forget about it. I’d just as soon go play Hello Kitty Online while listening to the Best of Abba.

  11. I think UO is a bad example of “sometimes what a game teaches you are painful lessons” though. UO was kinda broken in that respect, and the lesson for a long time was basically that you don’t play the game unless you want to be involved in the pvp game. The fact that people who wanted no part in the pvp kept playing despite that was very much a factor of there being no where else to go, and that’s what caused so much discomfort among that population; they were deliberately struggling against what the game was trying to “teach” them.

    If you’re going to have painful lessons they need to not be persistent parts of your gameplay experience, unless your gameplay experience is marginally repeatable. Other than those sorts of one off events, the difficulties do need to be conquerable in some way. In UO there was no way to “win”, so there was no way to improve the situation. You “got” it the first time you got pked, but there was very little you could do about it after that without playing their game and not the one you were interested in. And even if you did take up arms, there was no way to actually achieve any sort of lasting victory. You knocked em down they got back up again just as strong. It was a very hopeless situation. Is there some value in having gone through that? Sure, but it’s not sustainable in any way, not if you’re charging for it.

  12. In UO there was no way to “win”, so there was no way to improve the situation.

    You could have learned hacking, broken into their computer, installed a Trojan, and then nuked them at just the right moment.

    Oh wait, that’s cheating, huh? Go Goliath, go!

  13. I’d just as soon go play Hello Kitty Online while listening to the Best of Abba.

    Hey now, “The Winner Takes It All” is all about painful lessons. 😉

  14. @Michael, No that’s not even a solution, they’d be back the next day on a new rig with even better security! 😉

  15. Hey now, “The Winner Takes It All” is all about painful lessons.

    Hrmmm, good point. And then there’s “Fernando”…

    In UO there was no way to “win”, so there was no way to improve the situation.

    Arguably, the anti-PK forces achieved the ultimate win in UO when Trammel was introduced. And we did it by going outside the box – fighting Goliath on the message boards and mailing lists, rather than on his home field. Perhaps it would have happened without any player activism, given that the game started hemmoraging players the second that a non-PvP alternative was available, but I’d still count it as a decisive carebear victory.

    If Trammel had come sooner rather than later, we might have a greater number of sandbox worlds descended from UO rather than theme-park worlds descended from EverQuest. But that’s another type of painful lesson.

  16. If Trammel had come sooner rather than later, we might have a greater number of sandbox worlds descended from UO rather than theme-park worlds descended from EverQuest. But that’s another type of painful lesson.

    UO has steadily moved more to the theme park style since Trammel. It’s still got the major part of “Sandbox”, but playing it and reading about it now feels quite Themeparkish. They have quests for achievements, certain places to get certain things, “end game” like content, etc. Some of this isn’t really different, just fortified 12 ways. All the while, the social structure has changed entirely.

    This change was foreseeable. Not that it had to happen that way, but the mindset was obvious. Some people used to post on their forums that they are foolish to try to compete with WoW. But they did, and they lost that battle. Their boards are now gasping for air, their game world less active than some private servers, and the majority of their dwindling subs are on the other side of the world.

    The biggest thing was exactly what this topic is about. They went orthodox, and relied on copying. It was so bad that even their major event at one point was found to be copied from an event on a private server (Belo Ondariva, if I remember the name right?). They had no initiative left while trying to play the same game as the big guys.

  17. @Michael, No that’s not even a solution, they’d be back the next day on a new rig with even better security! 😉

    Pft. You just have to do it more subtly. Finesse, sir! I’d give you details, but I’ve never played UO.

    Also, see Yukon’s comment. The point is that you’re defining your method of winning as vital, which demands sticking to the rules of the currently dominant, which results in a staggering loss rate of 75.2%. That’s exactly what the article is saying is a bad idea.

  18. @Yukon, Trammel also created issues with the social game fabric of UO though, so in effect, while it solved the pk problem, it did it by sacrificing other genuinely good parts of the game as well. It was the right decision, but it was lamentable that that was the case.

    And that’s kinda the problem. The solution was to invalidate the situation, not to actually solve it. And Trammel caused a number of issues in the long term as well, both in terms of community building an in terms of economy. UO actually needed PKs in some number in order to function the way it was designed, they just needed to be on the losing side, not the constantly winning side.

    I’m also not sure that UO would have done better had there been a Trammel earlier on. UO’s pve game has always been horrible shallow, and the community ties wouldn’t have developed in the same way had Trammel been there from the start. I think they would’ve genuinely been weaker, and I think that would’ve caused it to spiral downwards much sooner. In order for sandbox games to have been bigger in the market place they would’ve needed to be more compelling on more levels than just not having to worry about getting ganked when you’re walking outside the guard zone.

    @Michael, eh, no. I don’t have a method per se, because there was no method in UO. What ended up happening was segergation, not success. They took the conflict out of the game. So the people that didn’t want to play the PVP game had somewhere to go, but at the cost of the pvp game existing. The PKs won, and the devs surrendered and removed them from the game for the good of the overall playerbase. If the PKs had lost, Trammel would not have been necessary. In the article this is like the naval game people banning the guy who keeps winning by using computer generated models. The PKs were David, and Goliath “won” by contriving them out of existence. The ruleset didn’t change to account for that though, and it caused issues.

    Remember, the rampant pking and not the pvp system were what broke UO. The pvp system actually has some valid elements to it. Those were stripped out along with the bad parts because a workable solution to maintain those good points was never reached. Unfortunately, I’ve yet to see a system that actually works for what UO was trying to do, and that’s the really sad part of this. Scale seems to break solutions that worked in MUDs, and not having a similar system changes the nature of the game into something else entirely. It really makes me think that games like UO just plain cannot function at the 100k plus users scale.

  19. Eolirin:
    Remember, the rampant pking and not the pvp system were what broke UO. The pvp system actually has some valid elements to it. Those were stripped out along with the bad parts because a workable solution to maintain those good points was never reached. Unfortunately, I’ve yet to see a system that actually works for what UO was trying to do, and that’s the really sad part of this. Scale seems to break solutions that worked in MUDs, and not having a similar system changes the nature of the game into something else entirely. It really makes me think that games like UO just plain cannot function at the 100k plus users scale.

    I think it can be done. But you have to add a few things that UO never placed in their game.

    1) Real player communities. Player built cities. There’s two important aspects here, the community by itself and the community in relationship to others as part of a world. There needs to be reasons for the community to exist on it’s own, but each community also needs to interact with others.
    So players need a benefit for belonging to a community, and at the same time be able to enhance that community. And each community needs to have reasons to interact with other communities, such as trade. Meaning territorial resources are important and need to be varied.

    2) Players need to be able to pursue not only justice, but be able to reinforce their own safety through said justice. It doesn’t do any good to stop injustice on an individual event basis. Players need to be able to stop, or send reeling, the causes of injustice. laws must be in place and enforceable.

    Only by combining these aspects together do you get the kind of cooperation and spirit required to make open PvP a viable option in an MMO.

  20. Almost every game since UO, including some notable “hardcore” titles, has incorporated segregation of PvP. The clear lesson from UO is if you shove PvP down the throats of the player base, they’ll walk.

    It can be a zone system, a flag system, or both in parallel. As long as the predators can’t gank random newbies minding their own business, it works.

    UO could have done it better, in addition to doing it sooner. Rather than “mirroring” the existing land-mass, they could have given us a new continent (mirroring was an elegant use of limited resources, but ultimately unsatisfying from the player perspective). Had they designed such a region layout with the needs of player communities in mind, and allowed existing communities to relocate as unified entities, and run with the “city stone” concept, and prioritized a lot of really cool stuff that they added too late…

    Then it would still be an isometric 2D title fighting against some damn sexy 3D eye candy. But it would have kept the solid, unique community focus that has, to this day, eluded those prettier games.

    Not only can you have a dynamic and vibrant community without any PvP to speak of, you can have it without any combat system at all. There will still be inter-player conflict, because any time people are free to interact, there’ll inevitably be conflict. But it’s a lot easier to channel conflict in a controlled manner, one that will encourage the losers to learn, improve, and try again, if you don’t have random gank squads killing off your miners and lumberjacks. Give the citizens security and give the fighters a nice, fair, balanced tournament ladder in the city arena.

    If Goliath’s an anti-social jerk that’s costing you subscriptions, there’s nothing wrong and a great deal right with the designer putting a rock right between his eyes. But if you want players to wield the sling, let’s talk pay and benefits package… ’cause hell will freeze over before it happens for free.

  21. If Goliath’s an anti-social jerk that’s costing you subscriptions, there’s nothing wrong and a great deal right with the designer putting a rock right between his eyes. But if you want players to wield the sling, let’s talk pay and benefits package… ’cause hell will freeze over before it happens for free.

    Would you accept that pay and benefits package in-game?

  22. Remember, the rampant pking and not the pvp system were what broke UO.

    I can’t remember that. :/ You guys grew up with UO; I only found out about it well after all the major phases had become history. I’m not aware of any wiki or book that actually records and gives context to all the major lessons learned from UO that I can go and buy.

    Someone needs to make that.

  23. I can’t remember that. :/ You guys grew up with UO; I only found out about it well after all the major phases had become history. I’m not aware of any wiki or book that actually records and gives context to all the major lessons learned from UO that I can go and buy.

    Someone needs to make that.

    Not the lessons, but the things players did and the tools the game gave them to allow them to do these things. The lessons were all handled wrong, at least the major lessons. A fresh look at solving the problems that arose from the marvelous mechanics is sorely needed.

  24. We’ll need a trilogy of books if I keep rehashing Trammel.

    What I was originally getting at is this: unlike the real world, in the virtual world it’s sometimes impossible to change the tactical framework in your favor due to the game physics (barring cheats and exploits, which are undesirable on a number of levels).

    If you’re David in that situation (the players at a major systemic disadvantage to Goliath) sometimes your best option is to nudge the dev team and get the rules fixed. And if they don’t respond quickly, go find a more balanced game.

  25. Another way to approach this is to even ask yourself if you need to fight Goliath at all.

    You got your agon in my paideia! No, you got your paideia in my agon!

Sorry, the comment form is closed at this time.