GM who lost the least?

Does anyone know which GM in the history of chess had the lowest loss percentage? Let's only count GM's with at least 100 tournament games. It doesn't matter how many wins they had, or how many draws. I'm only interested in the lose percentage. So, since it's the percentage that matters, then 50 lost games out of 500 (10 %), is better than 30 lost games out of 150 games (20 %). Would it be Capablanca? Or...?

Gonna be tough to find GMs without at least 100 tournament games...

I think so too. I just wanted to be sure someone didn't suggest someone who played just five games and won them all, since this wouldn't be statistically significant. 

It also depends on the opposition. I guess Capablanca, Petrosian, Karpov, or Kramnik would be good candidates, but so are WC - Fischer, Kasparov who just won every game they played

Pretty sure it's gonna be capablanca he had a win streak of eight years and as the above poster said he rarely lost even before or after that. Just imagine...not losing for eight years...amazing.

Steve..., that's not true. According to 365Chess, Capablanca lost 199 of the 1,194 games in their database. That's 17 % of his games. that Kasparov lost less than 8 % of his games in the same database. But I wonder... did anyone lose even less than Kasparov?

That 34 games of Capablanca's is a pretty well-known statistic (and 26 of them in tournaments, if I recall)...so I'm very skeptical about this 365Chess figure, which could include simuls and whatnot.  I seem to remember his career loss record was 4%.

Think about it...Alekhine accounted for 6 of those losses in one match!  No wonder nobody expected him to win (who's gonna be able to beat Capa 6 times?!).

Karpov btw was known for having a loss percentage of 5% (but this was before Kasparov came along, as I recall).

Petrosian only had a 22% decisive result in his games (and I'm sure not too many of those were losses, since he too was famously invincible).

Actually, I think Steinitz may have gone longer without a loss (during his 25-game winning streak).

Which brings up another point:  it was less of a strain to go without a loss back when there were a lot less tournaments (not to mention the unevenness of the competition in earlier times).

Now, I've tried to find the well-known 34 losses of Capablanca. Can't find it anywhere. On chessgames.com his number of losses is 47, which certainly is low (7.0 % of the 670 games in their database). In the same database, Karpov's lose % is 9.9, Petrosian's is 8.2, Kasparov's is 7.3, and Steinitz's is 24.5. But are those numbers only from tournament games??? 

RoseQ..., does it need to have a purpose. Isn't it okay just to be curious? That said, to me it does have a purpose.

From chessgames.com (exhibition games excluded)

Capablanca: 7.01 %

Kasparov: 7.29 %

Petrosian: 8.16 %

Kramnik: 8.89 %

Spassky: 9.63 %

Karpov: 9.88 %

Tal: 11.07 %

Anand: 11.26 %

Fischer: 11.35 %

Smyslov: 11.84 %

Botvinnik: 11.97 %

Alekhine: 13.83 %

Lasker: 14.83 %

Euwe: 16.21 %

Steinitz: 24.51 %

Nice, Arctor! Did you calculate all those (as I did with some of them), or did you find a list?

I thought Capablanca Is the one that never got mated. He would resign the games before that happened.

 Yes, I just calculated from the data on each players profile page. For someone who took so many risks, Tal's relatively low losing percentage was one that surprised me

You're right, that was kind of surprising. Good work, Arctor.

I've seen it listed in books several times (along with the 26 losses in tourneys).

However, it is I suppose possible that the list was not complete and may have left out some of his later events.

Still, you can definitely not go by that chessgames.com statistic.  Scroll down to the first page listing the games and you will find game #1 is the odds game he played and won when he was 4! lol  Also that whole first page has a bunch of irrelevant games played before the Corzo match (which is the first event that most anybody would count in his official list of results).

Now that I think about it, I seem to recall that Capa was at 5% and (for a while anyway) Karpov was at 4%, at least until he ran into that Kasparov brick wall.  Still, Karp only lost to the guy 21 times in 144 games (and won 19)...not too shabby, I would say.

I think Bobby Fisher has lost the least.

Thanks, AndyC. Interesting!

It all depends on the competition if you look at the ratings wherein these GM's played. The highest ratio would go to Bobby. Against Taimanov 6-0, same thing with Larsen 6-0. He's also the kinda guy who would forfeit a World championship match game and continue trailing Spassky 2-0! If we can paraphrase the OP question by asking who's the hardest player to beat? Everyone even Capablanca if he met Bobby would say it is HIM! 

Steinitz went nine year without losses in serious games 1873-82, as with Capa mainly because he played very little but there's also that +7 -0 =0 match against Blackburne in 1876. Capa played more but still only five events in those eight years, and in a couple of them the opposition was so weak that it was obvious that he would be undefeated, like Hastings 1919 and the match against Kostic. Kramnik went undefeated a much shorter time in years but the number of games was much higher and the opposition stronger. Still Capa was of course extremely hard to beat.

Bogdan Lalic from Croatia probably has the lowest loss percentage ever:

http://ratings.fide.com/chess_statistics.phtml?event=409081