In top level chess when both sides are happy with a draw

How does it generally work, if both sides before the game know that they only need a draw, but a loss would be bad? I assume that it’s unseemly to just shake hands and agree a draw over a pristine board, are there variants that are de facto draws, so if both sides follow the path you know that you’re quickly headed for a drawn position?

You’re always happier with a win than with a draw, and you’re always happier with a draw than with a loss. White has a significant advantage over black, and competition is always based on multiple games, with white and black alternating, so the conventional wisdom is that you try to draw as black, and try to win as white. While agreed-upon draws are a common way for games to end, they only happen when both sides agree that they can’t realistically expect to win. White would have no reason to agree to a draw, over a pristine board.

In World Championship Title matches of the past the format was x number of points, draws not counting. This could lead to a marathon event lasting months. In that marathon it was common for the players to have a tacit agreement to draw games quickly, just to give them a rest.

I’m not really talking about title matches. I know that chess, like football, occasionally uses aggregated score over a group of players, so there will be occasional times where both players are 0.5 points from a cut-off point; and 0.5 or 1 mean the same to them, but 0 would be worse. Not that it would happen very often but it would happen.

Chess requires a lot of preparation and a lot of energy. There are times during a tournament when it is beneficial for both players to obtain a draw. For example, a top player has already secured the tournament championship and only needs a draw to finish up. He may be playing a weaker player who would be happy to get a draw against a superior opponent. The top player can play an opening and variation that is very “drawish” in nature, and the weaker player happily follows along. After a dozen or so moves, a draw offer is made and accepted.

There have been draws agreed after very low (even 0) numbers of moves, but as you say, it’s not considered the ‘done thing’. Which is odd in a way, since the usual way round it, as Jasmine states, is to just play for a draw from the outset. Another way players sometimes achieve this is to copy the moves of whatever famous drawn game from chess history that they fancy.

It’s certainly not uncommon for this situation to arise in regular chess tournaments, which are often played over a set number of rounds. Say two players starts their last round game against each other both exactly one point ahead of their nearest challengers in the rankings. Both know that a draw (half a point) guarantees them their position, with a corresponding amount of prize money. Whereas a loss would risk being caught be the chasing player(s) and having to share the prize money. In such a case they are unlikely to want to risk losing by playing to win, even with White.

How does the chess enthusiast community regard this kind of situation? Is it business as usual, bad form, magnanimous?

I don’t play in chess tournaments but I think it’s seen as accepted practice by nearly all. Some parts of the chess community would like to ban agreed draws (at any stage - not just those effectively agreed before the game starts) on the basis that it’s bad for the image of chess, but I think that’s part of an effort to make chess more of a paying spectator sports. For sports where this already applies, e.g. soccer, this is indeed a known (and largely solved) problem that tournament organisers take steps to avoid, by arranging the format such that a pre-arranged draw is rarely if ever a desirable outcome. But in chess tournaments, the vast majority of which have no spectators and virtually none of those that do having paying spectators, it’s not such an issue. After all, if you’ve outclassed the rest of the field such that a draw in your last game will secure the championship, that is your rightful reward. It happens in plenty of other sports and leagues, too (or indeed where a loss in the last game or few is irrelevant).

Pre-arranged draws are frowned upon. It is considered bad form to meta-game with other players in a tournament about what outcome you want to produce. However, there are a number of lines that become very drawing very quickly (and can soon become forced draws), and if both players choose to play into such lines, a draw can be effectively arranged in real time. The difference here is between:

(1) The game is on, and it’s just between those two players in that moment, and they each independently decide that a draw is optimal for them.

or…

(2) The players team up ahead of time to plan on a mutually advantageous result, perhaps even conspiring for or against other players’ standings in the event.

The overarching philosophy is that it is meant to be every player for herself/himself, not informal teams. Bobby Fischer famously maligned the Soviet federation for, in his eyes, colluding against him through match fixing (including draws).

Many high-level tournaments disallow draw offers before a certain number of moves (say, 30) to prevent “1. e4 (=)”, although draws by repetition can occur at lower move counts still.

I will defer to your obviously greater knowledge and experience of tournament chess etiquette. But it seems to me the practical difference between 1 and 2 can be virtually indistinguishable. For example, in the situation I described, where both players stand to benefit significantly from a draw but may suffer if they lose, whether they discuss an agreed draw beforehand or come up with one over the board (whether that be in 10 or 31 moves) is pretty irrelevant, it seems to me. What really is the point of both players choosing a drawish line and then indeed agreeing to a draw on move 30, other than keeping up appearances (and even that fails anything more than a cursory inspection)?

Having said that, I read about a case from the World Junior Chess Championship (or some such) a few years ago at which in the last round, a Dutch player (who was half a point ahead of his nearest rival, from China), came to a draw against his opponent after 40 or so moves, at which point his Chinese rival’s opponent (who was also Chinese) resigned in an equal position, giving th Chinese player the title on countback. That seems wrong, but again, what can be done to combat it effectively (other than the obvious solution of trying to avoid pairing friends in the last round, but that may not always be possible).

I suppose you could keep the players ignorant of the overall standings, so they don’t know how far ahead or behind they are compared to the competition, and so have an incentive to always play their best.

In the long, rigorous, 82 game NBA season, it is understood and accepted that, on some nights, teams are just going to “mail it in” and lose. It’s understood that, in the long term strategy of playoff seeding and the quest for a title, it is just part of
how things go. I would say that these kinds of draws in chess would equate to that.

There isn’t a whole lot you can do directly to prevent players from colluding, or at least not always giving full effort in every game. It’s not much different from other sports, where at the end of a season there are often teams without anything to play for, or where a draw is as good as a win, in terms of their position in the standings.

One thing with chess in particular is that there are often multiple different goals or incentives for players in the same tournament. There are often category prizes, so you get situations where the tournament leader is matched against someone with a chance for one of the other prizes, such as top junior player. If a draw means they each win their prize, then of course they’ll want to go for it. Also, players want to increase their rating, and to get “norms” for titles, and they can be willing to sacrifice their tournament chances to reach these other goals.

One approach to avoiding much of this is to structure tournaments as knockouts, instead of using a cumulative system like round-robin or swiss. In a KO, you’re playing a match against your single opponent and you must win or you’re out. But many players won’t go to a tournament where they can be knocked out early; they want to be guaranteed a number of games.

This wouldn’t work because many draws result because of the simple reason that outstanding players often and honestly play each other to a standstill. Neither player can actually win the game unless one of them just says, “the hell with it”, and makes an intentional blunder.

“Grandmaster draws“ have been around about as long as organized chess. Here’s a good article from more than a decade ago trying to talk to possible solutions.

You are right, they are virtually indistinguishable. It’s a spirit-of-the-game thing. Playing for a draw per se is not considered bad, especially when the tournament structure or other incentives make that the correct play for a player. It’s just that deciding on such a strategy in concert with other players ahead of time is considered poor, as the scope of collusion is much broader there. I admit, it’s a fine and often immeasurable line.

Yeah, that’s a perfect example. The resigning player presumably got (and would deserve to get, IMHO) some pretty severe stink-eye from the community. There isn’t really an avenue for formally stopping such behavior, which is even more why it’s overtly frowned upon, to at least provide some back-pressure against it.

Knockouts are tricky in chess but possible. For instance, you don’t want each pairing to be decided by one game, draws notwithstanding, because the black player would be at a disadvantage. A typical scheme is to play a mini-match with an even number of games between the players (perhaps just two games), alternating the colors and assigning the usual points of win=1, draw=0.5, loss=0. If the mini-match is tied at the end, you can continue with pairs of tie-break games or, especially in shorter time controls, play an “Armageddon” round, where a draw counts as a black win, but black starts with less time on the clock. (Black is said to have “draw odds” while white has “time odds”. Common is 5 minutes vs. 4 minutes in blitz chess.)

Because of the need for mini-matches in each round, you can’t really do a knockout unless it’s a fast time control.

KO tournaments are fairly common - each round is played as a match rather than a single game, as Pasta described. The World Cup KO is one of the biggest tournaments, and in the past knockouts have often been used to select the challenger for the world championship match. Many recent online tournaments have been KOs, or at least included KO portions. Playing online avoids most of the issue of attending a tournament only to be knocked out in the first round.