Would a 2000 ELO player beat Kasparov if he had a piece advantage?

Although one piece would still be a tough game for the 2000, such a comment is a wonderful display of ignorance.

For some real information, see pfren's comments on the first two pages.

Forget Kasparvo and change the question to : can a 2000 player beat Houshreddybka 3000.0 running on the NSA computers ; the answer is still "yes in most of the cases".

(I am not yet 2000 FIDE, but I oscillate between 1900 and 2000)

Terence Chapman (who is around 2200 i think) played an odds match against Kasparov a while back - odds of 2 pawns. He lost. However won one very smooth game with a and b pawn odds.

http://www.chessgames.com/perl/chessgame?gid=1238128

Excellent question eh. 

I think it is all scaled.  I am same scale to a 400 ELO player as a 2000 is to a 2700, give or take.  In a sense, players who sacrifice are doing just what the title suggests, going down a piece.  Even if Garry agreed to start off one piece less I beleive he has the ability to work around it.

Closely, kinda, relates to the thread of can a 1300* player beat a GM.  Of course, but, to have a GM, of that calibur, lose to someone several hundered ELO lower...I am not sure lol.  Great question.

As you did not understand my point : I am almost sure I would score significantly positively against "best play" with one piece odds. And I am 100% sure to score 100% with queen odds.

I expect players who broke 2000 to do better.

One piece is just too much. If you don't reckon so, I am forced to pull the rating card and say that you have no clue.

FYI, I played two GMs (hence >2500) in my chess life, including Aronian, got killed each time (without odds), but if you take any of his pieces off the board in the first 20 moves the GM becomes completely losing. They have no magical powers, you know.

Lets try be more realistic, three pawns instead and GK plays white.

Remove h2, a2, and the d2 pawn.

Wouldn't the result depend of the opening ? There are some very tactical line (Najdorf, Dragon, King's gambit and so on) where being a piece ahead is not such an advantage. I don't play the Najdorf or the Dragon, but I play the King's gambit and in my experience (well, I'm 1500-1600, not really an expert.), white's Queen's rook is quite difficult to develop. In fact many game are won or lost before it move at all. So Kasparov playing the King's gambit a rook down might win the game. I'don't think an expert is level with Kasparov in this kind of complicated position.
 And it's quite difficult to play solidly against the King's gambit.

Of course black could chose the Caro-Khan or the French and the situation would be quite different.
 

Yes, but if you are one piece up from the beginning, you are not going into those lines.

For instance (White starting with nothing on b1), after 1.e4 e5 2.f4 Black can just play 2...d6 (which would be a playable but poor move in normal circumstances) and the tactical lines disappear.

@CLINTEASTWOOD : I would be much less comfortable with three pawns odds than with a piece - open lines are valuable too. I would prefer (for Black) b2,c2,h2 odds for instance to your a2,h2 and d2.

True, as white ~2800 player vs 2000 = piece of cake.

Three pawns is more possible because one piece down and we can occasionally beat houdini 3 with some time on the clock.

(To post #85) Thanks, <Elubas>, for the recognition :-) It does make me feel better - and now I'm willing to face anybody up to 2700 up a piece. Let's see what damage they can do!

But somebody here suggested playing Kasparov up a Queen - if the guy agrees to play me like that, I don't seriously expect him to get one draw out of fifty games.

No offense against Garry the Giant, as I'm sure he'd just laugh it off and agree with this comment himself, after verifying my playing standard. 

Most of the time.

<MrKornKid>

if you are rated 1100, then indeed, the distance from you to 400, is the same as that from 2700 to 2000. 

That much is true. But what does it apply to?

It applies to your odds of scoring against that player, in a game which is a standard game of chess - no odds. 

One difference is, that the more chess players rise in skill (and elo rating), the more a material advantage means, as they are more capable of using it to keep and increase their advantage. 

So - up a piece, 2700 to 2000 is NOT NOT NOT the same as 1100 to 400. 

As 1100, the 400 player will give you back the piece, and probably much more, before too long... as he doesn't have a minimal concept of keeping his pieces and ignores threats all the time. 

However, a 2000 isn't that generous... not even when playing somebody 2700. He knows openings, strategies and structures. The result of the game is very much in question, of course, but nothing like in the match 1100 to 400. Hope this helps :-)

And what is more, <MrKornKid>, there is something that you need to know about chess...

going down a piece with no compensation, is nothing "just like" a sacrifice. 

When you sacrifice, if you're a strong player, you know very well what you are getting for the material invested: you either get a mate at some point, or that same material back, or some improvement in the position, that will make sure that the extra piece isn't advantageous to your opponent. 

In some case, you are already losing, and the piece you're giving is given to make your opponent's life more difficult, because of some circumstances which you create - so again, it isn't "just for free". 

In the lovely sacrifices of Kasparov, a strong player can always explain to you what he gets for a piece. 

But just losing a piece for nothing, is a totally different story, and is generally translated to a loss - which can also happen over hundreds of points. 

That said, rating isn't utterly meaningless, and in practice not every game up "a piece for nothing" will actually be won, as any tournament player of any level knows all too well. 

An example that I still remember from the Israeli league games some 20 years back, had a 2340 player blunder a piece to a 2160 player on move 5, to a simple "Qa4 check" move in the opening, winning a hanging knight on e4, which has just taken the wrong pawn, and the story is well known...

However the 2340 player later won that game :-)

And I have personally saved or won countless games down a piece, against players who were totally my equals...

However, I would personally feel that I have chances, definitely, up a piece against anybody, let's say, not in the World top fifty :-)

I think I would score 100% against Kasparov in a ten game queen odds match, but I'm not entirely certain :)

Though a piece matters much more at higher levels, a 2800 elo player can do more down a piece than people might think -- seeing tons and tons of possibilities really does help! A piece won't do so much until the endgame, so it gives the 2800 a lot of time to at least go down with a bang.

Queen odds is sure absurd, Elubas stop that nonsense.

2000 rated players should have equal chances due to piece odds equals 800 ELO points.

Well, that's what I am speculating, although I'm too lazy to test it against houdini (maybe I should!). If I lose concentration, being put under concrete stress (my opponent), it's not outside of my realm of possibility that I could make a huge blunder or two that puts me at risk of a draw or loss, given 10 full games for that to happen.

I think that perception comes from me watching 1 minute bullet games where a guy would sacrifice his queen very early on (usually for the f2 pawn), and still not only beat his opponents, but generally crush them (in fact he would usually finish up 9 points!). His opponents would not infrequently be titled players! You'd think a queen would hardly even allow problems to be posed, yet pretty good bullet players would make oversight after oversight. Long chess if of course quite different, but it showed me how even minor pieces can do tricky stuff if played by an opponent strong enough to take advantage of every single possibility -- chess is a very tactical game after all. In fact if you asked me about queen odds before these sorts of experiences, I would probably think it would be unquestionably absurd that I couldn't beat anyone with queen odds 100% of the time. Now I would say it's probably absurd, but not unquestionably :)

Of course this player I am talking about was caught for cheating -- I guess he had the expertise and free time to get a bot to make all of his moves for him.

Testing this hypothesis by playing against an engine doesn't prove anything. I think 2800+ players would play very differently when a piece down. They will try to avoid piece trades and will play very tricky. Therefore I think 2800+ players will still win with even a rook down against a 2000 player.

2000 players are no professionals.

No  No  No  No  Nooooooooooooo!!!!

Just for science, it would be interesting to establish a statistics of agreement with the title's affirmation vs. rating.

<Bobby4785>

When I was about 1900 strength, I tried a series of games against Rybka, up a rook, at blitz time controls (but I was rather generous with myself, and gave myself somewhat higher thinking times - 5 minutes + 10 seconds increments - and Rybka only got 5 minutes). 

Up a rook, Rybka lost to me almost all of the games. I didn't try up a queen, as there seemed to be no point in it. 

2000 players are no professionals, it is true... :-) However, each 100 points from your own rating, towards the professional level (which is, probably something in the order of 2400) makes a player that much more professional than you... :-) and the differences acquired with each 100 points, can be quite striking, as you'll often find out.