Monday
Aug092010
Endgame Blunders at the Rapid World Championship
Monday, August 9, 2010 at 3:48PM Two games from this weekend's Grenke Rapid World Championship caught my eye as examples of what can happen to a player who is pressing for the win and gets excited about his chances - so excited he loses his sense of danger. Have a look.
Reader Comments (6)
And what is your solution for fixing 'stupid moves on DGT boards' problems?
It is hard to expect from arbiters that they will be able to carefully check all the games in rapid tournament. Usually the break is only long enough to reset the boards, check the cables and store PGN. And you cannot expect the software to prune the last move if it doesn't make sense.
DGT software has a lot of more or less annoying bugs (including being easily fooled by a quick piece exchange) but I don't see a good solution here.
Here's one very simple solution: put a three position switch on the side of the board, possibly with a little "confirm" button. It can be small enough not to ruin the aesthetics (and anyway, with wires coming out of the board and clock that doesn't seem like such a huge factor anyway) and from a technological standpoint it's trivially easy.
First (and obvious) problem is that your solution will work only for new boards, and with price $400 (or something like that) per board, nobody will buy new one just for that. So, only software solutions are realistic.
Second, why moves like Bc8 are registered? Exactly because players forget to set the result (by putting kings in the center). They will forget to do it using the switch as well, and you will get those Bc8 moves again.
One thing that can be easily fixed is removing leading Ke4/Kd4/Kd5/Ke5 when it is matched in a second or two with putting the other king nearby. But that accounts for a small percentage of DGT-added moves (king has to be on one of the adjoining squares and move has to be legal).
Of course, suggestions are welcomed. I did a DGT relay many times and would love to see some improvements. For now I manually checking each game in standard tournaments, but in rapids there is no time for that.
Yes, I realize that ...Bc8 is a problem, but it's not one I'm too worried about. That has happened pretty infrequently in my experience (it's the first example of the sort I can recall), but fake king moves to the center are recorded constantly. It's a small percentage, sure, but it's far from trivial.
I recognize that the switch is only a solution going forward, but I've been proposing this to the company for at least six years, going back to a time when there were far fewer boards out there. There are software solutions available too, like this: when K-center square is legal, then it is impossible to use it to record a game result. The following alternatives would be sufficient: kings on corner squares (h1/a8 = White win, a1/h8 = Black win, h1/h8 = draw), and if that's beset by legal problems as well then use the e1/d1/e8/d8 quartet. Necessarily, one of the three quartets will be available to the kings without inserting any ambiguous moves.
The only reason I mention the switch first is that it's idiot (= arbiter) - proof.
I saw plenty of move-to-the-first-rank errors, it is a surprise to me that you haven't seen it more often. If the players forget to set the result, any legal piece-to-its-beginning-square is recorded correctly. Probably the reason is that you follow mostly top-level tournament when players do remember to set the kings. In open tournaments the first-rank error is much more common. On the other hand, I am surprise that in top-level tournaments arbiters forget to check the games.
One more very common source of errors I forget to add is players starting to analyze without setting the result.
Also, there is no idiot-proof solution here. Your solution is not good as players will set the king the old way anyway, the only hint for arbiters is that there is no result set.
The best I can think of is an option not to record the moves after the clock was stopped. Of course, the players sometimes don't stop the clock either, but from my experience I can tell that this will cover at least 90% of cases.
Also, there are worse things that final spurious king move. See this game from World U20 Championship (Tomczak - Blomqvist, round 11):
1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bc4 e6 7. Bb3 b5 8. O-O Be7 9. Qf3 Qb6 10. Be3 Qb7 11. Qg3 Nbd7 12. f3 Nc5 13. a3 Nxb3 14. Nxb3 O-O 15. Rad1 a5 16. a4 bxa4 17. Nxa4 Qc6 18. Ra1 Qxc2 19. Nd4 Qd3 20. Qf2 Qa6 21. Nb3 Nd7 22. Nb6 Nxb6 23. Rxa5 Qb7 24. Bxb6 Rb8 25. Be3 Ra8 26. Bb6 Rb8 27. Ba7 Qxb3 28. Bxb8 Qxb8 29. Qa7 Qxb2 30. Ra2 Qb3 31. Ra3 Qb4 32. Ra4 Qb5 33. Ra5 Qc6 34. Raa1 Bg5 35. Rfd1 d5 36. e5 Qc3 0-1
AZ: As I said, the idiot-proof solution is to add the switch for future sets. Still, if arbiters try and fail to execute the score because it doesn't function the old way, they'll eventually learn or are at least more likely to. As it is now, there's nothing to prompt them.