Chess.com sneaks a cheap shot

This isn't a whim of Chessbase/Playchess, that's the official FIDE rules. Playchess is an online service provided by the German company Chessbase, thus it is rather natural that they implement the official FIDE rule set rather than some USCF spinoff.

Here a full FIDE vs USCF comparison of how the mating material assessment rules (that also determine the game result when one side exceeds time) effectively work:

FIDE rule set:

A side is considered to have mating material in a given position if the opponent could be checkmated as a result of any series of legal moves (this includes helpmates).

The presence or lack of mating material for particular material configurations of the side under consideration is assessed as follows:

K - no mating material present (no check, let alone checkmate, is possible) in any combination of opponent material;

K + N - mating material present (due to possible helpmates), unless the opponent has a bare king, or a king plus any number of queens (and no other pieces or pawns);

K + B - mating material present (due to possible helpmates), unless the opponent has a bare king, or the opponent's material consists of any combination of a king and any number of queens, rooks, or bishops bound to same square color as the own bishop (and no other pieces and pawns);

K + NN - mating material present (due to possible helpmates) in any combination of opponent material (no opponent chessman is necessary to block an escape square of his king).

In any material combinations not covered above mating material is considered to be present.

USCF rule set:

A side is considered to have mating material in a given position only if that side's material (including pieces obtained from potential pawn promotions) is sufficient to force checkmate on a bare king.

A side whose material is reduced to any the following combinations (that don't allow delivering forced checkmate to a bare king) is, irrespectively of opponent material, considered to not have mating material under the USCF ruleset:

K

K + N

K + B (specifically, king and any number of bishops running on the same square color);

K + NN.

The only exception from the above is if the side under consideration can deliver forced checkmate from the given position.

In any material combinations not covered above mating material is considered to be present.

The following rules apply under both rule sets:

• as soon as both sides lack mating material (by the considerations of the given rule set), the game ends in a draw by insufficient material;
• if a side exceeds time while the opponent lacks mating material (by the considerations of the given rule set), the game results in a draw; if a side exceeds time while mating material is present on the opponent's side, the opponent is awarded a win.

Playchess, a European service, implements the FIDE rule set.

On the other hand, Chess.com implements the USCF rule set sans the exception, as no check for a possible forced mate (which would require a position analysis) in the given position is performed.

What I posted is not a direct quote from any of the rule sets, but my own summary of how their respective mating material assessment rules effectively work. As such, I don't worry about lawsuits. ;)

(If "rule set copyrighting" was to be taken to the extreme, two persons would only be allowed to discuss or even mention the rules if they both bought the rule book, and that only in a private conversation without the presentce of any third party - what sense would that make?)

The USCF does not make their complete rule set available online, but a summary of differences to the FIDE rule set is available (from there you can learn that the USCF sanctions the use of upside-down rooks as queens in official play, as well as moving the rook first when castling:

http://www.uschess.org/content/view/11750/668/

From there you can learn things like the USCF sanctioning the use of upside-down rooks as queens in official play, as well as moving the rook first when castling (doesn't make sense really given that castling is a king move, but oh well).

I agree that it would be better if a common rule set was accepted wordwide. Most national chess associations do adopt the FIDE rule set, but somehow the USCF must have their own ways. Go figure.

Hello ALL,

   I can't believe an outraged but ill-informed forum query instigator can generate four pages of comments, perhaps more by the time I'm through, even though the angry player had only a bare King.  It seems that 95% of the respondents on the first three pages are only talking through their hats and haven't a clue about what they're commenting on because USCF BLITZ CHESS Rules apply . . . according to which . . . .

A game is drawn:

When either one of the kings is stalemated.

or after agreement between the players.

or If the flag of one player falls after the flag of the other player has already fallen and a win has not already been claimed, unless either side mates before noticing that both flags are down.

OR whenever one player has insufficient mating material when the opponent’s flag falls.  Could it be more straightforward?

Bob

ps

Other niceties like, three-fold repetition draws and  50-move rule draws usually are out of place in blitz, because of the near impossibility of the player himself (rather than a website) keeping an accurate record.

pps

There is some controversy over one aspect of the rule:  a player can get a draw after his opponent's flag falls  "or makes an illegal move."  It is NOT clear who makes the illegal move in question . . . a very poorly written clause, eh?

Huh? Are you sure about that? that's kind of a weak rule then:

So according to chess.com rules black can make a draw here by waiting untill his time runs out?

That's the most important rule given the OP's query, however the problem is that whether a side has sufficient mating material in a given position may depend on what rule set (and implementation of a chess website) is in use, and determining that can actually get pretty complicated (as you can see from my lengthy post above, that focuses precisely on an in-depth analysis of that).

It is rather useful to know these peculiarities in order to avoid surprises in situations such as this:

"It is possible [under FIDE rules] to lose on time in situations that are a draw under USCF rules. For instance, GM Nakamura lost on time with a king and rook vs. king and knight. Under the FIDE laws of chess, the game is drawn when one player runs out of time only if there is no legal sequence of moves by which the opponent could checkmate the player. Since there is a helpmate that allows a king and one knight to checkmate a player with a king and rook, GM Nakamura lost."

It would appear so. As per the USCF rules (which Chess.com is oriented on) white has no mating material, and as far as I know no analysis is done to check whether a forced mate exists in the final position. This might be worth checking or consulting with a staff member who is competent to advise on that though.

Huh, the position is mate, it's black to move. he has to play h2 and then Nf2 is mate.

besides, it might not have been a draw before either (that black king could have taken a rook on h1 for instance)

I'm all for having a draw in clear drawn stuff, it's just that the rules posted before means that some clear wins are now draws, that's why I gave the example.

Here, I'll give another one to clarify:

@FirebrandX:

I'm not sure what you are tempted to disagree with in particular. All that I said about the determination of a presence/lack of mating material applies, based on FIDE Laws of Chess rule 6.9:

"Except where one of the Articles: 5.1.a, 5.1.b, 5.2.a, 5.2.b, 5.2.c applies, if a player does not complete the prescribed number of moves in the allotted time, the game is lost by the player. However, the game is drawn, if the position is such that the opponent cannot checkmate the player’s king by any possible series of legal moves."

 This is confirmed by a quote from the USCF - FIDE Rule Differences, "Obscure differences", #8 (I am posting it for the third time):

"It is possible [under FIDE rules] to lose on time in situations that are a draw under USCF rules. For instance, GM Nakamura lost on time with a king and rook vs. king and knight. Under the FIDE laws of chess, the game is drawn when one player runs out of time only if there is no legal sequence of moves by which the opponent could checkmate the player. Since there is a helpmate that allows a king and one knight to checkmate a player with a king and rook, GM Nakamura lost."

What you mention is covered by FIDE Laws of Chess Article 10, and it is much more specific - pertains only to a sudden death (no per-move increment) time controls, and the player wishing to claim a draw in that way must have less than two minutes on the clock:

Article 10: Quickplay Finish

Again, rule 10.2 specifies certain strict conditions under which a draw might be claimed, whereas rule 6.9 is general and independent from the above.

Nobody had hallucinations when Nakamura lost on time against a K + N when playing under FIDE rules.

If you find it "unfair", that's an opinion. Regardless, that's how the FIDE rules work. Please don't imply I have provided incorrect information just because you don't like these rules.

If the conditions for rule 10.2 to apply were met, he could have gotten a draw. But that's a specific rule, and if it does not apply or no claim is made, then 6.9 applies and the player who exceeds time loses if the opponent has mating material according to FIDE rules (i.e. considering helpmates, which I presented a thorough analysis of).

Ah ok, that's a good point. I didn't really look at the problems of the alternatives, I just thought it was odd that some clear wins were now draws.

Normal human play in time pressure is full of blunders. There are plenty of helpmates in real games.

I am not defending anything. I have made an effort to present in detail how the presence/lack of mating material is determined under FIDE and USCF rules. This was my focus. Please don't complain to me or say that the info I gave is incorrect, because it isn't. There are minor additional rulings like the USCF "insufficient losing chances", or FIDE Article 10, but they pertain to specific situations where an arbiter would be needed to adjudicate the game, and as such these rules cannot really be implemented in online play anyway.

If I remember correctly, Chess.com used to implement the FIDE way of determining mating material as well, until people started whining that they "should not have lost on time to a K + minor piece", or something similar. It was then eventually changed to adapt the USCF rules, but not flawlessly, because this would require a position analysis to catch any forced mates, which is not implemented (vide the corner case presented by HattrickStinkyduiker).

The arbiter will overrule Nakamura's super-GM opponent by claiming the guy isn't trying to win normally? How can you tell for sure if someone's trying to win 'normally' anyway?

'Just shuffling the Knight' could be a brilliant redeployment of the N to create zugzwang, for all that arbiter knows.

Please hold your horses, sir. As I explained several times already, I objectively analyzed the mating material conditions (which happen to also determine the result when a flag falls) for both rule sets and presented the results, in order to at least clear up on what and how is determined where.

It so happens that you seem not to be fond of the FIDE rules (which are accepted almost everywhere except in the US). In that case, be happy that Chess.com employs a (flawed) version of the USCF rules instead, and complain to Chessbase/Playchess or even better to FIDE directly if you dislike their rules so much. I think that they were very intentional to include possible helpmates (which may also be a result of blunders, as BigDoggProblem correctly pointed out), as that approach does not make any assumptions as to how the game would have continued, it only takes into account how it could have (and is admittedly easier to implement on chess servers, as it does not require position analysis to rule out possible "forced helpmates" as in HattrickStinkyduiker's example ).

As for taking the liberty to call the USCF rule set a spin-off - well, FIDE is the World Chess Association, and as such their rules are authoritative almost everywhere, except for the USCF (a national chess organization) which chose to go renegade and have their own rules. Admittedly, if we see the USCF allowing the use of inverted rooks as queens in official games, or allowing moving the rook first when castling, it's easy to have a little less faith in them than in FIDE.

@HattrickStinkyduiker:

The position you gave in post #129 is actually quite interesting - it's very study-like. As it turns out, Ng4 is the only winning move - the hanging white rook is not only poisoned, but pulling it out of en prise will allow Black to get away with a draw.

As for the original position:

If Black exceeds time, then:

• by FIDE rules, Black loses (White has mating material judging by material alone, as the black pawn can assist in a helpmate against Black);
• by USCF rules, Black loses (White normally wouldn't be considered to have mating material with his material reduced to K+N, but in this particular position he has a forced win, i.e. 1. ... h2 2. Nf2#);
• by Chess.com "house rules", it's a draw (the server considers White not to have mating material having only K+N, the position is not taken into consideration and as such the forced win for White is ignored).

I would very much like to see how rules can be "copyrighted". Their cording is considered a work of art ?