I swear, do you guys even think about things? what part of 2900 did you not understand. It's an estimation that assumes you're not making a living playing chess. Ask your math teacher about the concepts of a domain and estimation sometime
Twice as good = rating*2?
Sort:
Chess ratings are more like Richter's magnitude scale
yeah, when I get serious. You gotta duck for cover when that happens
Sofademon wrote:
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.
Actually, temperatures in negative Kelvin do exist, going by the technical definition of thermodynamic temperature. Though as a fun fact, a system with a temperature in negative Kelvin is in some sense "hotter" than any system with a positive temperature. By the time you get to that point though, it's getting pretty deep into the bits of physics which bore most people out of their skulls, so let's not go there.
TL;DR: /nerdfest.
Let's talk about baking waffles instead. Or scholar's mate.
uhohspaghettio wrote:
psyduck wrote:
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
Psy????
Oh, I think I understand what you mean. Your curve treats the distribution in such a way that it is ratio rather than difference in rating, which is interesting... I never knew we had a clever person on here...?
Thank you, I like being complimented. Of course that means that someone in the middle of the pack is only twice as good as the worst chess player, which I think is about right
Dec 6, 2010
0
#28
Vance917 wrote:
psyduck wrote:
Chess ratings follow a Gaussian curve,
In the real world, the normal distribution has about as much relevance as Santa or the tooth fairy. If the statement were true, then there would be a positive probability for a negative chess score, or for one over a trillion.
by this logic you have have people that are negative tall, too.
TadDude wrote:
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.
Oh, I thought Chess.com used elo... D: Never heard of Glicko before
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.
There should be a true zero. If you reverse Rybka, so that the engine always picks the worse move, it should be pretty close to zero in rating.
woodshover wrote
Good God NO!!!!!!!!!!
Huh?
tonydal wrote:
Say OP, did you get this concept from the "Humiliating game" thread (featuring the following comment by Bojan)?: "Recently i`ve played a game with a 2100 rated player on chesscube, he was saying he is twice better than me, and i told him i am gonna humiliate him in front of hundreads spectators and so I did."
I lol'ed it then, and I'll do the same now.
Acutally not, I just got the question in my head; is it really possible to claim to be twice as good as someone, and if it is, how does this translate in rating?
fezzik, if beating someone 99.7% of the time makes you ~1.414 times better. If Tony=X(Fish) Chucky= X(Tony), Chucky=X^2(Fish). If Chucky = 2 Fish, then x is ~1.414
fezzik, I might mix up a lot of things, but not multiplication. you were adding, but look at the title-"rating*2"
tonydal wrote:
Ah, and now we go tumbling back into geekville...
It's a chess website...
tonydal wrote:
Ah, and now we go tumbling back into geekville...
In Soviet Russia,
geekville goes tumbling back into you!
Chess ratings follow a Gaussian curve,
In the real world, the normal distribution has about as much relevance as Santa or the tooth fairy. If the statement were true, then there would be a positive probability for a negative chess score, or for one over a trillion.