Clint Ballard is on a campaign to end draws. While we sympathize with his goal, and hope he finds a way, his current system isn’t it.

He’s created yet another scoring system that eliminates the zero-sum aspect of chess. In this system Black wins are worth 3 points, Black draws are worth 1 point, White wins are worth two points, and every else, including White draws, are worth nothing. The aim of the system is to encourage White to play for a win (note, it does nothing much to encourage Black to do so, but that’s not relevant to my point).

On the surface of it, it’s a good try, and we applaud his efforts and his willingness to experiment. But we are skeptical. Why? Let’s do the math:

Current GM level stats are 28+17-55= for White. Running the statistics through his point system we get White socring on average .56 points per game, while Black scores on average 1.06 points per game, giving Black about a half-point edge in tournaments where colors don’t balance (all 5-ss weekenders fit that category, as do all other odd round total events). But we haven’t forgotten the goal of this system is to reduce the number of draws.

We’ll look at two hypotheses and see how the numbers change. First, let’s assume that by buckling down on both sides, the pile of draws is divided evently between the three possible results (draws will still happen, because sometines they’re the only way to avoid losing). So now we have 46+37-18= for our baseline, and that yields average scores per round of .92 for White, 1.29 for Black, still an edge to Black of about a third of a point in odd-round events.

Now let’s assume Black picks up a disproportionate share of the formerly drawn games. I consider this the more likely outcome, as every chessplayer understands what usually happens when they try to win a drawn game. So we’ll divide the draw pie into 5 slices, in which White will win 20%, Black 60%, and the remaining 20% will still be drawn. This makes our baseline 39+50-11=, and that yields .78 points per round for White, 1.61 points per round for Black, giving Black a whopping .83 point edge in an odd-round event.

There is, of course, a point at which the scoring system becomes neutral. Will Mr. Ballard’s event find the “sweet spot?” We don’t know, and we welcome the chance at additional data. But we doubt it.

And we also believe that even if it works, it will be, at best, a temporary “fix.” Chessplayers are, after all, intelligent gamers, and it will be just a matter of time before they would begin to “game” this system, quite probably by throwing games to one another in order to arrange the final results, rather than playing quick draws as they do today. Would that be an improvement?

We at The Chessmill wish you luck, Mr Ballard, and truly hope you’re on to something. But the numbers do not make us optimistic.

This entry was posted on Sunday, August 27th, 2006 at 9:20 am and is filed under Chess News, General. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.