lichess.org
Donate

A few words about Lichess "win (or draw) estimation system", and how it defers from official FIDE rules

FIDE Rules of chess 6.7.a

During the game each player, having made his move on the chessboard, shall stop his own clock and start his opponent’s clock. A player must always be allowed to stop hisclock. His move is not considered to have been completed until he has done so, unless the move that was made ends the game. (See the Articles 5.1.a, 5.2.a, 5.2.b, 5.2.c and 9.6)

As would say Magnus Carlsen, a friend of mine, who was with me in a nightcub in Oslo this week-end: please stop talking bullshit
#7 - So, you are eager to turn a polite discussion into a insulting quarrel? Is it the way you usually do?
Well, I'll not be one of those involved into your ego-trip suffering dramas, but don't worry, you'll find plenty of others who'll gladly make you a company there, so your wishes would soon be granted for your utter satisfaction.

Or, as my old pal Korchnoi, with whom I spent a lot of vodka-filled nights, say: Fuck off, my ignorant camarad, and please be so polite to get off from my dick, will you?
Lemuel Gulliver: "if we respect the updated official rules by FIDE, which say that INSUFFICIENT MATING MATERIAL OF YOUR OPPONENT IS NOT ENOUGH FOR YOU TO NOT TO LOSE THE GAME (BUT DRAW), IF YOU RAN OUT OF YOUR OFFICIAL TIME!"

You should actually read the official fide rules here (especially 6.9):
www.fide.com/fide/handbook.html?id=171&view=article

Instead the following is correct:
IF IT IS NOT POSSIBLE TO MATE THE OPPONENT EVEN IF OPP MAKES WORST POSSIBLE MOVES [1], IT IS A DRAW. OTHERWISE IT IS NOT.

Here are koncrete examples: http://www.e4ec.org/immr.html

[1] There are well known cases where referee wrongly declared a draw, even if a mate was theoretically possible.
Addon: The algorithm which correctly implements the FIDE rule has not been written yet. Other chess sites also do not correctly implement it. The algorithm needs to be fast, so it needs to calculate such 'absolute draw' positions (aka draw positions where even with worst possible play a mate can not be achieved) in advance and put it into some kind of tablebase. And as you can see in the second link of my previous post (the pawn ending), maybe the algorithm will not be able to calculate all possible 'absolute draw' positions.

There are tons of not yet implemented algorithms, for example algorithms which can detect fortresses. Work in this field has not even started. And it is very probably a gigantic field. Making me laugh hard at the dudes saying 'chess is solved'.

Even a min max algorithm which correctly detects move repetitions (aka strongly connected components) in the tree of position graphs aka chess is not implemented right in most chess engines, for performance reasons. Which explains why some engines sometimes move back and forth but give a clear advantage. Chess is not solved.)

This topic has been archived and can no longer be replied to.