11

Magnus Carlsen just reached a rating of 2889.2 today. Considering the way that Elo works, inflation and deflation and relative strength of Carlsen, I was wondering whether it is theoretically possible for him to reach Elo 3000? If so, is it practically possible?1


1. I know my second question is a bit opinion based, but I actually want to know your opinions!

  • I highly doubt it – Lynob Apr 22 '14 at 11:56
  • It is absolutely theoretically possible. It's very unlikely, though. – dfan Apr 22 '14 at 13:33
  • Not if Grischuk does it first! ;) – Jeff Davis Apr 22 '14 at 14:47
7

The average Elo rating of the e.g. top 10 players in the world is rising every year. I think that the relative distance between the number one rated player and the other 9 players in the top 10 becomes relevant in this context. This distance can be estimated by subtracting the average of the ratings of the 9 players (ranked 2-10 in the world) from the rating of the top player (today, Carlsen) and dividing this by the rating of the top player.

Increasing this distance is difficult for the top rated player, because every new game played by the top player has a certain amount of risk and uncertainty. I think that Carlsen will have a hard time reaching an Elo rating of 3000 exclusively by increasing the distance to the players ranked 2-10. Yet, if the relative distance stays the same of the following years and the average rating of the top 10 continues to increase, then it is definitely possible for Carlsen to reach Elo 3000. The second ranked player at that moment would probably have around Elo 2950.

  • So are you saying he should throw a couple games? J/k, but perhaps the top 9 players should have higher ratings, so when they do play him if he wins it would bring him closer to 3k. – MDMoore313 Apr 22 '14 at 19:20
3

It's not so hard: he only needs to beat 2800 players 50% of the time ( drawing the other 50% which will give him 75% of total points).

Given his record, this does not seem practically likely.

0

Theoretically, yes very much so, if Carlsen organised a 100 game match between himself and a 2100 rated player(who he theoretically has a 99.99% chance of beating in each game) and for each win he'd get a rating gain of one elo (by Fide calculations), then that'd be a 100 elo gain putting him at say 2977 or thereabout...so all he's got to do is organise a 150 game match.

  • I think it is only 0.8 points. So he has to play more games or each game separately. And why take on a 2100 player? A 1500 player will do even better. Or did they change some rules recently? – BlindKungFuMaster May 15 '15 at 8:01
  • True BKFM, a win against a 1500 or 2100 would fetch him the same 0.8 gain regardless as both are over 400 elo weaker, so then a 1500 it is, he'd up it to a 180 game match then, no need to have separate games, theyr all encompassed under one match over a period of say 10 months. – Rickka May 15 '15 at 8:12
  • 1
    This is incorrect in so many ways. 1. The match would be rated as a whole, not game-by-game, so Carlsen's score of +100-0=0 against a player he'd be expected to beat 99.99% of the time would result in no significant rating gain. 2. Games after the match is won don't count, so Carlsen would only get rated for +51-0=0. 3. Only FIDE-sanctioned matches and tournaments are included on the FIDE rating list; the FIDE Qualification Commission would obviously refuse to rate this match. – David Richerby Nov 25 '18 at 12:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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