Wednesday, February 07, 2007

Gaming by the Numbers

As I write this, I’m reminded of the scene in Dead Poet’s Society where the teacher mocks the analysis of poetry using numbers and a graph to represent a poem’s quality. However, as I played a game of Louis XIV last night, I was struck by the idea of trying to analyze a game using a numerical scale for a number of different axes. I was contemplating that Louis XIV is not confrontational, and thus not of interest to several players in our club as they are very much into confrontation. As I thought about other games I realized I might be able to rank games onto a scale of 1 to 10 (or -5 to +5, whatever) to signify how much direct player confrontation occurs. I use the word confrontation, as cooperative games like Lord of the Rings allow some player interaction, but it isn’t one player trying to take advantage of another. Cooperative games of this nature would surely fall on the far end of the confrontation scale. Zero sum games where anytime one player wins another loses, like wargames, would then be on the high end of the scale (10). I’m sure the topic would be great fodder for gamer discussion on just how confrontational a game might be. I’m sure this is not a new discussion at all, but was got me interested enough to write about it is the idea of looking for additional game properties that might also be laid out on an axis (1 to 10) to further categorize games. If such a system could be constructed, gamers could figure out what their preferences are, and then try to match them to games that closely fit their style. If a player has no preference for a particular axis (say, they think confrontation is fine, but they don’t need it to have fun) then they could just ignore that axis when trying to match a good game in “game space”. Here is a half-hearted attempt to try to quantify independent attributes about a game. This attempt to analyze games is not to weed out the stinker games from the good, but to compare solid designs to each other to match an individual player’s likes and dislikes. Games with rules universally regarded as unplayable should be considered “broken” and not placed on the chart at all.

I’ve already talked a little about this, but this is simply a ranking on how much players are able to get in each other’s way. Pure cooperative games like Vanished Planet would score a 1, while more vicious games like Diplomacy or a wargame would score a 10. Note that some people may find their preferences to lie on the two extremes. While Settlers isn’t very confrontational (say, a 3 or 4), the few times it occurs (via the Robber), may make the game less fun. Thus, a player might be fine with confrontation, as long as it is the expected norm for the game – they don’t like to feel singled out. Games like Can’t Stop or Goa would also be fairly low on the scale with Puerto Rico scoring middle of the road due to strategic role selection. I’m the Boss would be a fine example of a highly confrontational non-wargame.

A perennial favorite for discussion, clearly the element of chance is anathema to some and a welcome injection of “fun” for others. It is also fairly easy to measure. To judge this, we first have to rule out any effect due to other player’s decisions – just because one opponent plays poorly enough to give a second opponent an advantage doesn’t mean the game contains a chance element. Since this is a “game-y” audience, lets put 1 as a completely random game like CandyLand, while a 10 would be something with no randomness at all – like Caylus. I also want to keep the ranking separate from the player’s ability to mitigate the fickle hand of fate. A game that is pure luck with no way to manage the randomness is probably not a very good game, and could be taken care of with a separate category. For example, Poker is a very random game – if a single hand is played. So, Poker would be a ranked very low on this axis, say a 2, since chance plays a big part of the game moment by moment.

By now all the poker players are up in arms. Hopefully the addition of this category will make you happy. Control represents a player’s ability to control the outcome of events within the game. In the long haul, a poker player can manage their resources, read people’s bluffs and so on… Thus, poker should fall somewhere in the midrange of the control axis. Meanwhile, CandyLand still remains at a measly 1. Other games may score high in the Chance category (having little randomness), but still have very little control. An example would be a multiplayer game where the game situation is always changing so players can only respond tactically to a given situation and it is difficult to plan out any long-term strategy. Citadels may be a good example of a game with only a bit of randomness, as one can’t affect any other player directly but it is still hard for a player to have a high degree of control of the situation.

Theme often comes up in game discussions, and it clearly affects many players’ opinions of a game. This category can also serve as double-duty, since a game with poor components will also tend to have a poorly executed theme. Some gamers may find less theme acceptable and thus be very fond of abstract games or pasted-on themes, while other gamers find they do enjoy a bit more story behind their game. Pure abstracts might score a 1 or 2 simply because they don’t try to advance a theme. Meanwhile a game with even a pasted-on theme may rank as high as a 4 or even a 5 if it is executed in a nice manner. (Again, sort of adding in components to the mix as I feel quality components can sometimes make up for a drier theme as I will help the theme along if I enjoy the components enough.) Games like Tikal would be medium-high possibly a 6 or 7 as it is a fairly abstract game but the components help to carry the theme forward. Around the World in 80 Days has a pretty good theme and would come in at an 8 or 9. Die Macher is a dry game that still seems to pull off gameplay that does feel as if one is negotiating to win elections – I’d give it at least a 7.

Time to Play
Taken by itself, the length of a game is probably not a great measure of a person’s preferences. Sure, some may prefer to avoid particularly long games while others find a game that is too short really won’t have enough time to develop. However, the length of a game is clearly significant to many players, as reviews will often mention a game to be too long for its content or too short to develop. Taken as just one attribute of a game and then compared with others can make it a valuable measure of a person’s preferences. Perhaps one player is willing to play a game with a great deal of chance, as long as it is short. Similarly, one might prefer a game with a longer duration to have more control, and not be a series of independent exercises in optimization.

I’m sure the list of attributes could go on and on – at some point in the future I may attempt to make a somewhat definitive list. However, I think even the few examples I present may provide food for thought. Remember, each category is not a ranking of better or worse, just different. Some may prefer certain ranges while others prefer another. It would be interesting to gather a pile of data, having players score a set of games according to the above criteria, and then compare the results to ranking on the Geek to see if there are definite trends in opinions about games. The results would probably only reveal things that are already conventional wisdom.

How about you? Think there are any defining attributes that I missed? Is the whole concept impossible? There will always be games that come along with a mechanism that some find intolerable, blind bidding for example. That can’t be properly contained in my rickety scoring structure (although it could affect the games overall Control score). However, putting in additional on/off binary flags for a particular game mechanism (or attribute) seems to be moving further from my ideal dream of each gamer having their own preferred n-dimensional game space.


huzonfirst said...

This is all interesting, Matt, but I suspect such ratings would mostly show how differently people view the various games. For example, I consider Louis XIV as fairly confrontational (it's hard to view it otherwise when an opponent smacks down three pieces at the site you placed two on earlier). The infrastructure building portion of Goa may have no confrontation, but there's a good deal of it in the auction portion, particularly when you determine which tiles get auctioned off. Similarly, I'd say Tikal has a much stronger theme than 80 Days. And so on.

Sometimes it's hard enough just getting the facts straight on games, as well as a fairly straightforward number like a rating, without having to worry about the more subjective and less defined measures. As I say, it's interesting, but the number of disagreements it would foster might make it mostly meaningless.

Anonymous said...

I attempted something similar and posted it on the Board Game Designer's Forum at BGDF a while back. I never really did anything with it since then, but I thought I would let you know as it appears we had similar lines of thought.

Dr. Matt J. Carlson said...

Good comments.

I realized later that Louis XIV has a lot of confrontation, and so I think you are correct.

However, I'd disagree with your assessment of Goa. I find determining which tiles get auctioned off to be only a very small portion of the game. True, the first player can have a small amount of control and the 3rd or so player sometimes can mess with the last player, but otherwise I find most categories of tiles to be fairly easy for people to reach. I would allow that there is conflict in the auction portion. I just haven't seen tile selection process be all that confrontational. (Most people can use most of the options, even though there are a few juicy tiles out there and you might try to prevent someone from getting to select one. Unless everyone tries to do so, usually _someone_ can tag a juicy tile.)

As for the value of the ratings, I think it isn't anything to crow about. However, the same could be said for ratings in general on the BGG. The idea behind them is that they can become statistically significant if enough data points are used. I would expect those who find Goa confrontational will tend to rank more games confrontational than other players. Thus bringing most games' confrontation ranking slightly higher. This would do no harm to the concept... games would still be ranked from less to more confrontational. A given player would then adjust what "level" of confrontation they care to play....

Smatt said...

Hi Matt,

Let me first tell you that I use an informal ranking system that you described for all of my customers. It addresses the problems mentioned by HUZONFIRST, as the information obtained is always reflected back to the customer.

Basically, when trying to find a game for a customer (usually one who is overwhelmed by the selection), I try to establish a gaming history. I ask what games they like, played last, played most, etc. From this line of questioning, I get low/med/high strategy preferences, often low/med/high time preferences, and always low/med/high commitment to learning the rules. When I have that information, I can then give some suggestions based on games that fall under the same categories (at least to me). In this way, I usually score big because I'm able to send them off with a game that works well 99 times out of 100. Many times, people have come back and thanked me for my suggestion. I don't think they really understand that I'm profiling them and the games they like, just that I recommended a fun game, but whatever.

One couple was in disbelief that I could recommend a game based on such little information (they told me two or three games they liked). I explained myself, then showed them Blokus. It met all their requirements (at least in my mind): mild strategy, low time commitment, few rules to learn. As they left, the man said to the woman, "I really like that game." They didn't buy, but Blokus appealed to their sensibilities.

Anyway, nice article. Keep up the good work.

huzonfirst said...

I guess we're looking at a difference in playing styles with the tile selection in Goa, Matt. We really do try to avoid giving people juicy tiles to auction off, since money supply is such a big part of the game. Particularly evil starting positions are always grudgingly applauded. I just find the auction half of the game to be very important and there's a considerable amount of interaction there. I don't view Goa as anything near group solitaire.

Dr. Matt J. Carlson said...

I enjoy Goa immensely. I will be perfectly willing to acknowlege the auctions can make or break a game for a good player.

However, my point on the tile selection is that I find the mechanism very limiting for player choice. Sure, there are times where a tile selection round can be set up to be particularly vicous, but I find that to be the exception rather than the norm. With what, 8 auctions in the game, I would say that at most 4 or so give a player significant choice to affect other players. Those typically occur when one is the first player or next to last. The layout of the tiles typically allow sufficient freedom for players to consistently make reasonablly decent choices. (With a group effort, the juiciest tiles sometimes can be avoided, and in very rare cases can a first player set themselves up for one later...)

Anonymous said...

This sort of ranking might be useful if people agreed on how games ranked. But your rankings are very different from mine. For example, I consider Louis XIV to be one of the most confrontational multiplayer games that I enjoy playing. If I play 3 cubes on a space where only the player who plays the most cubes gets a reward, and you play four cubes on that same square, rendering all the turns I spend playing the 3 cubes completely useless, this feels to me about as confrontational as it gets. How could it be more confrontational, unless you got to reduce my cubes to sawdust or punch me in the nose after you achieved a majority?

And one of the reasons I love Die Macher is that it is so strongly themed. I feel exactly like an incredibly slimy politician when playing it. War of the Ring is the only game I can think of offhand that I find to be more strongly themed than Die Macher.

Number of players is a very important axis you left out, which can interact strongly with the other axes in determining preferences. For example, I don't like extremely confrontational multiplayer games, but I have no problem with confrontation in two-player games.

Dr. Matt J. Carlson said...

Hmm, Die Macher has a good theme, but I find the scoring to be somewhat abstract (manipulating the cubes etc...) It has a moderately strong theme but not as high as I would rank other games. Yes, there are lots of elections, but how they are affected are somewhat abstract to me.

In Louis XIV, if you spend 3 and someone spends 4. Yes, you didn't win, but the opponent loses all 4 pieces for next round and you get yours back. (Thus they have to spend more than a single action to get them back. Adding up it took more than 2 to place them anyway. So the opponent spent almost 4 actions to win. Whereas you spent 1 and lost - since you got your pieces back.) A pretty good trade. I need to think about Louis XIV more, somehow I find that competition more palatable. I guess that might be the key - it is a competition and not a confrontation. If I lose an opportunity, I am not also penalized for it. I just fail to gain. Whereas in a wargame, if I lose a fight I'm also typically penalized by losing units as well.