Kasparov agrees on Mickey Mouse

[John Fairbairn]

As long-time forum members know, I've always been critical of Mickey Mouse time limits - blitz or close-to-blitz games.

Korea led the way, obviously with an eye to tv audiences, but many pros have shared my concern and there has over the past year or so been a shift back towards longer time limits.

But it was interesting to me that ex-world chess champion Garry Kasparov put a figure on the decline in performance with MM limits. He was, of course, speaking after the end of the chess world championship here in London, which had to go to rapid games to break the deadlock in all-drawn classical chess games.

He said:

Carlsen’s consistent level of play in rapid chess is phenomenal. We all play worse as we play faster and faster, but his ratio may be the smallest ever, perhaps only a 15% drop off. Huge advantage in this format.

Two points there:

1. I'm surprised to see such a large drop-off figure - and of course he is implying it is much larger for other chess pros.

2. These games were at "rapid" format which is a bit slower than "blitz." Presumably at blitz the % drop-off is enormous.

Unless fans actually prefer this crap-shoot format, which would be beyond my ken, I think there's a lesson for go there, though - as I said - it seems already to being heeded.

[Javaness2]

Media coverage of the World Chess Championship was actually really rather good in some quarters. I am thinking particularly of theGrauniad here, who had a slick chess app embedded in their live page. Via Olimpiu Urcan I learnt of an amusing remark made by Magnus in the press conference "As for the opinions of Garry [Kasparov] and Vlad [Kramnik], I mean, they are entitled to their stupid opinions. That’s all I have to say." Of course, he wasn't talking about the time limits, only about the offering (taking) of draws. Nobody, surely nobody, is a fan of the Yudina 'Armageddon' System. The question is what alternative can FIDE offer in the modern world.
* An extended match length presumably requires more money.
* The condition, of champion retains the title if the match is a draw has seemingly fallen out of favour.

I don't know where Kasparov gets his numbers from, but I would argue that 25 minutes with 10 seconds increment isn't exactly Mickey Mouse, surely Blue Peter would be more apt. What he is right about is that Carlsen has excellent nerves and instinct.

[HermanHiddema]

The way I'm reading the news around the chess world championship, many people there are in fact clamouring for shorter time limits after the main event saw 12 draws in 12 games. I've not seen Kasparov's opinion on it, but I'm not sure that his comments on the lower level of play in rapid automatically mean he disapproves of them.

[dfan]

What units is "15% drop off" in?

[bernds]

I think it's a stretch to interpret Kasparov's comments that way.

Chess in particular is in a very different situation from Go. Openings are analyzed to a point where a large portion of top-level games is just Stockfish playing against Stockfish, when it should be human against human, and the most likely outcome in long time controls is a draw. I believe we just saw the first chess world championship in history where no games in classical time controls had a decisive result.

So maybe the long games had a high quality, but also little risk-taking, because winning is hard. In contrast, the shorter games showed very clearly who the best player in the world is. So maybe they didn't produce perfectly games that will be remembered for eternity, but Magnus in particular is quite capable of delivering a performance in short time controls that is worth remembering. Whether it's clobbering Fabiano 3-0 in long-ish rapid, or destroying the field in Lichess bullet arenas, there is no doubt who is the strongest chess player in the world.

In fact, I'd go so far that if you want to see brilliant tactics on the board in chess, you need fast time controls, because then the chance that only one player sees them is higher.

In Go, it's more fun to watch long games because engine lines probably still don't occupy the entire first half of the game, and no one is just going to simplify the game to a draw.

[bernds]

In fact, most commentators seemed to think the twelve draws were not very good for the game. After game 12 there was Svidler's "Press F to pay respect to classical chess", and Grischuk's "This is the cherry on top of the coffin of classical chess".

[Gomoto]

In go I am definitly more interested in pro games with longer time controls. Even Amateurs play in Europe with 2 hours main time for each player at the yearly congress.

[topazg]

As a spectator, I'm always more interested in interesting games, particularly complicated ones that lead to a decisive result. In chess, not only is the opening well enough theorised that games at the top level end up being "draws from the beginning to the end" or "someone surprised someone else to get a marginal advantage that they may or may not convert over the following 4 hours", or a single relatively large mistake turning a drawn game into a non drawn game.

It's rare in classical games for someone to play a super sharp and risky line that looks complicated and interesting, because given an hour to analyse the complicated chunk of the next 10 moves, the opponent may find all the good counter-resources and end up ahead, so both often just play straightforward lines where it's a draw unless one player errs. In rapid and blitz this is much rarer. I personally think blitz is fast enough that straight out errors tend to take over a bit more, but my biggest criticism of the format is that it's very hard to follow what's going on when players are smashing the clock on 10 seconds left.

Rapid has always felt like the ideal format to me - short enough to fit multiple games in (which IMO is always a plus for spectators and fans), and long enough to avoid the most obvious of the blunders. If we want very high level chess, we stop watching GMs and watch AlphaZero play Stockfish again or whatever, and give them a day analysis per move. It'll be boring as hell, but the quality will be high.

If we want characters that can blunder, fight, show emotion, and dazzle us with the occasional brilliancies in otherwise high level enough play to be far better than us mortals, then rapid chess does this just fine. People can have favourites, see a steadily altering tournament +/- scoreline (as opposed to minor centi-pawn adjustments of a single game move by move). For what it's worth, I also watch most of the top level classical tournaments anyway, but I can't sit down for 6 hours straight, so I'm checking in while doing work, housework and all the other things RL makes me do. Rapid I can "check in" to see individual games and replay for interesting bits, and watch entire games when I have time.

Mickey Mouse is a very pejorative term to use. Kasparov certainly didn't, his entire tweet that you quoted was simply remarking on Carlsen's ability to maintain playing quality at shorter controls, not criticising shorter controls. In fact, all of the events he's participated in himself more recently that have involved other GMs have been shorter time controls (rapid and blitz), so he's obviously not that offended by them.

[Uberdude]

After Carlsen offered the draw in the final classical game and before he won the rapids, Kasparov said:

In light of this shocking draw offer from Magnus in a superior position with more time, I reconsider my evaluation of him being the favorite in rapids. Tiebreaks require tremendous nerves and he seems to be losing his.

Was he being ironic, or wrong?

P.S. Something else that struck me in reading about the 2018 World Chess Championship was the relatively equal split in prize money: if it was won in the 12 classical games the winner takes 60%, loser 40%, if after tiebreaks as happened 55% vs 45%. I know we've talked about prize splits for Japanese title matches here before, winner generally got around 3-4 times the loser IIRC, and Lee Sedol apparently pressed for the highly asymmetric 5,000,000 CNY vs 200,000 CNY (ratio 25:1) for his match with Gu Li.

[jlt]

Carlsen’s consistent level of play in rapid chess is phenomenal. We all play worse as we play faster and faster, but his ratio may be the smallest ever, perhaps only a 15% drop off.

Does the "15% drop off" refer to the percentage of moves that match the first choice of Stockfish?

https://chess-db.com/public/enginefirstline.jsp

[Charlie]

In this case, I don't think that Go and Chess are really comparable.

After reading a lot of coverage of the Chess Championship, I get the impression that knowledge of their upcoming opponent, databases and supercomputer research for half a year before the match all combine to create a situation in which the "classic" part of the match is a "first to blunder" affair -- one blunder equals one loss and, after that, your opponent only needs to force a draw for the remainder of the games. The result is a style that it excessively conservative and pretty boring. Chess fanatics might disagree but I can't even see that the game records from that are very useful -- surely the supercomputers make them redundant curiosities rather than study material.

For Go, the draw rate is basically zero (outside of Europe in which the Komi is seemingly determined by a random number generator. I've played with 6.5, 7.0, 7.5 and even 6.0 just this year -- all under Japanese rules) and I don't think one-sided matches are all that common. Certainly, longer time controls don't create a situation where every game goes to white by half a point.

Even the advent of strong A.I. has not changed the situation very much. Perhaps it is because pros have learned new Joseki, one or two new Fuseki configurations and new concepts from A.I. but have not really reached a stage where they can be said to be playing directly from a super-computer's opening book.

I predict that A.I. might affect Komi but that is all. In a game with the high branching factor of Go, I do not believe that the human brain is capable of digesting super-computer lines so comprehensively as to be robust to disruptive tactics by one's opponent. As black, one might gain an advantage in a single game by playing a well-researched opening but, over the course of a match, it will always be possible for white to approach a corner or pincer something and deny it.

As long as this remains true, I think that human Go will remain creative no matter how much time is on the clock. As long as the game is creative, more time can only be good.

With regard to the outcome of game 12, I find it hard to have anything but respect for Carlson's draw offer. Perhaps he was tired, perhaps he had lost concentration, perhaps he had simply lost the will to fight -- whatever the case, his admission that he could still lose that game shows remarkable self awareness and his willingness to hand Caruana a second chance showed courage and confidence in his rapid skills.

People laugh about "Get Strong at Resigning" but, honestly, I learned my lesson this year: "1"s and "0"s on the wall-list are secondary to one's mental state and a tournament spans more than a single game.

[topazg]

After Carlsen offered the draw in the final classical game and before he won the rapids, Kasparov said:

In light of this shocking draw offer from Magnus in a superior position with more time, I reconsider my evaluation of him being the favorite in rapids. Tiebreaks require tremendous nerves and he seems to be losing his.

Was he being ironic, or wrong?

Wrong I think. A lot of chess commentators were highly critical of turning what was a solid-but-not-decisive advantage into a draw offer. Magnus in interviews and the press conferences has now said a number of times that he'd already decided for the 12th game that if the game was even or marginally in his favour he was going to offer a draw rather than throw 3 hours into a Caruana who was defending difficult positions extremely well, and rest / prepare to take the game in the rapids where he's recognised as super dominant ( https://2700chess.com/rapid ) .. I'm pretty sure if he wasn't favourite in the rapids he'd have pushed a lot harder, but it was clearly a very deliberate strategic decision being made rather than simply a case of "nerves getting too much". Certainly Kasparov would never have made that draw offer in a million years, and it's always easy to judge someone else's actions based on your own feelings.

I guess there's also the psychological element of "you know what? I know I'm better here, you know I'm better here, but it doesn't really matter as I'm going to win the tiebreaks anyway, so here, have a draw and I'll see you on Wednesday".

[gowan]

In chess or go, for me the question is who played the better chess or go. I consider anything faster than "classical" is a handicap to playing well. Fast time limits makes the game one of minimizing your own mistakes or blunders rather than actually understanding and finding the best moves. For me that is the point of playing, finding good moves. For me there is too much emphasis on winning over playing well. Really fast games are not fun to watch, either, because there are too many mistakes by both sides. Do you enjoy watching a fast game between weak players, where almost every move is a mistake? Even in a fast game between pros there is no time for the viewer to study the position and make judgements about the moves so what can you get from it but vapid entertainment? I enjoy watching real time matches on line in the two-day Japanese title games.

As for measuring the decline in playing level due to short time limits, at one US Go Congress I happened to play some blitz games with one of the visiting pros. He said I should take two more stones than I usually would in a non-blitz game; even so I only won one out of four games. At another US Go Congress, Jiang Zhujiu 9p (Jujo) played blitz handicap games with various strong amateurs and found that his opponents seldom made sente plays. So many people want to improve at playing go but if that's your attitude how much do you think you will improve by playing two or more stones weaker than your "actual" strength?

I consider that Carlsen and Caruana are effectively equal in strength at classical chess. In the historical world championship matches they went on for 20 plus games. Why do we have to have the championship match only six games long (I don't count rapid games)? If chess is being reduced to short time limits, it is more of a game of nerves rather than a thinking game.

[Elom]

I've mentioned this before, but I think we have to make an important distinction between blitz and classical.

In fact, I was just about to post on another one of my funny tournament ideas.

(I feel like if I think of something in my brain, it will appear in the universe...)

Anyway going off topic. Please pardon me if I'm haphazard or sound like a brat with a mobius hat. I'll use hide tags (sorry).

As I've said before, I've been confused by the attitude of comparing chess to go and trying to draw chess players into go by showing Go is better. I know but little of the history of chess but from what I've seen chess has done well in Europe** for an area that has little respect for mindsports in comparison to Asia. One could even say what has been accomplished in Asia is impressive considering global apathy towards mindsports.

However, it doesn't make sense to set one's standards by the level of accomplishment of another mindsport, yet alone compete in any sense. This is because all mindsports face common problems (I think) in public perception. Maybe you could say it is un-golike to focus on a local position when the global state calls for different local plays.

One may ha wise to view chess as a cycle in time before go. Rather than mimic chess^, it may be best to learn lessons.

Combine it with a study of physical sports (Croatian football team?)

And mental arts (Japanese archery?)

Shorter limits tend towards sport (in the importance of pre-preparedness and it's nature as a show).

And longer limits tend towards the mind (the 'cap') the emotion and the marathon. More of an event for great games than a tournament.

Now I'm going to put an opinion on Chess timings, but of course I'm not knowledgeable in any way on chess timings and only give an opinion based on first impressions. Draws in chess seem to often be handled via rapid or blitz games. I think this is unfortunate. It might be better to apply 'black wins all draws' in tiebreakers instead or something along those lines (which is essentially what go does for white with the half point) but I suspect the larger difference in advantage the second player gets after having a win-draws-guarantee is more in chess to the point of being impractical.

I would ideally start on faster limits, and end on longer ones (the stronger players will do better earlier on but when they meet the longer time settings even things out. Maybe).

Faster Time limits are a lot (I think) about:

1-Personalities and their growth, like a story. Faster time limits lean heavier on the knowledge-perception base, the crystalised 'chessness' (or 'goness' if you will. Blegrh...) of the player. I've heard that the advantage to the stronger player increases the shorter the limits, because the weaker player has less time to read out unfamilar situations. the would be more unfamiliar situations in a game for the weaker player than the stronger player that has more shapes positions memorised and unforgotten in their brain, a natural neural net (combination).

2-Action. I heard that players take more risks in shorter limits because they cannot so easily read out risky sequences that fail.

Fighting spirit— Go for it!

Long time Limits:
Creating high-quality games that go apprentices can study, that future students can look at like today's do at the great games of the past and go

wow

But it's hard to do that when there are no international tournaments with 4-8 hour limits.

Now I know that there comes a point where increasing time increases quality less and less and less. It's a logarithmic scale going both ways, perhaps a sweet spot's somewhere there.

But I want every final to be like the Gu Lee Jubango!


Rapid games I think are a nice mix of all aspects.

We need shorter limits to draw in people who would otherwise pass an eye over mindsports and longer limits for the limitlessness of go. One speed of play shan't push out the other! is along the lines of where I'm thinking...

Regarding prize money... well I shouldn't really be commenting on this but I think doubling the prize money every round is best for pro tournaments. In general.

^Chess institutions

*even a 9x9 board has more possible combinations than atoma in the universe, I'm sure. For chess, it's best to use the estimated entire universe which is 150,000,000,000,000 times larger or so. And for go on 19x19 if every atom in the observable universe had a corresponding universe (i.e. the observable universe squared) one may get closer to the 19x19 games. I need to check this but I don't have time right now. 9x9 is like a duel, but 19x19 is like a war.

So Chess should be compared with 13x13 go. In fact I think partnerships with chess clubs should be based on the 13x13 board (similar board sizes, etc). Okay I'm barely speaking common sense here...

**the other law of the internet .

[John Fairbairn]

The way I'm reading the news around the chess world championship, many people there are in fact clamouring for shorter time limits after the main event saw 12 draws in 12 games.

There may be a clamour but what you don't hear is the silence of those who've lost interest because of the new format. Time limits may or may not be a significant factor. Gauging chess popularity is complicated anyway by the fact that draws are a turn-off, and there's also the problem of digital cheating (chess seems to be well ahead of go on this curve). But I do think (as some comments already on this thread suggest) that there's a significant number of fans who expect and prefer a mind game where the mind is given time to operate properly.

There's also the effect this has on sponsors. I suspect their eyes gleam at the possibility of getting tv coverage and they don't care if tv insists on inviting the wee sleekit tim'rous beastie to set the time limits. But what does seem to cause a panic in the breastie - and for the players and not the mouse in this case - is when tv drops their coverage as audiences are always clamouring for something new. We have seen several major Korean go events ruined by this trend. I obviously don't know the details of why these events folded, but the pattern of very short time limits seems to be common to all or almost all.

Yes, this made the world news for chess (Armageddon was obviously the trigger word yet we didn't get that far) but all that glitters isn't gold, today's news is tomorrow's fish-and-chips paper, and it's a long lane that has no loaf on the bread.