2016 Candidates Moscow 10-29 March

[JustinHorton]

I think if he'd played the correct move 105. Rb2 he ends up taking the rook on move 114 so would have had a couple of moves to spare.

In the comments on the instant chess.com report a poster tells us:

Perfect play by both sides: Caruana versus Svidler
105. Rb2 Rh3
106 Bf2 Rf3
107 Bc5 Rf4
108 Bd4 Rf3
109 Rb4+ Ka3
110 Rb8 Ka2
111 Rb2+ Ka3
112 Re2 Rf4
113 Re1 Rxd4. Here the move count resets, well before 50 moves are up.
114 Kxd4 and it's a simple win with K & R vs K with 50 moves to do it.

[JustinHorton]

Meanwhile continuing my reivew of crosstables, Chess.com's is fine:



so why can't the official site produce anything comprehensible?

[Alex Holowczak]

Them’s the rules but it’s not altogether cool bananas. Not quite in the same territory as determining world cup qualification based on who gives San Marino the biggest kicking, but it’s on the way.

Football has changed that rule, results against the last place team normally don't count now as a tiebreak

Alas, no. You've confused two things.

Firstly, they've changed the rules so that the result of the head-to-head matches matters first, so you sort of create a mini league of the tied teams. At that point, you then go to Points, Goal Difference, Goals Scored and away Goals; and that determines who is placed higher. If two teams tied on points and drew their matches 0-0 against each other, none of these tie-breaks work. Only at that point do you only go to goal difference. It's against all of the other teams, but that is effectively only tie-break number 5.

Separately from that, the format means the top 2 teams in each group, plus the best 3rd placed team in all groups, qualify automatically. Therefore, it becomes necessary to remove the team that finishes last from the team's record, but only because some groups have 5 teams, and some groups have 6 teams. So the rationale is not to avoid the possibility of Team A in a group thrashing the team that finishes last by more than Team B in a group; but because Team A might thrash the team that finishes last in its group, but there's no corresponding team to thrash in some of the other groups.

--

The thread went on to discuss alternative formats.

The Premier League darts competition works as follows: 10 players play a single APA, and after 9 rounds, the bottom 2 players are kicked out. All results carry forward, and the remaining 8 players play another APA. After 16 rounds, the top 4 qualify for the knockout playoffs.

I thought one idea might be to tweak that format slightly for chess. 10 players play a single APA, and after 9 rounds, the bottom 4 players are kicked out. You only carry forward the results against the other 5 players (because of the possibility for qualifying players to have more whites than blacks - in darts this is mitigated by playing 12 legs per match, so the corresponding advantage - the right to throw first - is mitigated). Then the 6 remaining players play the other 5 players with the colour they haven't played. Thus after 14 rounds, you have a final table containing 10 games, one of each colour against the 5 who made the cut.

This ticks a few boxes:
(1) You avoid Jon Bryant's concern about two players at the bottom who don't care about the result having an effect on tie-break, or the impact of how certain players do against the tournament's stragglers.
(2) You add some mid-tournament excitement as to who may or may not make the cut.
(3) It still only needs 14 rounds, so you don't make the tournament any longer. The current pattern of play with a rest day after round 9 is also convenient in the overall schedule.
(4) You add two more spots in the tournament, which might be useful.
(5) You may be able to increase the prizes for the winners if you take the golf approach whereby you don't get any prize money if you don't make the cut.

So that's what I'd do if I had any influence in this sort of thing at all.

[IM Jack Rudd]

I can think of a few interesting features that tournament format has, one of which is that it's possible to get a 9-5 colour split in actual games played.

[Alex Holowczak]

I can think of a few interesting features that tournament format has, one of which is that it's possible to get a 9-5 colour split in actual games played.

Indeed ... but so what so long as it's 5-5 in the bit that counts?

[IM Jack Rudd]

I can think of a few interesting features that tournament format has, one of which is that it's possible to get a 9-5 colour split in actual games played.

Indeed ... but so what so long as it's 5-5 in the bit that counts?

Because all of it counts. You have to actually play those games against the back markers, you know.

There are more subtle issues, as well, such as the issues arising when there is a tie just on the cut between two players who'd carry forward vastly different results. You have the 4NCL as an excellent example of the strange results this can give in practice.

[Alex Holowczak]

I can think of a few interesting features that tournament format has, one of which is that it's possible to get a 9-5 colour split in actual games played.

Indeed ... but so what so long as it's 5-5 in the bit that counts?

Because all of it counts. You have to actually play those games against the back markers, you know.

There are more subtle issues, as well, such as the issues arising when there is a tie just on the cut between two players who'd carry forward vastly different results. You have the 4NCL as an excellent example of the strange results this can give in practice.

It doesn't all count, because you'd be 5-4 or 4-5 over the first 9 rounds - which of course would be better if it were equal - and then 5-5 in the final 6 section.

I don't see the subtle issue of teams carrying forward different results as a problem at all. That is just there to avoid having to actually play them twice over, similar to the 4NCL problem.

[IM Jack Rudd]

It does all count, unless you think a top player thinks his FIDE rating is an irrelevance.

[Alex Holowczak]

It does all count, unless you think a top player thinks his FIDE rating is an irrelevance.

I mean count towards the tournament, not the player's rating. I'm not much interested in the impact on the rating. There are some very imbalanced colour sequences in team events, for instance.

[IM Jack Rudd]

Here's a fun one. You do this format with the last London Chess Classic. Which of Caruana and Adams goes through?

[Chris Rice]

Caruana seems to be taking a bit of a hammering on social media for missing the win but this is such a difficult ending to play. This clip though from a Carlsen-Svidler match with the Arkell ending shows how it can be done under severe pressure. https://www.youtube.com/watch?v=HY3SM3GKjd8

[MartinCarpenter]

If nothing else he'd presumably basically given up by then - if someone has defended this ending for 30+ moves, no real reason to expect them to suddenly slip!

He had of course done really rather well to get to R & B vs R.

[NickFaulks]

I missed most of yesterday. Do I gather there were confusing announcements out of Moscow re tiebreaks?

Fortunately, it now seems clear that everything will be decided on the first tiebreak, so no problem.

[Mick Norris]

chess.com report

TWIC

Chess Mind

I think the advantage of the slight possibility of Vishy influencing the outcome, is that Karjakin can't be certain a draw will do, so he might not risk playing a drawing opening as it were - however, Caruana similarly might not know whether to go all out for it from the opening, although to be fair he does seem to have been trying all tournament based on his opening choice

Anyway, just over 3 hours until the start of play

[JustinHorton]

chess.com report

Giri tends to take a piece or pawn that he took from his opponent with him, and play with it while walking around. Many times in his career (including once here in Moscow) he arrived in playing halls, finding a piece from the previous round in his pocket.

What?