Probably a brief thought on probability

Civilization: Revolution

Sid Meier and Rob Pardo of Blizzard gave a couple of talks recently at GDC 2010,  and they inadvertently talked about similar things; fudging the math away from “truthful” math under the nose of the player, usually to the player’s advantage.  Meier, for example, reported frustrations in playtests when players lost fights where the game had given them a 50:50 chance to succeed, and even more so when they lost consecutively.  Mathematically, it’s a completely reasonable scenario but that didn’t matter to the beaten and battered player.  They decided to fudge the math that actually increased chances of success after failures above and beyond what the statistics would report (the player’s unit strength vs. their opponent’s).  Players responded well, and now the game launched with that system in place.

The debate here is that fudging the math like that seems to be undercutting the strengths of gaming to pushing players’ abilities and thought process to higher levels than before the challenge is presented and their experience has stagnated for it.  Jaron linked a good overview of that argument on the Game Design Advance blog here.  The points being made there are valid and I don’t disagree with them at all.  A good game, like a good book or a good movie, isn’t afraid to keep shoving you towards its own intentions, no matter how uncomfortable you might get, until you start to learn, grow, and better yourself through its obstacles (or stories or messages or whatever).

My only comment is that context is crucial in a design decision like this, especially to their targeted audiences.  Meier’s talk was referring to the development of Civilization: Revolution, a console-specific “port” of the concept of his more-hardcore Civilization games.  You can compare just the presentation layer of Civ:Rev to something like Civilization 4 and see what each title is going for.  Civ4 aims for an absolute, ruthless emulation of a rise and fall of civilizations and their interactions with other civilizations in between those times.  Civ:Rev adds a little brevity to that aim and to make it approachable for those who don’t wish to invest in a serious simulation and minimize any walls towards the “fun factors” like combat or just saying “Look at all these cities I own!”.  Frustration should be a minimized factor in a game with objectives like Civ:Rev has, so fudging the math there with the intention to frustrate players less, no matter how “wrong” or “right” the player’s perception of the game space is, seems reasonable to me.  If they did this with a Civilization game, on the other hand, that would seem to violate that game’s aim of a hyper-realistically-operating game world.

In Blizzard and Rob Pardo’s case, players in World of Warcraft rebelled against a system that reduced their experience game over time to reduce mindless “grinding” gameplay.  Pardo decided to, instead of starting at 100% experience gain and falling to 50% gain, start at 200% gain and reduce to “regular” 100% gain, failing to mention that all experience point goals were doubled.  Everyone cheered, and Pardo achieved his intended design philosophy.  While that can be seemingly deceitful as a practice, the context of WoW includes it being an expansive, immersive, all-inclusive (almost to the point of casual) experience.  He wanted players on more footing, instead of heavily weighting rewards towards those who spend impossible amount of time in the world and reduce burnout among players not “hardcore” enough to handle grinding.  This seems like a reasonable objective.  Now if Blizzard decided to arbitrarily adjust the math behind interactions in Starcraft, an exacting, stressful, hyper-competitive environment, they would be asking for trouble.

My point here is that game design decisions like this can be easy to pick on when we all want amazing games that push our limits of thinking, recreate the most affecting simulations of amazing situations, and affect how we view the world and each other and our relationships and so on and so forth.  The truth of the matter, however, is that game design decisions have to fit into the context of the modern reality of games today; they all have different objectives, different markets or niches to fit into, and all different audiences to emphasize.  Debate on decisions like probability are great for expanding our view on when we should or should not wield it in a design.  We shouldn’t throw out a tool out of our toolkit wholesale because we don’t tend to favor what that tool usually builds; there will probably be a great fit for it on a project somewhere down the line.

When someone asks you what kind of processor is better, a 2 year 0ld dual core or an eight-core monster, the most accurate answer is “It depends.”  The Crysis 2 player will want the high-end processor where the the two high-school sisters can use a cheaper processor so they can enjoy Farmville reliably without unnecessarily breaking the bank.  It’s all in the context.

Also, that wasn’t a very brief thought I had there.


3 Responses to “Probably a brief thought on probability”

  1. March 30, 2010 at 2:50 pm

    I see where you’re coming from, but I don’t think fudging the math in the player’s favor is the problem as much as outright lying to them. What Blizzard did with the WoW experience system is fine, it’s just playing off of human psychology (http://www.psychologyofgames.com/2010/03/16/framing-and-world-of-warcrafts-rest-system/). My problem is with a game telling somebody “you have a 2:1 chance to win” if it’s not actually a 2:1 chance. If you want to increase the odds each time, that’s fine, but I’d prefer the stat reflect the internal fudging. As a gamer I don’t like having to second-guess the hard information (numbers and the like) the game is giving me. Potentially far worse, though, is that games teach and inform us, often from a young age. If it becomes enough of a trend in this heavily trend-based industry, people who grow up with games that do this may develop an even more warped perception of statistics than than the average person. A significant portion of our lives, including gaming, is based on intuition; it just seems like a step in the wrong direction.

    • March 30, 2010 at 3:28 pm

      Where’s the line between the lie (Civ:Rev) and the deceit (WoW)? Both methods are still an intentional obscuration of what’s going on behind the calculators. The bottom line in these endeavors is the whole “just make it fun” deal. Black holes of fun were literally rooted out, analyzed, and then some retooled method put in its place and then fun happens again, proven through rigorous playtesting or feedback. Games have the capability to teach and inform, and it’s great, but what games actually do teach doesn’t have to be pigeonholed into obvious truths, and sacrificing fun factor for a targeted audience in a product that should be entertaining to them seems overboard. Civ:Rev may be giving you slightly off numbers on a battle productivity, but it’s not devoid of teaching you something like recognizing patterns and altering your play style to capitalize. Once again, a decision like this has to be framed in each context separately to see where it would help or where it goes south; there are a million different kinds of ‘truth’ possible in a game and that’s part of the beauty.

      As far as warping perceptions of growing minds, you have to admit that was pretty heavy-handed of you. 🙂 Intuition is gained throughout most life experiences, and a person playing a game is going to bring their own brand of it before they even start playing anything. If games feature a slight deceit, I think their intuition will get just as sharp discovering them and manipulating these systems to an advantage as much as receiving probability data and weighing it at face value.

      • March 30, 2010 at 3:45 pm

        The line is that WoW is still being entirely honest about what’s going on; it’s merely playing with human psychology. As far as I understand it, Civ: Rev is actually giving you misinformation. Again, I’m not saying Civ: Rev shouldn’t tweak it, but I don’t see a problem with being transparent about it. If you tilt it every time the player fails, tell them, and explain it away with game fiction if need be (“Your troops are now even more determined after their last defeat!”).

        I didn’t mean it as heavy-handedly as it came across, for sure. But some people don’t have much of their own brand to bring to the table if it’s something they have little experience with, regardless of age. It just seems kind of counter-intuitive to reinforce people’s misconceptions; it seems like a bit of a potential slippery slope /if enough games start doing it/. I’m not saying the sky is falling, just that I can see some possible harm, however minor, in something that seems unnecessary. Again, this is a heavily trend-based industry, and it seems to be one hell of a double-edged sword.

        Number Munchers made me care a helluva lot more about math than my times tables ever did. 🙂

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


My name is Anthony Munar, a computer programmer in Utah. I also play a bunch of video games every now and then. I talk and think a lot about them, but I never really solidify those thoughts anywhere, and writing is something I like doing, so I thought I'd do it right here. I don't intend to be high-and-mighty authoritative about what I say and I don't really have any sort of standing in the games industry. This is just for me to muse about games when I want to.

Naming a blog these days was harder than I thought. In calculus, the inflection point on a curve is where its concavity changes between upwards and down. So, maybe, the inflection pixel is the pixel which represents something that turns my opinion around on a game, like the pixels representing a beam cannon firing in FreeSpace 2, the pixels representing a flying car wreck in Burnout, or the pixels representing my own sentry gun holding off an army in Team Fortress 2.

Using the word 'pixel' in naming something game-related seems clichéd, so sorry about that.


March 2010
    May »

%d bloggers like this: