It depends on how you calculate it. As I recall, under the original formula for ELO, if you were more than 400 points above your opponent, your rating would drop even with a win. The current formula used by FIDE doesn't allow this, and caps off at 400. What this means is that with repeated wins, your rating will only be 400 above your opponent. Larger player pools mean the top players are likely to have higher ratings even if, objectively, they're no better than their predecessors. Given an infinite pool of players, ELO ratings have no upper bound. It's not a perfect analogy, but think of pouring sugar into a pile. The width of the pile is like the ELO spread. As the pile rises higher, it also spreads out. At the same time, it's only going to spread so far because the sugar rolling down the side will only pick up so much speed in the process. But, if you poured sugar forever, the pile would never stop spreading outward.
Maximum Possible Elo?
That sugar analogy actually makes sense if I think of it like a normal distribution. If you have infinite players, there will be someone with an infinitely high rating, as well as those with a rating of basically zero, even though the majority or mean will be basically the same as if you had far fewer players.
Does this mean that we might one day see a player break 3000? Carlsen is at 2850 or so right now.
4.
Yeah 3000 is beaten by hikaru see his profile
see his profile
He wont, he closed the account 3 years ago
That sugar analogy actually makes sense if I think of it like a normal distribution. If you have infinite players, there will be someone with an infinitely high rating, as well as those with a rating of basically zero, even though the majority or mean will be basically the same as if you had far fewer players.
Does this mean that we might one day see a player break 3000? Carlsen is at 2850 or so right now.
Carlsen is at 2882
That sugar analogy actually makes sense if I think of it like a normal distribution. If you have infinite players, there will be someone with an infinitely high rating, as well as those with a rating of basically zero, even though the majority or mean will be basically the same as if you had far fewer players.
Does this mean that we might one day see a player break 3000? Carlsen is at 2850 or so right now.
Carlsen is at 2882
dw some young prodigies will hit 6700 soon, I give it until 2040
That sugar analogy actually makes sense if I think of it like a normal distribution. If you have infinite players, there will be someone with an infinitely high rating, as well as those with a rating of basically zero, even though the majority or mean will be basically the same as if you had far fewer players.
Does this mean that we might one day see a player break 3000? Carlsen is at 2850 or so right now.
Carlsen is at 2882
dw some young prodigies will hit 6700 soon, I give it until 2040
I don’t think so, 12 years to almost triple Magnus
That sugar analogy actually makes sense if I think of it like a normal distribution. If you have infinite players, there will be someone with an infinitely high rating, as well as those with a rating of basically zero, even though the majority or mean will be basically the same as if you had far fewer players.
Does this mean that we might one day see a player break 3000? Carlsen is at 2850 or so right now.
Carlsen is at 2882
dw some young prodigies will hit 6700 soon, I give it until 2040
I don’t think so, 12 years to almost triple Magnus
my dog is already 2300, I think humans can hit it very soon as we are intelligent creatures. I think as we explore more sharp openings like the bird we will see ratings skyrocket
i am bad at chess and dont understand it but on chess.com is prob 3200 i dont think it codes to go higher
i also dont understand code
Assuming that Chess *can* be solved for now (and I highly doubt it can), what would be the maximum possible Elo of a player, assuming that they basically never lose, or have perfect play. Is it infinite, or is there a limit.
As an aside, say a 2000 rated player (or 2800, I guess it matters little) played a 1200 rated player (or computer, again, it doesn't matter much) 10,000 times or so, winning all games. At what point would their rating stop increasing, if ever? Is it possible for an Elo to go down by winning, or would it continue indefinitely at an exponentially slower rate (or would it just get to +0 for winning)?
Sorry if this is a) a load of stupid questions or b) has been asked before, I'm just curious. For the record my rating is low, it's just hypothetical for me.
Maximum possible Elo is 999999999 ( I is 1000000000).
[...] As I recall, under the original formula for ELO, if you were more than 400 points above your opponent, your rating would drop even with a win. [...]
That is definitely not the case, and never was.
In fact if the rating difference is merely 400, then in 100 games its likely that sooner or later the 400 points below opponent will score a win. At this point it would be rare, but it wouldnt be impossible at all.
People fight with people 400 points below all the time, for example if Magnus Carlsen fights with an IM or FM.
That sugar analogy actually makes sense if I think of it like a normal distribution. If you have infinite players, there will be someone with an infinitely high rating, as well as those with a rating of basically zero, even though the majority or mean will be basically the same as if you had far fewer players.
Does this mean that we might one day see a player break 3000? Carlsen is at 2850 or so right now.
Carlsen is at 2882
Carlsen WAS a 2882, about a decade ago, at age 27.
Right now he is something like 2830.
About the question of maximum ELO, I'm not aware that thats known.
ELO measures the relative difference between players. It doesnt rate against a perfect player, because we dont have such a perfect player and no way to even estimate how well a perfect player would perform.
If we would have such a perfect player, which in my understanding would require a quantum computer of enormous capabilities, then we could measure how well this perfect player would perform against Stockfish, or whatever chess program is the strongest at that time. Assuming that Stockfish would be any close to this perfect player, we could get a percentage of losses of Stockfish and know the exact distance to perfection. We know the approximate strength of Stockfish, currently with version 18 about 3700 (it of course depends upon the speed of the computer, too), so then we would know the upper limit of ELO that is possible.
Chances are of course that Stockfish is still so bad that the perfect player still consistently wins. Then we dont know the rating of the perfect player, only that it has to be at least about 800 points better than Stockfish. So that would be currently 4500.
I would guess that Stockfish is now so close to perfect play that it is probably not so bad that it always loses to perfect play. So the perfect play would, I would guess, to be somewhere between 4000 and 4500.
gddddddddddddd!!!
cueb
i am bad at chess and dont understand it but on chess.com is prob 3200 i dont think it codes to go higher
i also dont understand code
then y talk
Assuming that Chess *can* be solved for now (and I highly doubt it can), what would be the maximum possible Elo of a player, assuming that they basically never lose, or have perfect play. Is it infinite, or is there a limit.
As an aside, say a 2000 rated player (or 2800, I guess it matters little) played a 1200 rated player (or computer, again, it doesn't matter much) 10,000 times or so, winning all games. At what point would their rating stop increasing, if ever? Is it possible for an Elo to go down by winning, or would it continue indefinitely at an exponentially slower rate (or would it just get to +0 for winning)?
Sorry if this is a) a load of stupid questions or b) has been asked before, I'm just curious. For the record my rating is low, it's just hypothetical for me.