mcmodern wrote:
It is not impossible, but we will not see it for a long long time.
It could happen today, and again tomorrow. But once in a million years still beats the odds.
Is there any chance that a 1300 rated player can beat a 2700 rated player?
Sort:
mcmodern wrote:
BMeck wrote:
mcmodern wrote:
It is not impossible, but we will not see it for a long long time.
That does not make any sense. What would account for the average strength of a 1300 increasing while the average strength of a 2700 decreasing over time?
What I mean is that if you try say for a trillion times, you might hit that one time. This event has a probability of > 0, so if you try enough, it will happen, it just won't happen anytime soon.
Anything is, theoretically, possible when it comes down to it. But saying it wont happen for a long time still doesnt make sense. The games are independent of one another. Given the problem, the 1300 has the same probability to win the first game that he or she does the billionth.
So, if 1300 keeps playing 2700 their odds of winning increase because their info increases.
Umm in that case, they would not be 1300 anymore right? 1300 is a relative strength, if the player knowledge improves, they would not be 1300. The best chance for this to happen would be some very young and talented player, who is very under rated, maybe someone like Magnus or Wei Yi when they are about 8 or 9 or maybe even younger, but are rated 1300. I know Wei Yi played in Chinese pro league when was 8 or 9, so he was not 1300 at that age.
mcmodern wrote:
So, if 1300 keeps playing 2700 their odds of winning increase because their info increases.
Umm in that case, they would not be 1300 anymore right? 1300 is a relative strength, if the player knowledge improves, they would not be 1300. The best chance for this to happen would be some very young and talented player, who is very under rated, maybe someone like Magnus or Wei Yi when they are about 8 or 9 or maybe even younger, but are rated 1300. I know Wei Yi played in Chinese pro league when was 8 or 9, so he was not 1300 at that age.
"1300 is a relative strength, if the player knowledge improves, they would not be 1300."
That is what I meant by "Given the problem." Earlier in the thread (I dont expect to read it) we settled on the 1300 player being a true 1300. Meaning his or her playing strength is always 1300.
Did that diving guy break his neck?
A person rated 200 points higher than another person has around 75% of winning
So either 0.(25^-7) or 25^-7%
Yeah, not gonna happen
thread*must*continue....
didibrian: The math doesn't work like that. We're dealing with bell curves. Chess ratings are designed to correspond to a normal bell curve with a standard deviation of 200 * √2, which is about 283. A 1400 point rating difference would be about 4.95 standard deviations, giving the higher rated player a winning percentage of 99.999963%. If they play 2.7 million games, the lower rated player can be expected to score one point. (Of course, we are assuming the 1300 player continues to play like a 1300 player, not learning from the losses.)
I don't think this is an unrealistic figure. 2.7 million is a big number, although we see numbers in the billions and trillions so much nowadays that 2.7 million doesn't sound that big. Think about this, however: The odds of being dealt a royal flush in five-card poker is less than a million to 1. When was the last time you were dealt a royal flush?
Based on the elo formula the winning probabilty would be 0.000316%
Charetter115 wrote:
Number too small. Not registering. Try 3.16 × 10^-4.
Based on the elo formula the winning probabilty would be 0.000316%
I do agree with GreedPawnEater. The chance is far higher than 0.000316 because the GM´s blunders more often than that. I expect at least one big blunder in 1000 games among average 2700.
But it is difficult to meet a GM in tournaments for a 1300. I guess it is necessary to beat a 1900 and a 2200 first, and then maybe meet the GM in the 3. round. To beat a 2200 it is circa necessary to have 1700+ real strenght (which is possible for a 1300 kid taking GM-lessons)
Yes of course!
We're not just talking about GMs. We're talking about 2700+. Much stronger than most GMs, and I doubt they are likely to blunder in such a way that they would lose to a 1300. It would have to be something completely ridiculous, because I think most people that level could beat a 1300 even a queen down, it'd have to be something like overlooking a mate in 1, which is almost unheard of.
GreedyPawnEater wrote:
The chance is significant. Every grandmaster blunders once in two games. If the 1300 player is good enough to punish the blunder the Gm will lose.
2700+ GMs blunder less often. Every 1300 blunders an average of six times in twenty moves.
GM blunders concede a vital square. 1300 blunders drop a piece.
3 years and 3000 posts on this topic?
Alexander Pope was correct:
Hope spring eternal in the human breast.
Jimmykay wrote:
3 years and 3000 posts on this topic?
Alexander Pope was correct:
Hope spring eternal in the human breast.
There's really only three dozen endlessly repeated.
Say we have an average of 40 different moves, out of which average of 3 are good. So a probability of playing good move on random is around 0.075. You need to play around 70 moves to win a game(GM won't resing earlier). Giving us very small but existing probability of winning - 2.15239444 × 10^-112
It is not impossible, but we will not see it for a long long time.
That does not make any sense. What would account for the average strength of a 1300 increasing while the average strength of a 2700 decreasing over time?
What I mean is that if you try say for a trillion times, you might hit that one time. This event has a probability of > 0, so if you try enough, it will happen, it just won't happen anytime soon.