I think that’s good advice - if you can learn something, play on.
Otherwise it’s pretty depressing trying to think of a good move each time in a totally lost position.
Yes, that ending can take over 30 moves (by each side) to win.
However it is relatively easy to drive the lone King to the edge of the board.
(N.B. The best defence is for the lone King to head for a corner square of the opposite colour to the one the bishop moves on, since you can’t be mated there.)
So if you are keen enough, learn the procedure to drive the lone King from one corner to another of the opposite colour - and mate soon follows.
I had that position against a player rated about 2150 once. I was pleased to finally use my knowledge to force mate quickly.
Afterwards my opponent confessed that he thought it was a drawn position - and wondered why I was playing on!
That’s a very interesting use of game theory in the meta-game of the whole tournament.
How do such considerations sit relative to the ethos of the game? Golf for example famously has a culture that any cheating is only cheating yourself. I’m not enough of a golfer to know whether that’s more honored or honored in the breech.
Is one “duty-bound” to play each game to the max, considering only the 64 squares in front of you, or are larger meta-strategies considered just one more aspect of the high level game to be exploited as and when possible?
I could imagine in some multiple simultaneous game scenario dumping a game or two to buy the mental time & space to win a tough one. Or alternatively resigning from the tough one to gain the brain bytes to reliably sweep the field of all the lesser opponents.
A fascinating area overall, and one I have zero clue about.
I can’t speak much to chess specifically, but I think the prevailing view amongst most sportspeople is that you play to maximise your chances of winning the overall event in question. So another example from chess would be, if you have a 1 v 1 World Championship match (played as a series of games), and one player wins an early game, I don’t think they are under any obligation to paly aggressively from then on, given that they will win the title if all the games are draws.
There are similar cases from more high-profile events, for example in the football (soccer) World Cup there have been instances where it suited both teams to draw, and they duly did so. Then there was the Badminton competition in the Olympics 5 (I think) years ago, when both sides realised the seeding of the subsequent rounds made it in their interests to lose. That last case is slightly different in that it actually involved playing to lose, rather than draw. The most famous one, which I have mentioned at least once on the boards before, was a Caribbean soccer competition in which the rules stated drawn games would be resolved via a penalty shoot-out, and a win via penalty shoot-out would be scored 2-0 to the winning team. In the closing minutes of a late-stage game, a team winning by 1 goal, but needing to win by 2 to progress, deliberately scored in their own net to level the game, hoping to win the penalty shoot-out and thus secure the 2-goal winning margin needed. The final few minutes then involved the other team trying to score in either goal!
My view is that in all the cases, the onus is on the competition organisers to design the competition in such a way as to obviate such behaviour, which is usually possible with a little foresight and effort. For example, glee’s chess tournament could have had a rule that if 2 players were tied on points, first place goes to the lower-rated player - in which case, the GM would of course have played to avoid a draw. Or the rule could have been that ties on points would be resolved with a blitz game, in which case (I assume) glee would have tried to win the rapid game rather than offering a draw (unless the GM was known to be significantly weaker at blitz than rapid, which would be rather unusual).
For events with few spectators and that don’t rely on their ticket money to finance them (such as glee’s chess example), it doesn’t really matter. But if the organisers of the World Chess Championship want it to produce exciting games that people will watch and talk about, they need to think about how to incentivise that kind of chess over the typical grandmaster draws that we see a lot of. And to some extent they have done so, by means of a relatively (compared to past contests) short series of classical chess, followed by rapid tiebreaks.
Yeah. I’m fully familiar with the game within the game, perverse incentives, tie breaker shenanigans, and all the rest. Though I always enjoy hearing you retell the story of the two soccer/football teams trying to out-own-goal each other into a loss. Like the 3 Stooges at their finest.
I was specifically curious how high level chess treats this area. I think of chess as simultaneously ruthlessly aggressive and yet somehow genteel, above the sweaty testosterone-fueled norms of lesser games. Which of the two opposing impulses carries the day is the question.
The fact @glee made the offer of a draw implies chess is OK with the idea the meta game is what really matters. The fact his opponent wasn’t shocked, shocked I say! at getting the offer says the same.
That the GM essentially then filibustered until he had no choice but to accept the draw offer strikes ignorant me as a face-saving or covering-up move. Surely he understood the full implications of the proposal within seconds (if that many) of glee making it. Which filibuster, if I read it right, leans the opposite way: accepting the draw would be unsportsmanlike or beneath the dignity of his (legitimately) high station.
I’ve gone through many grandmaster games, and I can’t remember seeing one played through to checkmate. One reason is that they are on a level where they can see their own impending doom well before the rest of us mere mortals. Another is that it is uncouth to force a talented opponent to play out the rest of the game that a club player could win. It’s like being down 20 points in a basketball game with less than a minute left and calling timeouts. Another reason is that it is embarrassing to allow yourself to be mated. It’s purely an amateur move and not worthy of that level of play.
This genuinely surprises me. Back in high school chess club, we’d practice the K+2B and K+B+N v. K mates, and yeah, the latter is a lot more work. (Not that I ever had to use either of these in a real game. But if it had happened back then, I’d have been able to pull it off.) Hard for me to believe someone with a >2000 rating didn’t know you could force mate with either of these (if you know what you’re doing).
But of course you can’t force mate with K+2N v. K.
In terms of game ethics, I don’t consider we did anything wrong:
we had achieved our leading positions with full-length wins in the previous 5 rounds
although my opponent was rated higher than me, I was known as a strong speed chess player*
if one of us lost, their amount of prize-money would drop considerably**
most importantly, we hadn’t agreed anything before the game
Afterwards the tournament referee and the players who came 3rd = were mildly amused by my action and congratulated us on winning.
*I had won the British Speed Chess Championship a couple of years previously
** my opponent was a professional player, who needed the money far more than me (I was a computer programmer)
I’ve got a busy couple hours of work coming up, but I’m gonna have to look at this afterwards. Having not thought about this problem for a half-century, my WAG is that this enables the player with 2 knights to avoid stalemating the other player.
Well in an international tournament with many entrants, the same situation as in my game arises.
Players in the leading positions will win a lot with a win; a decent amount with a draw; and very little if they lose.
If your main source of income is playing in tournaments, then I can understand why there are short draws between closely-matched professionals who are in the lead heading into the last round.
Similarly in qualifying events for World Championships, professional players are probably far more interested in finishing in the qualifying places than actually winning the event.
Once you get to the final two playing for the World Championship title, then the problem (as has been mentioned earlier) can be that a player who takes the lead has no incentive to win any more (if that gives them the title.)
On a lighter note here are two anecdotes from my career as effectively an ‘amateur’ player (i.e. I would happily take any prize-money, but my income came from a full-time job outside chess.)
In the last round of an international event, my opponent and I were just above the middle. A respectable result, but no chance of any prize-money.
After 6 moves, my adversary offered a draw. I politely declined. (All around us, other games were finishing as swift draws and players were heading off to the bar!)
My opponent was astonished and I found out later that he asked a club-mate of mine what was going on. “Oh, glee just likes playing chess!” came the answer,
Well we were one of the last games to finish and the result was … a draw! (But it was a good game.)
In one of my best-ever results, I was in clear second position with one game to go and a draw would qualify me for the all-expenses paid Master section next year.
I was playing … a club-mate of mine, who was leading the whole thing (and similarly heading for the Master section.)
Before the game, other players in the event were congratulating me on qualifying and my mate on winning the tournament. “How long before the draw?!” they wanted to know.
They were astonished to hear that we not only hadn’t agreed anything beforehand (which shows what many players think is ‘ethical’), but that we were going to play a full game.
And it was a pretty tough game … which ended in a draw, but again was one of the last to finish.
I am confident that the GM (who I knew well) was just surprised by my offer.
There were no spectators or reporters and the games were not recorded or published - so there was no need to ‘put on a show’.
I think he literally lost track of time (he may have been thinking about his bank balance, because his rating was around 2500 and in those days you didn’t earn much money at that level.)
The win is pretty difficult … and impossible if the pawn has advanced too far.
You have to trap the lone King in a corner with your King and one Knight, whilst your other Knight blockades the pawn.
Then you release the pawn and the race is on between your Knight delivering mate and the pawn Queening.
The ending was analysed superbly (pre-computer days!) by the study specialist Alexey Troitsky [’ Alexey Troitsky - Wikipedia’], who showed just how far the pawn could be advanced (it depends on the file the pawn is on.) [’ Two knights endgame - Wikipedia’]
On the one hand…
I had an OTB game with an FM where he “blundered” a piece but it was a trap and I lost. The other Master at the club told me his position was actually completely lost and the cheap trick was his only chance. Hurts when you realize that, but that’s how it goes.
On the other hand…
My bar for resigning is much higher against lower-rated players. Players who are 100-200 ELO below me very often get into a better or winning position. It may even be as much as half of all such games, as many players bring trick openings that put me on the back foot right from the get go.
However, I still end up winning 2/3 or more of those “losing” games, because they’ll often blunder or just generally not know the end game theory.
This might stretch your definition, but at the 1967 Interzonal (the tournament to select world championship challengers), Bobby Fischer, who had already forfeited a game by not showing up, showed up with five minutes left on his clock before forfeit (so 55 minutes into his first move), and made a move. He proceeded to squash his opponent (Samuel Reshevsky, an American GM who already had many reasons to dislike Fischer), who dragged out a lost position and then filed an unsuccessful protest.
At the elite level in classical (long) time controls, playing to checkmate is rare but does happen. Here is a recent example (last few minutes) where Svidler allows Carlsen to mate purely because of the aesthetics of the sequence. (Svidler is a fairly laid back and dead-pan-humor sort of person, though, so it fits his persona.)
More often, though, resignation at elite levels happens rather far away from the actual checkmating sequence, once the position is clearly lost and counterplay will be ineffective.
In fast time controls, there is always a chance for mistakes, so games are taken to mate much more regularly.
At lower levels, however, players resign way too often. Mistakes happen constantly, and a common coaching tip to novice and club-level players is “never resign”. At worst, you get practice visualizing (e.g., calculating for stalemate tricks in rook endgames). Better, your opponent opens up the possibility for counterplay that he underestimates and you are able to make the position messy and mistake-prone. These aspects of the game are genuine aspects, and resigning before exercising them is leaving points on the cutting room floor.
I resign when I am sure my opponent can finish me off without even thinking and/or I have nothing to learn from the process of continuing and/or I am tired and see no practical path to victory and/or I am playing in a multiple-rounds-per-day event and have better tournament EV by readying for the next round than by spending another hour hoping for a 3% chance of a weaseling out a draw.
To be sure: if I have any chance for counterplay in a technically lost position, I will not resign until that counterplay has been squashed.
It certainly isn’t cheating. But it is reasonable to ask what sort of practice different puzzle formats provide. Personally I feel that knowing ahead of time that a specific theme is key takes away much of the benefit. Knowing that it is a mating sequence isn’t as bad, since forced mating sequences can very well require significant calculation – which is good practice. For tuning your “tactics radar”, a broad pool of puzzles is clearly better, since you will have to evaluate the tension in the position and look for tactical themes present. A final level is fully mixed positions, where you need to decide if you are looking for an advantage, a critical defense, or a quiet positional improvement. Resources exist to study in this way. Here is one well-regarded book for students of the game. The site chesstempo.com also has mixed-mode tactics (though maybe not with positional answers, just “to win” or “not to lose”).
[quote]
2. Is the rating system on lichess(dot)org remotely comparable to what a player’s rating would be in real-life, rated tournaments?
[/quote[
Not even close. The Lichess rating system is notoriously higher than common OTB systems. This isn’t bad or good, of course. They’re just different systems. If you are a beginning player, your Lichess classical rating might be 1600 and your USCF classical rating might be 1100. If you’re a high-level club player, the difference is more like 200-ish points instead of 500.
If that’s Lichess classical 1300, then, yeah, maybe 800 or 900 UCSF.
Do it! It’s very fun and you’ll improve rapidly, much more so than by playing online with typical online time controls. You may already know this, but tournaments of medium to large size are typically split into rating sections, so you would play essentially a separate tournament with other newer players. Big tournaments might have six or more sections. Smaller ones, maybe two or three. Local chess club tourneys might be limited to one section for all players, but the Swiss tournament system means that after a couple of rounds you’ll be paired with similarly rated players anyway.
- and wondered why I was playing on!
