I think magnus carlssen is more than twice as good as a 1400 player =P.
Twice as good = rating*2?
Sort:
I think the way it works is that MC (live rating 2812) theoretically has the same probability of beating a player rated 1412 in any given game as a 1412-rated player has of beating a hypothetical player whose rating is 12. Not that I've heard of a player rated that low.
The difference between ratings is supposed to refer to one player's probability of beating another player in any given game.
veclock wrote:
If you are twice as good as someone, do you have twice his rating?
How do you calculate?
You have to think in a different way.
You are twice as likely to beat a player rated 100 points lower. 64% vs 36%
chess.com uses Glicko not Elo but the estimate should still apply.
http://www.chesselo.com/probabil.html
http://everything2.com/title/Elo+rating
When there is no true zero, there is no ratio scale. So 80 degrees Fahrenheit is not twice as hot at 40 degrees Fahrenheit.
Estragon wrote:
Vance917 wrote:
When there is no true zero, there is no ratio scale. So 80 degrees Fahrenheit is not twice as hot at 40 degrees Fahrenheit.
Dang! So all those people in Florida this winter might as well be in Minnesota? Who knew?
Actually there is, you just can't use the rating itself. TadDude hit upon it - with established ratings, you will outscore a player 100 points below you by 2-to-1 over a series of games. It is fair to say you are twice as good as him, but you always only say this to yourself, avoiding the scenario tonydal recounts where you mouth off and then the guy wins his one out of three in front of the hot chicks, humiliating you and impressing them.
Hot chicks at a chess match? you Sir, are off your bleeping head.
veclock wrote:
If you are twice as good as someone, do you have twice his rating?
How do you calculate?
Good God NO!!!!!!!!!!
0
#9
I heard strong players say that for every 100 elo/rating points you raise, you need to get twice as good.
tonydal wrote:
Still, there is absolute zero...so maybe we could get this thing to work.
Isnt -412 kelvin or something like that absolute zero?
no. I think ratings are more logrithmic than that.
Fezzik wrote:
CPawn wrote:
tonydal wrote:
Still, there is absolute zero...so maybe we could get this thing to work.
Isnt -412 kelvin or something like that absolute zero?
Thank you, CPawn, for confirming your nationality!
Youre welcome. Hail the North Pole!
Better than being a quitter.
Think of it this way:, just because one ship can cross the Atlantic in 6 days, doesn't mean 6 ships can do it in one day.
rooperi wrote:
Think of it this way:, just because one ship can cross the Atlantic in 6 days, doesn't mean 6 ships can do it in one day.
That analogy may not be apt. If one ship can carry one item in six days, then to carry six items it needs to make six (round) trips, so 12*6=72 days. But with six ships, each one can make just one round trip and we can do this in parallel (simultaneously), so that it takes only 12 days, which is one sixth the time. Here linearity actually does hold true.
BTW, there is no such thing as a negative Kelvin. 0 kelvin is absolute zero, no heat energy, no molecular motion. Its as cold as cold gets.
Vance917 wrote:
rooperi wrote:
Think of it this way:, just because one ship can cross the Atlantic in 6 days, doesn't mean 6 ships can do it in one day.
That analogy may not be apt. If one ship can carry one item in six days, then to carry six items it needs to make six (round) trips, so 12*6=72 days. But with six ships, each one can make just one round trip and we can do this in parallel (simultaneously), so that it takes only 12 days, which is one sixth the time. Here linearity actually does hold true.
Geez, you got to love pedantry:)
OK, if one seagull..........
Chess ratings follow a Gaussian curve, so all you have to do is plug in your rating in the formula f(x) = ae^[-(x-b)^2/2c^2]. where x is your rating, a is the highest number of players with the same rating, b is that most common rating, and c will be somewhere around 1200. find the area under the curve between your rating and and 2900 as well as your rating x2 and 2900. divide those two numbers to see how much better someone with twice as high ranking as you is.
Boom goes the dynamite
complicated? that? If you bothered to look at the math, you'd realize it's not a calculation for an abstract concept like skill, but a of world rankings (3000 is twice as good a 6000 in my evaluation). what part of bell curve didnt you understand lol
Please Sign Up to comment.
If you need help, please contact our Help and Support team.
If you are twice as good as someone, do you have twice his rating?
How do you calculate?