I might have a computerlike style in some games? Is it so that computermoves does require so deep calculations that humans wont go for them? I do calculate unpresicely on intuition or guesstimation, so I often choose moves that has a far out plan, without calculating enough. I think that if it works , it works, and the only way to find if it is right, is to try playing it.
Because of that style, and general lack of experience, I do not perform consistent, and might play one game at 2000 strenght, and the next at 1200 (blundering or miscalculating)
The way I see it is, if you really are capable of playing at 2000 and you play at 1200 in 50% of your games, that isn't normal. I seriously doubt your averages work out that way. I believe the creators of chess rating systems themselves don't agree that is how games would be played considering this is how the math breaks down.
Arpad Elo set the ratings table up so that the standard deviation of performance in a single game is defined to be 200 rating points. He assumed that player's performance in a single game would vary according to a "normal" or "Gaussian" distribution. This would mean that about 68% of the time a player would perform between -200 and +200 points around his rating in a single game.
For multiple games the deviation becomes less and it becomes less as the square root of the number of games. So the "standard error" in five games is 200 over the square root of five, or about +-89 points, and in ten games +-63 rating points. If your performance is actually better or worse than that then you'll gain or lose points accordingly.
About 86% of the time you should perform within +- 300 rating points of your rating in a single game if your rating is accurate.
In fact that looks like either sandbagging or intermittent cheating of engine use, or perhaps some of both....
2000 and 1200 are ca top and bottom. Everybody is at 1200 or something if they loose an officer. I did beat a 1815 Fide in otb-tournament in september, and 1815 Fide is stronger than 2000 online. My friend Vegma is at 1950 online and maye 1650 fide. He is a bit stronger than me.
Because I did compete a year, 37 years ago, I have some intuitive chess in my blood, which can make strong players bleeding, and because I have been away for 37 years, and only read one chessbook back in my youth, I have a huge amount of knowledgeholes to fill. Therefore I am inconsistent.
I also have lost a couple of games online on tactics, giving away officer , to win something some moves later, but forgotten what I was doing, opening the game three days after, and been playing 30 other games-moves inbetween.
I usually have interesting games online against 1500-1900-players, but one mistake, and the 1400`s bites me.
I belong among 2000+ players, but I have some work to do to get there and stabilize on that level. My plan ist to get 2000 online for christmas 2015 (and 1800 for christmas this year). These days I feel like beeing behind my plan.
I regretfully inform you that if you make one mistake against 1400's here and lose, you don't belong amongst the 2000's. You may want to see yourself there or think because of you having a good game while your stronger friend had a bad one that, automatically grandfathers you by association but, it doesn't work that way.
I have beaten an 1800+ personlity on here and a chessmaster game I have, drew a few other 1800's here, and a 1900 here in 960 random chess, but that doesn't make me think I deserve or belong with the 1700's, even though I have beaten a handful of them here too.
In fact, a 2000+ will still make some mistakes but, the mistakes the 1400 makes are going to be amplified so much by the profound moves played against them, the mistake of 2000 wont matter. A true 2000, easily could gambit a pawn or liquidate one and crush a 1400, nearly everytime they play, as long as they both have established ratings.
Here, my last otb-tournament:
I am Erlend Sæteren. I got 3 points. My victories were against 1814 Fide (Aarnes) and fide 1638 (Sjoberg), and unrated (Gusarov)
My friend Vegard Martinsen got 3,5 points. He is 1950 online. I am not at his level yet, but there is only a lot hard work to get there in ca one year.
http://turneringsservice.sjakklubb.no/standings.aspx?TID=Hostturneringen2014NordstrandS-NordstrandSjakklubb
ALso, you have to consider that, you maybe are really familiar with your friends' tendencies but, that won't necessarily get you far when confronted continually by players at 2000, who are putting you in unfamiliar positions...
I might have a computerlike style in some games? Is it so that computermoves does require so deep calculations that humans wont go for them? I do calculate unpresicely on intuition or guesstimation, so I often choose moves that has a far out plan, without calculating enough. I think that if it works , it works, and the only way to find if it is right, is to try playing it.
Because of that style, and general lack of experience, I do not perform consistent, and might play one game at 2000 strenght, and the next at 1200 (blundering or miscalculating)
The way I see it is, if you really are capable of playing at 2000 and you play at 1200 in 50% of your games, that isn't normal. I seriously doubt your averages work out that way. I believe the creators of chess rating systems themselves don't agree that is how games would be played considering this is how the math breaks down.
Arpad Elo set the ratings table up so that the standard deviation of performance in a single game is defined to be 200 rating points. He assumed that player's performance in a single game would vary according to a "normal" or "Gaussian" distribution. This would mean that about 68% of the time a player would perform between -200 and +200 points around his rating in a single game.
For multiple games the deviation becomes less and it becomes less as the square root of the number of games. So the "standard error" in five games is 200 over the square root of five, or about +-89 points, and in ten games +-63 rating points. If your performance is actually better or worse than that then you'll gain or lose points accordingly.
About 86% of the time you should perform within +- 300 rating points of your rating in a single game if your rating is accurate.
In fact that looks like either sandbagging or intermittent cheating of engine use, or perhaps some of both....
2000 and 1200 are ca top and bottom. Everybody is at 1200 or something if they loose an officer. I did beat a 1815 Fide in otb-tournament in september, and 1815 Fide is stronger than 2000 online. My friend Vegma is at 1950 online and maye 1650 fide. He is a bit stronger than me.
Because I did compete a year, 37 years ago, I have some intuitive chess in my blood, which can make strong players bleeding, and because I have been away for 37 years, and only read one chessbook back in my youth, I have a huge amount of knowledgeholes to fill. Therefore I am inconsistent.
I also have lost a couple of games online on tactics, giving away officer , to win something some moves later, but forgotten what I was doing, opening the game three days after, and been playing 30 other games-moves inbetween.
I usually have interesting games online against 1500-1900-players, but one mistake, and the 1400`s bites me.
I belong among 2000+ players, but I have some work to do to get there and stabilize on that level. My plan ist to get 2000 online for christmas 2015 (and 1800 for christmas this year). These days I feel like beeing behind my plan.
I regretfully inform you that if you make one mistake against 1400's here and lose, you don't belong amongst the 2000's. You may want to see yourself there or think because of you having a good game while your stronger friend had a bad one that, automatically grandfathers you by association but, it doesn't work that way.
I have beaten an 1800+ personlity on here and a chessmaster game I have, drew a few other 1800's here, and a 1900 here in 960 random chess, but that doesn't make me think I deserve or belong with the 1700's, even though I have beaten a handful of them here too.
In fact, a 2000+ will still make some mistakes but, the mistakes the 1400 makes are going to be amplified so much by the profound moves played against them, the mistake of 2000 wont matter. A true 2000, easily could gambit a pawn or liquidate one and crush a 1400, nearly everytime they play, as long as they both have established ratings.
Here, my last otb-tournament:
I am Erlend Sæteren. I got 3 points. My victories were against 1814 Fide (Aarnes) and fide 1638 (Sjoberg), and unrated (Gusarov)
My friend Vegard Martinsen got 3,5 points. He is 1950 online. I am not at his level yet, but there is only a lot hard work to get there in ca one year.
http://turneringsservice.sjakklubb.no/standings.aspx?TID=Hostturneringen2014NordstrandS-NordstrandSjakklubb