How extensive is this calculus that puts pieces on the same plane as differences in rating? It would be nice to have a table that would allow us to look up any two ratings and see an approximate difference in terms of pieces. I wonder, for example, how much of an advantage a 1700 would need to be favored over Kasparov.
Would a 2000 ELO player beat Kasparov if he had a piece advantage?
I doubt it, Kasparov gave a simul to the Czech team, and won. How can he not beat a 2000 player down a knight?
I doubt it, Kasparov gave a simul to the Czech team, and won. How can he not beat a 2000 player down a knight?
Because while he may be a whole universe beyond me in positional and tactical understanding and in calculation, a piece is a piece is a piece.
A piece is a piece... so i would expect a 2000 ELO could beat Kasparov. Unless, of course, Kasparov sends some critical attack that dominates the 2000 ELO guy.
Perhaps we can try an experiment. Pfren or Joeydvivre should try playing against one of the strong chess engines at a strength of around 2800 and see if they can win an odds game (without takeback). Then post the game here.
It depends on the piece and scenario.
His color + piece he loses
White + knight = win or draw (for Kasparov)
White + bishop = win or draw
Black + knight = any result
Black + bishop = draw or loss
Nimzowitch was known to give earthlings à rook advantage but asked A2-A3 in return. Hè then always played italian game where THE pawn on A3 made that Bb4 was not possible after THE central Exchange and people had to play Bb6 and hè demonstrated then THE well known pawn roller plus king attack and always won... Hè was à trickster...
À new chess rating system! In summary I learned today on chess.com :
Fide ELO = USCF + 150
USCF 50 = 1/2 pawn
So fide 1850 = 20 pawns !
How many pawns are you?
I wouldn't find it hard to convert an extra piece into a win, even if I were playing Kasparov. Provided I have much time to think (the regular 1:30 +15s.) I don't think I'd make huge blunders that would give Kasparov an edge. It's just about simplyfying.
You also have to differentiate -- Kasparov now or Kasparov in his prime. In his prime I can't image one piece would be enough for any 2000 ELO player to beat him. At that time he was arguably the best chess player in the history of the game. My guess is he would get it out of book quickly and use his huge tactical edge to get the piece back. If the 2000 could simplify he would have a chance but obviously Kasparov would make that difficult.
I think I could convert an extra piece, even if my opponent were Kasparov.
I totally get the inclination that when you have an extra piece you should have too much control over the game for things to go wrong, but I think you might get a clearer picture if you try to beat your engine with piece odds (Maybe take away the engine's queen's knight or something). If you do, that would be extremely impressive.
In my experience, the engine would find ways to keep your pieces from going to where they want to go due to tactics. I wouldn't blunder material (at least not for a while!), but I would keep having to make positional concessions because I'd have to put my pieces on squares that defend all of the potential threats the engine could come up with. By the time my position was merely slightly better than the engine's even still in the middlegame, I quit, figuring I'd probably end up losing! Granted, that was a few years ago, but even now I don't think I could do it.
I recall reading some mathematical calculation (I have no idea how they do it btw; it's the same mystery of how they determine the possible ratings of players way into the past; it might not be accurate) that Kasparov in his prime would be a slight favorite to win against a 2100 whom was given piece odds.
Bottom line, I recommend anyone who is confident they can beat anyone at piece odds to play the strongest engine they can find under those conditions.
Kaufman will lose pretty easily against a 1800 player with knight odds, me too, of course. Talking about 1800 FIDE, not USCF, which is inflated by 100-120 points, at least.
Speaking as a player who is actually the equivalent of the 1800s in FIDE rating, I could easily see myself gradually making enough mistakes to give my GM opponent a gradually better position, resulting in some sort of position where my opponent may have won a pawn back and created some kind of fortress situation where despite my piece it's so hard to break through without precise play. If I were to predict my result in a piece odds game against GM Kaufman, I would be slightly leaning toward a draw, although a win would be possible; also note, I would even consider losing that game a definite possibility.
Let's say a bishop, or a knight.
What do you think? (I'm asking those players who have at least such ELO)