| |
BGonline.org Forums
Effect of Stalemate=Win rule
Posted By: Nack Ballard In Response To: Effect of Stalemate=Win rule (Timothy Chow)
Date: Sunday, 9 December 2018, at 6:35 a.m.
If there are fewer drawn positions and the players are not steering for what remains, it seems obvious to me that there will be fewer draws with the rule change. That is why I said "you seem to be assuming" they are steering for the draws (trying harder to achieve them); otherwise as high a draw percentage would not be maintained. That said, your intended meaning in that paragraph and my interpretation of it seems like a side conversation that would distract from the main issue.
My main takeaway from your previous message was this statement:
The set of drawn positions will indeed be noticeably reduced. But chess will still be fundamentally a drawn game, and you'll still need a large advantage to win a game, which will happen at about the same rate as currently.
To which I replied...
We seem to agree that the set of drawn positions will be noticeably reduced, but we arrive at a different conclusion. To my mind, it logically follows there will be fewer draws.
To which you replied...
I'll grant you that, but I think that at the top level, this particular attempt to reduce the number of draws will result in an insignificant change in the draw rate. It will change the details of how players jockey around when they're near the draw/win boundary, because there are new shoals to avoid, but it won't fundamentally change the draw rate of top-level games.
I have a few comments. First, it is precisely the top level on which I've been focusing (in case you thought otherwise). I believe that is where the greatest change in the draw percentage will take place. The lower the level (still assuming near-equal competition), the higher percentage of decisive results there are; with a smaller draw window, the effect of the rule change there will be smaller. (I'd be surprised if you don't agree with that direction of the effect, though your phrasing makes me unsure.)
Second, I'm not sure on what evidence you are basing your opinion, or if you're just speculating. Also, jockeying to avoid shoals is a game that two people can play (contrary to each other's interest), and in any case sounds like an obtuse justification to somehow end up with almost exactly the same percentage of draws despite what we've agreed is a noticeably smaller window.
I used in part the actual end-positions of the mere-12-game World Championship match on which to base my opinion, including an actual drawn position that would have been a win, and cited an example variation to make my point. Also, even at my level I remember being on both sides of many a drawn endgame that could easily be reduced to K+P vs K, which would be a win with a stalemate=win rule change. I've also done enough endgame analysis that I think I have a pretty good feel for the frequency.
More closely examining Carlssen's "won" (by future stalemate) game 3 position: Sure, Caruana would not have gone into that exact variation, but I did not see another way to force a draw, including Nxc5 and then going after the KRP, which holds out longer but gets the N trapped), and the engines did not show any drawing alternative to eventually trading the knight for the last non-rook pawn.
Even if there was a path even earlier in Game 3 that led to an either-way draw (which there may well be), there are plenty of counterexamples of positions that are currently given up as drawn instead of being played out to reach the pivotal K+P vs K reason that I discussed (or other pivotal position). Games 4 and 5 are potentially such positions, though if they are stalemate=win wins, I'm not good enough to prove it), and there are countless examples from GM tournaments.
Tinkering with small rule changes...
That's just it. At the top level, stalemate=win is not a small change at all. It only sounds small because one almost never sees games actually played out to stalemate. They are agreed drawn before that happens; with stalemate=win, such agreed draws would become resinations as soon as the writing is on the wall. (Just like checkmate rarely occurs at the top level; resignation occurs once impending checkmate -- however far ahead -- is recognized.)
I do agree that we won't know with absolute certainty unless and until the rule is put into practice.
Let's take backgammon. You have seen, and suggested yourself, many many proposed rule changes that are designed to increase the chances that the more skilled player will win more often. And you know that unless you make a drastic change, it doesn't ultimately affect the win rate very much. The same is true in chess.
Not even close, IMO. Chess is deterministic. Backgammon is not. In the latter, luck overwhelms the result, even with what may seem like a major rule change.
Nack
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.