The Two Cultures

by Bob Newell

I am a former, somewhat serious tournament chess player (even though it was quite a few years back). I am also a semi-serious, limited-talent checker player. Then, at the end of January 2004, I learned to play backgammon, expecting a familiar experience on a different board.

I got nothing of the kind.

While I took to backgammon reasonably quickly, moving out of the novice phase and into the varying beginner stages in rapid order, the backgammon experience was totally new and differed from chess and checkers in more ways than I ever would have suspected.

Now, I must qualify my remarks. Today, given my geographical location, most all of my play of any game is over the Internet. I do some turn-based play and some real-time play, but very little live play, if any. And, my experience with backgammon is only of a few months' duration. Finally, I don't claim to be a good player at any of the games in question. But I do have decades of experience with board games of all types, in all venues.

So, for better or worse, here is a summary of my observations to date. These are just things as I see them. Feel free to disagree, make corrections, or simply blow it all off. But I think there really are "Two Cultures" and it came as a bit of a surprise.

I'll make my comparisons between backgammon and checkers, my two current interests. But the same applies to Chess, Go, Xiang Qi, and other abstract board games of strategy.

Determinism

The first thing a transplanted checker player notices with backgammon is the presence of dice: a randomizing element which throws our beloved determinism out the door.

In checkers, you're in control of your destiny. You make whatever move you wish, and your opponent may do the same. No such thing in backgammon. You move according to the possibilities allowed by the roll of the dice. In other words, there is a substantial luck factor.

Now, I found out quickly that backgammon is inherently very much a game of skill; a superior player won't win all the time, but will in the long run win much of the time. It's the short run, though, that makes backgammon both fascinating and frustrating.

In checkers, you might spend twenty or thirty moves building up a narrow edge. Slowly but surely, through careful play, the narrow edge widens. Eventually you bring through a hard-earned win.

In backgammon, you might have built up a substantial lead. You played carefully and correctly. Then all of a sudden your opponent rolls double sixes twice in a row and the game is over. You've lost. Your hard work and effort were to no avail. (It works in reverse, too, but we tend to forget those instances.)

Checkers is deterministic; backgammon is probabalistic. There's all the difference in the world, and I believe that it is this very difference that drives the rest of the cultural divergences that I'm about to describe.

Depth of Analysis

A common question asked of an expert checker player is, "How many moves ahead do you look?" While the answer depends on the situation, master checker players can visualize play quite far in advance, sometimes twenty or more moves per side, or forty or more plies, to use computer game programming terminology.

In backgammon, a lookahead of such range would involve an intractable number of potential situations. In checkers, only a single move is made per ply, and the range of possible moves can be easily enumerated.

But in backgammon, not only is more than one piece moved, as a rule, but the range of possible moves is described by the range of possible dice rolls, twenty-one in all. So each ply might involve a universe of something like 300 possible moves (in rough terms, seven or so checkers, times two possible moves each, times 21 possible dice rolls). A 3-ply analysis is already pretty large (perhaps twenty million or more positions to consider).

In addition, these potential moves have associated probabilities, a complication not faced by checker players. In backgammon, you might play the odds, making a certain move because a potential damaging reply is of low probability. Not so in checkers!

This makes the approach to play very different indeed, and calls into play a different skill set; in addition to visualization, mathematical calculation and intuition are required, plus a certain element of daring.

The Cube

It is said that the invention of the doubling cube has had a profound effect on backgammon, and indeed, skill with the cube is one of the most difficult aspects of the modern game. It is a device found in no other classic board game that I know about.

Could you picture a doubling cube in checkers? You get an advantage of some kind, perhaps an early king, and you say "double" and the opponent folds quietly? I can't picture it either. Why, then, does it work so well in backgammon?

I think there are two major reasons. One is the previously discussed probabalistic nature of the game. Making, taking, or passing on a double is in effect an evaluation of your odds of winning or losing, given the fact that dice are involved. A double when leading is often an attempt to bypass the dice, saying, I'm likely to win, do you care to risk twice the score on my being wrong? There simply isn't any such concept in checkers. (But doubles and redoubles exist in bridge, another probabalistic game.)

The second reason will be treated at more length below, but is based on backgammon being played in two modes, either "money" or "match." In a money game, doubling increases the stakes. In a match game, the stakes are likewise increased but in terms of match points rather than cash. Checkers is played a game at a time, for a point at a time, period. If there is a match, it will be a certain number of games, and a certain number of wins and draws will take the match. Each game counts the same. In, say, an 11 point backgammon match, doubling can make an enormous difference. With a couple of doubles, a single game could amount to four or eight of the 11 points, or in an extreme case, even the whole match.

Speed of Play

What is amazing to a transplanted chess or checker player is the speed at which backgammon is played (in head-to-head competition) and the impatience most players have with slow play. Tournament checkers might require 50 moves in an hour in modern play; vintage era checkers required just one move in five minutes. No such thing with backgammon. Averaging ten seconds on a move is considered slow play and five seconds per move is closer to the norm (and there are stats to bear this out).

What drives this lightning speed, which to me as a newcomer, seems to require the coordination and reflexes of a fighter pilot? It is very much part of the culture; I believe it has to do with a desire to get in a lot of play in a short time, due to the presence of money stakes (about which more soon). Time is money, literally.

To be sure, even checker play on the Internet is much faster than over-the-board or tournament play. This has more to do with the impatience factor of Internet users and players than with anything else; but head-to-head backgammon on the Internet experiences the double whammy of Internet impatience and the backgammon speed culture.

The Money Culture

Here we come to a major distinguishing factor between checker and backgammon play. Live (non-Internet) backgammon play almost invariably involves money, sometimes a great deal of it. Live checker tournament play may have cash prizes, but that isn't the point. Backgammon tournament players seek "equity" while checker players seek play for the sake of play (at least mostly).

Let's look at typical tournament entry fees. Checker tournaments run anywhere from $10 to $50 or so to enter, usually closer to the bottom end of that range. Backgammon tournament fees, beyond novice class, seem to be in the $100 or higher range, and more like $200 for "open" class events. Master class events can be $500 or $1,000 or even more. Tournament brochures talk about "90% return" meaning that 90% of entry fees are returned as prizes. What's going on here?

First, the tournament organizers are being rewarded for their efforts (that's where the other 10% goes, after expenses). Second, the players want a goodly sized prize fund, and they pay to get it. In addition to entry fees there will be side pools and auctions. What checker tournament has such things? What checker tournament organizer ever expects to make money for himself?

Live-play backgammon is all about gambling.

In fact, as noted earlier, non-match play is called "money" play. A wager of so much per point usually rides on each game.

To say that backgammon is about gambling is not a criticism; other games of skill, such as poker, are about gambling. In fact the correlation between probabalistic games of skill, and gambling, is quite clear.

This is not to say that wagers are never made on games of checkers or that cash prizes are not offered. They are, in both cases; and at the world championship level, a purse is put up to (at least) cover the expenses of the contestants. But the very nature is different. Checkers is simply not a gambling game.

For one thing, in checkers, a player who is noticeably better than another is likely to win every game. How long would one continue to bet when the outcome is very easily predicted? At least in backgammon, there are the dice, and a much inferior player will win once in a while.

There are side effects to the gambling and money culture nature of backgammon. Backgammon books sell for much higher prices than checker books even though backgammon books are in more plentiful supply. Why? It's about equity. How much do you expect to increase your backgammon earnings after you've studied a particular book? That will in part determine what you're willing to pay for it. A checker or chess book, on the other hand, is not likely to bring you very much playing income. I am told, too, that backgammon lessons from backgammon pros cost more than, say, bridge or Go lessons. While I don't have direct information on this, certainly the equity issue could drive such pricing.

Now, Internet backgammon play seems to be at least somewhat counter-culture. Much Internet play is simply play. There are some sites that offer play for money; I don't at present have a gauge for how popular this is. I would certainly imagine that the perceived risks in Internet gambling would limit this type of play. Of course, there are also a few Internet checker sites that offer play for money but my observation on repeat visits is that this type of Internet checker play is pretty much a non-event.

Conclusion

In the board game world, there are indeed two cultures. The split is very clearly along probabalistic vs. deterministic lines. Probabalistic games are in a world apart from deterministic games. The primary difference lies in the money culture of probabalistic play, but there are other differences in approach and analysis.

There are many instances of successful crossover play; and at least one backgammon pro is also a chess grandmaster. But if you have been a long-time deterministic gamer, such as a checker or chess player, be prepared for true culture shock when entering the world of the probabalistic board game, such as backgammon. I imagine the reverse might be true as well, although I have no direct experience there.

There is one question I have not approached, and that is the question of "Which is better?" That is in fact an invalid question because the answer is purely subjective. What is better is what you like best, whatever that may be. In my own case, I'm most likely to continue with Internet (non-gambling) backgammon, while largely avoiding live, gambling-related play. In fact, this is an adaptive strategy that may well suit deterministic to probabalistic crossover players, at least initially.

Of course, I won't give up my game of checkers, duffer though I may be. And who, having once experienced backgammon, could possibly leave it behind?

I think, in fact, that I've learned how to get the best from both of the Two Cultures.

Thank you to Bob Newell for his kind permission to reproduce this article here.
Bob's web site is http://www.chungkuo.org/.