Average opponent rating...

Originally posted by XanthosNZ
As I am an expert in upper-division string theory I declare this string pointless.

Well, it certainly is now. Maybe in a couple months when Doctor Scribbles up there has grown tired of herassing chess enthusiasts and moves on to a bingo website or something then we can have a reasonable discussion about the idea.

Originally posted by ItalyBoyBlue
Well, it certainly is now. Maybe in a couple months when Doctor Scribbles up there has grown tired of herassing chess enthusiasts and moves on to a bingo website or something then we can have a reasonable discussion about the idea.

Don't worry I'll still be here to point out that your idea has no mathmatical merit. Did you even read Rag's post where he asked for some information?

I don't like the ELO system, like Opponent's average rating, but do suggest Italyetc has a point that people miss.

A 1400 who plays 1000 rated vs. 1400 who plays 1800... the 1400 (who played 1800s) would have an ADVANTAGE: they would be able to pick up some strategies that the elite (1800) played, and use it against the 1400 (1000) who would NOT have that advantage. It is a minor advantage, so I don't think ratings differ too much. But I do agree that a 1400 (1800 opponent) would be more likely to beat the 1400 (1000 opp).
That seems logical:

NO new rate system. I like this one
YES show opp average rating - how can it hurt?
YES a 1400 (1800 opp) has advantages over a 1400 (1000 opp).

Originally posted by Ramned
I don't like the ELO system, like Opponent's average rating, but do suggest Italyetc has a point that people miss.

A 1400 who plays 1000 rated vs. 1400 who plays 1800... the 1400 (who played 1800s) would have an ADVANTAGE: they would be able to pick up some strategies that the elite (1800) played, and use it against the 1400 (1000) who would NOT have that ...[text shortened]... erage rating - how can it hurt?
YES a 1400 (1800 opp) has advantages over a 1400 (1000 opp).

The 1400 player picks up better play by osmosis and then would increase his true rating accordingly. The current system deals with this already.

Originally posted by Ramned
But I do agree that a 1400 (1800 opponent) would be more likely to beat the 1400 (1000 opp).
That seems logical

But it's incorrect. In the Elo system, any two equally rated players expect even equity in a game between them, no matter how those ratings were achieved. That's just what an Elo rating means.

If you're willing to give me better than even odds in a wager on the two players in question, I'll take whichever player you want me to and I'll have a positive expectation.

Thankyou. Also I'd like to say that low level players make a lot more mistakes and people who intentionally try to play low level players simply because it is a more immediately gratifying experience become used to the sort of play that depends on their opponents making aggregious mistakes (i.e. losing a queen to a fork) rather than tactical mistakes (i.e. failing to control a portion of the center squares without ample counter strategy, or failing to protect the King for combination attacks) Because of this the two players would also differ in the comprehension of the true fundementals of chess.

Originally posted by DoctorScribbles
But it's incorrect. In the Elo system, any two equally rated players expect even equity in a game between them, no matter how those ratings were achieved. That's just what an Elo rating means.

Your using the system to justify the system, and ignoring the argument. It's like using the word in the definition of the word, it's faulty logic. You're not really comprehending the argument, you just like to have your say. It's because you're silly.

Originally posted by ItalyBoyBlue
Your using the system to justify the system, and ignoring the argument. It's like using the word in the definition of the word, it's faulty logic. You're not really comprehending the argument, you just like to have your say. It's because you're silly.

Put your money where your mouth is. Let's wager.

You find a 1400 player who has played primarily lower opponents, and one who has played primarily higher opponents.

You offer me 1.1 to 1 odds if I pick the "lower" rated player to win.

They will play 5 games between them, and whichever side gets the most points wins the wager.

What do you say?

Originally posted by DoctorScribbles
Put your money where your mouth is. Let's wager.

You find a 1400 player who has played primarily lower opponents, and one who has played primarily higher opponents.

You offer me 1.1 to 1 odds if I pick the "lower" rated player to win.

They will play 5 games between them, and whichever side gets the most points wins the wager.

What do you say?

Eh, I say you're rediculous. Regardless of the outcome of the match we supposedly would waste our time orchestrating, I still would like to have the average opponent rating shown. You know why? Because I'm smart enought to utilize it!!! HAHAHAHA do you actually think I would make contact with a jerk like you so that I could wage money on a farcical situation, one that would take hours of effort and that you would more than likely attempt to rig. YOU'RE REDICULOUS. Don't blame me because your momma raised you stupid, just stop posting you child.

Originally posted by ItalyBoyBlue
Don't blame me because your momma raised you stupid, just stop posting you child.

😀😀

Originally posted by ItalyBoyBlue
Eh, I say you're rediculous. Regardless of the outcome of the match we supposedly would waste our time orchestrating, I still would like to have the average opponent rating shown. You know why? Because I'm smart enought to utilize it!!! HAHAHAHA do you actually think I would make contact with a jerk like you so that I could wage money on a farcical situ ...[text shortened]... EDICULOUS. Don't blame me because your momma raised you stupid, just stop posting you child.

There are only two ridiculous things in this thread.

1) Your spelling.
2) Your understanding of math.

Please excuse my inexcusable spelling. I know I am near iliterate. I hope you can even read this apology, the spelling used herein is likely most disastrous.

P.S. Please post a negative comment about me in the spelling forum rather than in here the site idea forum.

Originally posted by ItalyBoyBlue
Please excuse my inexcusable spelling. I know I am near iliterate. I hope you can even read this apology, the spelling used herein is likely most disastrous.

P.S. Please post a negative comment about me in the spelling forum rather than in here the site idea forum.

Please answer the questions Rag asked if you wish your stupid idea to be taken even remotely seriously.

Originally posted by Ragnorak
As requested by me before, could you provide the maths as to how this statistic would be of any use whatsoever? Show us what percentage of games a 1500 would have to win against exclusively 1200s to maintain 1500 rating, and what percentage of games a 1500 needs to win against exclusively 1800s to maintain a 1500 rating.

Also, can you explain how this s your reason for needing this stat keeps changing.

Do the maths, get back to me.

D

Ok, I did miss this post thanks to the numerous other rediculous posts I was distracted by. I will try to address all of the points in this comment. I will not waste my time doing the so called "maths" but I will submit you would only have to win a small fraction of games against 1800 players and only have to not lose only a small fraction of games against 1000 players to maintain your 1400 rating. This much is obvious.

Rag, is right that it only takes the throwing of a few games against low level opponents to greatly reduce your rating. Because of this the average opponent rating should be kept to about the last 20-25 completed games or so. The low ratings of those players will greatly reduce the average. Also this system would be more useful to indicate a rating that has been inflated/surpressed habitually rather than intentionally as purposed by Rag because their max rating within the last thirty days and all of their ratings over the last 300 games are already on display.

You may be right that there is only one best move to make. However that move may be relatively none better than many other different moves that lead to many other different lines of play. If someone where to always be able to assert the best move from any position they would never lose and chess itself would finally be unraveled. Luckily even the best super computers are unable to perform this task. Chess is not yet like checkers or tick tack toe where one player will win 100% of the time with the right strategy and the choice whether or not to go first. This is because the number of different moves and the number of different countermoves for every move and so on keep the number of different possibilites of play in chess truly astronomical.

Read the books written by the best chess players who have ever lived. They agree that much of chess is psychological, no matter how objective we try to keep it. The internet keeps chess very impersonal. For games like poker, internet play is absurd; as a large part of the game is judging your opponent's visual reaction to the game. In chess there is an element of this as well no matter how small. Since we all don't know each other like professional poker players and chess masters typically do this statistic would help indicate a player's character.

I do not believe that this implementation would be a drag on the servers. They would simply record the rating of your opponent at the time the game has begun and as that game finished it would be introduced into a log file and that file would be updated with every game using the simple mean formula and the 25 four digit numbers it would store. All this would drag down the servers? It would be certainly have to be then the LAST STRAW that broke the camels back, just a small few lines of code lumped in with many other larger lines of code truly responsible for slowing the server down.

Life is too complicated to be fully explainable mathematically. Thats what statistical formulas are all about. They do their best to describe certain facets of real situations. The ELO formula seeks to describe a player's potentional, and while it works very well in tournament play it can still be misleading in internet play.

Again I see no reason how the statistic would harm the functionality of the site, and if you think it is useless then ignore it if it is ever implemented. If it is ever put to a vote, vote against it. But please do not continuously post non-specific complaints, one is enough. This forum is for discussions not posting matches.

I apologize to Rag and everyone else besides DS for my profanity and my short tirades. I was initially very annoyed by the constant repitition of invalid statements and rude remarks that were littering this string.

Originally posted by ItalyBoyBlue
Ok, I did miss this post thanks to the numerous other rediculous posts I was distracted by. I will try to address all of the points in this comment. I will not waste my time doing the so called "maths" but I will submit you would only have to win a small fraction of games against 1800 players and only have to not lose only a small fraction of games agai onstant repitition of invalid statements and rude remarks that were littering this string.

Ok, I did miss this post thanks to the numerous other rediculous posts I was distracted by.
I've already shown you the correct way to spell ridiculous once on this page. At least make some effort to spell correctly.

I will try to address all of the points in this comment. I will not waste my time doing the so called "maths" but I will submit you would only have to win a small fraction of games against 1800 players and only have to not lose only a small fraction of games against 1000 players to maintain your 1400 rating. This much is obvious.
You don't want to waste your time with math and yet you claimed that your suggestion would help by appealing to math? And yes that is obvious and is in fact exactly the way in which the ELO system accounts for players playing different levels of opponents.

Rag, is right that it only takes the throwing of a few games against low level opponents to greatly reduce your rating. Because of this the average opponent rating should be kept to about the last 20-25 completed games or so.
How does that even help anything? So you know that the last 25 people I played had a 1400 rating. How would that help you determine if I was over-rated, I could have reached my current rating before those 25 games.

The low ratings of those players will greatly reduce the average. Also this system would be more useful to indicate a rating that has been inflated/surpressed habitually rather than intentionally as purposed by Rag because their max rating within the last thirty days and all of their ratings over the last 300 games are already on display.
The rating graph does a good job of showing people with lowered ratings and a planned extension of it will do even more. An average of the last 25 opponents will do nothing.

You may be right that there is only one best move to make. However that move may be relatively none better than many other different moves that lead to many other different lines of play. If someone where to always be able to assert the best move from any position they would never lose and chess itself would finally be unraveled. Luckily even the best super computers are unable to perform this task. Chess is not yet like checkers or tick tack toe where one player will win 100% of the time with the right strategy and the choice whether or not to go first. This is because the number of different moves and the number of different countermoves for every move and so on keep the number of different possibilites of play in chess truly astronomical.
Read the books written by the best chess players who have ever lived. They agree that much of chess is psychological, no matter how objective we try to keep it. The internet keeps chess very impersonal. For games like poker, internet play is absurd; as a large part of the game is judging your opponent's visual reaction to the game. In chess there is an element of this as well no matter how small. Since we all don't know each other like professional poker players and chess masters typically do this statistic would help indicate a player's character.
This whole section has nothing to do with the topic.

I do not believe that this implementation would be a drag on the servers. They would simply record the rating of your opponent at the time the game has begun and as that game finished it would be introduced into a log file and that file would be updated with every game using the simple mean formula and the 25 four digit numbers it would store. All this would drag down the servers? It would be certainly have to be then the LAST STRAW that broke the camels back, just a small few lines of code lumped in with many other larger lines of code truly responsible for slowing the server down.
So you argue for its inclusion because it wouldn't do any real harm? That's not a very good argument. There are actual real improvements that RHP needs well before we start adding useless stats for no reason other than that "they won't do any harm".

Life is too complicated to be fully explainable mathematically. Thats what statistical formulas are all about. They do their best to describe certain facets of real situations. The ELO formula seeks to describe a player's potentional, and while it works very well in tournament play it can still be misleading in internet play.
The ELO formula works just as well here as it does in tournaments. While the possibility of abuses exist (see the recently banned noddle killa for an example) this is true in OTB chess as well (see Claude Bloodgood for an example). Your rambling about math is just that. Rambling. I'm an engineer, the only true statistical math I use is Thermodynamics (behind a layer of abstraction) and experimental uncertainty. Almost every formula I use is indeed an approximation but that doesn't have anything to do with statistical formulas.

Again I see no reason how the statistic would harm the functionality of the site, and if you think it is useless then ignore it if it is ever implemented. If it is ever put to a vote, vote against it. But please do not continuously post non-specific complaints, one is enough. This forum is for discussions not posting matches.
Bad logic again.

I apologize to Rag and everyone else besides DS for my profanity and my short tirades. I was initially very annoyed by the constant repitition of invalid statements and rude remarks that were littering this string.
The only invalid statements in this thread have been yours. The only truely rude remarks in this thread have been yours. The only moderated posts in this thread were yours.