Roughly how much ELO does it cost a GM or top player to play blindfolded? For instance, if a 2600-rated GM were blindfolded, against what rating of opponent would they be expected to score roughly 50%? Or, what ELO games roughly correspond to the quality we find when looking at blindfolded games between top GMs (say, 2750-rated) like at Melody Amber?

I'm aware that there will be some variation (e.g. Korchnoi was terrible at blindfold), but there should also be consistency - most top GMs have no trouble playing blindfold and the fact that many top players enjoyed success at Melody Amber suggests none of that group are massively less affected by the blindfold than the rest.

Some estimates I've found online already:

  • A Chess.com thread where a user considers Larry Christiansen's performances and gauges the rating handicap of playing blindfolded to be about 500 points
  • A SE post concerning blind players. Here the number 600 points is thrown out there.

These strike me as overestimations (and are called out in the threads themselves for being arbitrary). For example, even the lower bound suggested in the blind players thread (450 points) would make the three top blind players all individually on par with Carlsen or better were they not blind.

As for Christiansen, that's the kind of thing I want to know more about. I presume there are no examples of 2800-rated players testing themselves in blindfold against 2200-rated players. (Examples from say Morphy's time would seem to strengthen my point - Morphy was much stronger than his competition, but his non-blindfold results make very obvious that he was NOT 600+ points stronger, or even close to that, while he had no trouble beating most opponents blindfold.) So maybe we can use examples of 2600+ rated GMs playing against opponents around 2000-2300 to make a judgement?

  • I don't have an answer but I have an idea on how we could get there. First, let's assume we can approximate the average rating of White (or Black) players given a large enough collection of rapid games, by looking at the average number of inaccuracies, mistakes and blunders (probably a better estimator than ACPL). Then we could look at all the games from the Amber Blindfold tournament and estimate that average. Now just compare it with the player's actual rapid rating average.
    – kamekura
    Commented Mar 19, 2021 at 8:41
  • @kamekura thanks for the intelligent response. Perhaps, given the recent interest this post seems to be getting, you could conduct this analysis and provide it in a reply? Commented Apr 8, 2021 at 21:02
  • To be honest, I'm not sure the assumption I made is true. We could look at CAPS instead. Either way it could be quite a lot of work.
    – kamekura
    Commented Apr 14, 2021 at 5:11

2 Answers 2


It really depends on the player, because some top players are used to think about their position without looking at the board (Nakamura, for instance is very often looking somewhere else during his games). For these players, it woud feel quite easy to play blindfold. But there are lots of players, who never think without looing at the position on the board. For those, it would be more difficult.

  • Not really what I was asking (I am looking for a way to quantify the average performance/game-quality drop, for an elite GM). 'Quantify' and 'average' are important keywords here. Commented Aug 4, 2021 at 21:27

It depends. Some can play multiple games at once. Others would struggle to play just one at a much lower level. There is no way to apply a correction factor that would work for all players. I played blindfold as an A player and beat an expert easily.

  • 2
    Please try to reference chess skill in a way that is familiar to all users. I had to check what "A" or "expert" mean.
    – David
    Commented Apr 6, 2021 at 11:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.