Insufficient material to mate.

In my opinion the one rule that is the most annoying personally is the rule that P+K vs N+K or P+K vs B+K is a loss if the side with a pawn times out (unless the pawn is an a or h pawn), for two reasons:
1. If the pawn is neither an a or h pawn, the king cannot be mated unless the pawn underpromotes, which in this endgame is incredibly stupid unless there is some situation I am not aware of that leads to stalemate if the pawn promotes to a queen.
2. In lichess, most players either have auto-promote to queen turned on or promote to queen if premoved turned on. Most cases where this rule applies is in time scrambles during bullet, in which case the player with the pawn will likely premove the promotion.

I understand the rule that by FIDE rules P+K vs B+K and P+K vs N+K flags are losses because it is possible to lose as the side with the pawn, but unlike OTB chess matches which normally have increment lichess games often don't, so this rule applies much more often on lichess than OTB.

Consistency is paramount. Keep it simple, stupid.

@Toadofsky That's an optional rule for USCF, similar to FIDE's rapidplay finishes. Their adjudication guidelines are what basically every site aside from Lichess uses:

"Defining a win. A game is won by the player who:

...

7c.) Who correctly points out that the opponent’s flag has fallen first, at any time before the game is otherwise ended, provided the player has mating material. Mating Material consists of (at a minimum) two minor pieces, a pawn, a rook, or a queen provided it isn’t a position where one could claim a draw under rule 8. If a player who claims a time forfeit states the claim with claimant’s flag still up, but then fails to stop the clock in time to avoid also exceeding the time limit, the claim will be void, unless the flag fall was observed by a director or independent witness."

The new USCF rules are an overkill. But taking autoqueen into account in adjudications (blitz+bullet) might make sense.

The new USCF rules also still contain the "7c) mating material" flaw whereby a player having forced mate with only a knight or bishop cannot claim a time win, for example:

White: Kc1, Nb4
Black: Ka1, Pa3 (time expires and white does not have mating material, despite having a forced mate)

@Toadofsky quite trolly there, eh? Heh. Of the millions (billions?) of games played on Lichess, let's consider all that have ended in a win for a player with exactly one minor piece. How often was this because of mate? How often was it because of Lichess' rule?

Even if we include the fanatical time controls and players of all ratings, I'd imagine you're going to see a ratio on the order of thousands to 1 of these games only ended up won because of Lichess' rule.

This is precisely why the USCF guidelines are the right way to go. Indeed you might get the wrong decision 1 in a million times. But the FIDE guideline is going to give the wrong decision the other 999,999 out of a million times. Should their online games ever start to gain any decent level of popularity (without increment), I'd expect to see a rapid change. It was a well intended guideline, but it works out pretty poorly for the short incrementless online games that make up most of online chess nowadays.

You're so right, USCF. A pawn is a pawn.



Your are entitled to your splendid probability arguments.

#56 I contacted my local USCF TDs regarding trying to fix this loophole (before reading your comment) since their rules have some merit. Specifically:

"Mating material" is more or less what you're asking, although again I stress that a player capable of checkmate (forced or otherwise) may be deprived of a win on time (which is what I'm petitioning USCF to consider somehow).

"Insufficient losing chances" seems a necessary counterpart to "Mating material": that a player may claim a draw in an obviously drawn (or won) position such as in #57. Since we don't have ILC rules, neither should we have "mating material" rules.
http://buttersafe.com/2013/03/28/chess/

(If Lichess ever gets ILC rules, then I'd favor adding "mating material" rules, but no sooner.)

@Toadofsky consider the 50 move rule. There are an immense number of endings that are technical wins that have neither a capture or pawn move for well beyond 50 moves. But the 50 move rule still works extremely well in practice. The reason is that in practice if you look at a million games that have gone beyond 50 moves without a pawn move or capture, they're nearly (if not literally) all just dead draws. So the technical existence of a possible contradiction to this guideline doesn't mean anything in practice.

And single piece mates are similar. Look at a million games where one side has a single piece and it's the same story. You'll see nearly (if not literally) no wins. The reason is that single piece mates are helpmates, but not like a one or two move 'oh shit' type helpmate - but complex, orchestrated, study like mates where one player generally has to extremely actively work to help their opponent mate him over a lengthy series of moves.

So this is why I don't really see insufficient mating material as any way related to insufficient losing chances. It seems much more closely related to the 50 move rule. Although I do see your point there - and at least applaud you for that! But in general I think insufficient losing chances is something that is for positions where the outcome is not, in general, self evident. For instance R + FGH pawn vs R + FGH pawn is one where a player might have claimed insufficient losing chances. It's extremely complex to determine and is going to rely on the skill and judgement of the adjudicator.

What about N versus a- or h-pawn, King in front of it? This occurs pretty often. I won it on playchess some time ago.

If you start to define single cases you will open a can of worms.