Will Carlsen be first past 2900?

Originally posted by greenpawn34
If the same bunch keep playing the same people again and again.
The best player's grade will continue to rise.

Try it in you club/clan.

Select 10 players. all graded approx 1600

A,B,C,D,E,F,G,H,I & K. with the instructions.

Everyone lose to A
Everyone lose to B except A
Everyone lose to C execpt B & A
Everyone lose to D except C,B,& A
...[text shortened]... key fart about his grade.
He knows how good he is and does not need a number to tell him so.

Yes, but in serious play one player does NOT win every game. If he does, then his rating must rise because he is BETTER than the other players. Ratings are comparisons. If a player has equal ability as a player rated 1800, then his rating will tend to 1800. Its very simple math.

What happens in clubs is you are restricting the pool of players. You can't compare ratings in clubs with overall ratings. If one player in the club is better than all the others, then his rating will rise (comparing to his overall ranking) because he is better than the other players, and his rating will tend towards a value, say 2000, that, in that pool, reflects his ability in comparison to the other players in the club.

Hope this is clear enough for you to understand.

So what's the difference between

A) always playing the same people in a club
B) always playing the same people outside a club

?

toet.

Originally posted by toeternitoe
So what's the difference between

A) always playing the same people in a club
B) always playing the same people outside a club

?

toet.

If he is better his rating will rise, if he is not it won't - his rating will show his ability in comparison to the pool - is all I meant by that.

The smallest club possible, me and my friend, if we play eachother, and only eachother, my rating (if I am the best of us two) will rice, but not infinitely. At some level I don't get anymore points of him, and lose 32 points if I lose agame with him, which I do once in a while according to probability. After a while we will enter a equilibrium, and the different in rating defines how much better I am compared to him.

Originally posted by greenpawn34
If the same bunch keep playing the same people again and again.
The best player's grade will continue to rise.

Try it in you club/clan.

Select 10 players. all graded approx 1600

A,B,C,D,E,F,G,H,I & K. with the instructions.

Everyone lose to A
Everyone lose to B except A
Everyone lose to C execpt B & A
Everyone lose to D except C,B,& A

key fart about his grade.
He knows how good he is and does not need a number to tell him so.

if A wins 1600s every time, he's not 1800 but MUCH stronger (infinitely strong). a 200 point difference in elo system corresponds to scoring 0.75, that's how the rating formula is rigged. it's the fundamental idea behind elo rating systems.

if you score better, your rating WILL rise. if you score worse your rating drops. if you score 0.75 you'll be exactly 200pts stronger than your opponent. the 1 lost game out of 4 will take exactly the points the 3 won games gained. as the rating difference grows, you'll gain less and less from every win, and lose more points from a loss. at the rating level which corresponds to your playing strength, the amount you win/lose meets the frequency of wins/losses thus forming an equilibrium. and a stable rating.

above that point your frequency of wins doesn't cover the amount of points a loss takes from you, and the rating drops towards the equilibrium point. below the equilibrium point you don't lose 'enough' so your rating goes up. your rating ALWAYS gravitates towards that equilibrium point.

that's how every elo rating system works, and why the rating can't ever 'escape the gravity'.


(and yeah, magnus probably doesn't give a rats ass about his rating...)

Originally posted by toeternitoe
So what's the difference between

A) always playing the same people in a club
B) always playing the same people outside a club

?

toet.

different player pools which are in a different phase of 'maturity'. a newly created pool is very different to an older pool.

the exact dynamics of a pool depends on a number of factors, which are in continuous flux, and cause the ratings to be incomparable to some other pool. even if the rating formula was exactly the same. (in practice there are small differences between formulas on different sites)

Originally posted by greenpawn34
Without a doubt there is false inflatiton going on.

If you get bunch of 1600 players playing each other again and again and again.
and one lad is better than the rest - say an 1800 player.

Then he will reach 2400. This does not make him a 2400 player.

It's not Fischer's fault that these incest tournaments did not exist then.

The whole thing ...[text shortened]... .
Though I'd expect a Carlsen victory.
His age, determination and skill wil come through.

Here is a good "real life" example of GP's "closed pool rating inflation" concept.

Claude Bloodgood was a murderer in prison in my home state, and he resulted in the USCF changing its rating system:

(From wikipedia)
"High rank possibly via manipulation
Bloodgood organized chess games within Powhatan Prison, which were by necessity with fellow inmates.[2] Many of these inmates were taught the game by Bloodgood, and thus began as unrated and inexperienced players. Bloodgood obtained USCF memberships for them. Some accused Bloodgood, with his intimate knowledge of the rating system, of rigging their ratings. The accusation was that he arranged for new prisoners to play rated games against other prisoners, who would deliberately lose, thus giving the new inmate an inflated USCF rating. Bloodgood, it is further alleged, then played rated games against the new highly-rated prisoner, and each time he won, gained a few more rating points. This continued for several years, and by 1996 his rating rose to 2702, making the 59-year-old Bloodgood the second-highest rated player in the nation. In comparison, at his retirement Bobby Fischer's rating was 2760, and several leading grandmasters were in the 2600s. Bloodgood's true strength at the time is not knowable but is likely to have been in the USCF Expert (2000-2200) range, though some have estimated that Bloodgood was of Senior Master strength (i.e., 2400+) when in his prime in the 1960s.[3]

This is all a matter of considerable controversy even today. Bloodgood himself vehemently denied these accusations, and said that he played chess in the only competitions available to him, prison tournaments, and won almost every game because he was the strongest player in the prison system. As his rating rose, he wrote the USCF to warn them that its system was prone to "closed pool" ratings inflation. However, nothing was done until Bloodgood's rating skyrocketed. He even qualified for entry into the U.S. Chess Championship, a prestigious invitation-only event intended for the best 16 players in the country. His high rating caused a crisis in the USCF, which debated extensively what to do about the situation. In the end, Bloodgood wasn't invited to the event (which he could not have attended anyway), and the USCF changed its ratings system rules to attempt to prevent "closed pool" ratings inflation."

I myself have some experience with this. In the 1990's under the old USCF system, various regions would inflate at different levels, and each year I would travel from Virginia with my 1700 OTB rating to the US Amateur Team East championship in New Jersey. I usually played board 1 on my team, and was paired against 2000-something rated experts from NY or NJ. I often won or drew, and returned to Virginia with a much higher rating than when I left.

I usually passed the points along months later to a "country player" from rural Virginia who was more underrated than I was!

Paul

Hi Orion.

"Yes, but in serious play one player does NOT win every game."

Of course but remember what I posted.

"A,B,C,D,E,F,G,H,I & K. with the instructions. "

The players are ordered or agree to lose.
Infact if everyone plays a Fools Mate it is possible for someone
who only knows how to move a pawn and a Queen to get a rating of 2400.

Hi Fabs and WW.

"I don't get anymore points of him, and lose 32 points if I lose agame with him."
(Fabs)

"above that point your frequency of wins doesn't cover the amount of points
a loss takes from you, "
(WW)

Again I tell you it's all a fix - that is why I set this scenario in a club.

Everyone agrees to let one player keep winning.
They think it will be good for their club/clan to have the No.1 player
on the site so come up with this plan.

So forget all the win/loss permutations. This is win win win.

I'm just trying to prove that the glorified and much worshipped numbers
mean absolutely sweet sod all.

Remember:

You are reading a post from someone who got a dog onto
the Scottish Grading System.

http://www.chessedinburgh.co.uk/chandlerarticle.php?ChandID=359

Hi Paul.

An excellent post (I take back my £5,000 brain gag - It's is being used).

A brilliant piece of investigated chess journalism.
(nit pick. You could have give it a quick edit to make it readable)

eg:

Bloodgood organized chess games within Powhatan Prison,
which were by necessity with fellow inmates.

Many of these inmates were taught the game by Bloodgood,
and thus began as unrated and inexperienced players.

Bloodgood obtained USCF memberships for them.

Some accused Bloodgood, with his intimate knowledge of the rating system,
of rigging their ratings..... etc etc.

I hate those wall to wall post. But 9/10 well done.

If you have not read it Bloodgood hit a grade of 2702 by doing
(allegedly) what I suggested.

The final sentence about passing on the high grade - (like the pox) - onto
rural players is brilliant.

By coincidence right after that piece on the Corner about the dog,
I do a piece about a player called Ken Bloodworthy. (here it's Bloodgood)

A good Corner that one.

Originally posted by greenpawn34
Hi Orion.

"Yes, but in serious play one player does NOT win every game."

Of course but remember what I posted.

"A,B,C,D,E,F,G,H,I & K. with the instructions. "

The players are ordered or agree to lose.
Infact if everyone plays a Fools Mate it is possible for someone
who only knows how to move a pawn and a Queen to get a rating of 2400.

...[text shortened]... a player called Ken Bloodworthy. (here it's Bloodgood)

A good Corner that one.

I bow to the pro.

As an aside, everyone in the South Mecklenburg State Penitentiary (where he was serving) ended up with inflated ratings ( if you are unrated, and you lose to a 2702 player, that's a one-game provisional of 2302, no matter how bad you suck), and all the other clubs in Virginia used to try to arrange club matches with the prison, because your players could pick up a large number of points.

Originally posted by greenpawn34


The players are ordered or agree to lose.

ah, okay, then you'll get the 'golden king' thing of course. it has nothing to do with the ratings of current super GMs though. nor inflation. just a basic scam.

Hi WW

Yes but it's the same basic scam the current top lot are pulling
by organising and playing in these 'closed shop' tournaments.

I'm now looking for twelve 1400 players to join Bates Motel
I'll make one of you famous.

Don't PM me, I'll PM you.

Yours

The newly crowned King of the Golden Scam. (The Bloodgood award).

Greenpawn 😏

Originally posted by greenpawn34
Hi WW

Yes but it's the same basic scam the current top lot are pulling
by organising and playing in these 'closed shop' tournaments.

I'm now looking for twelve 1400 players to join [b]Bates Motel
I'll make one of you famous.

Don't PM me, I'll PM you.

Yours

The newly crowned King of the Golden Scam. (The Bloodgood award).

Greenpawn 😏[/b]

the top guys play the 2750s all the time. the 2750s play the 2600s all the time. the 2600s play the 2500s all the time.

as a result, any possible rating discrepancies will propagate over the whole vertical rating scale very fast. the top GMs are not a separate closed rating pool.

Originally posted by orion25
If he is better his rating will rise, if he is not it won't - his rating will show his ability in comparison to the pool - is all I meant by that.

ok

Originally posted by greenpawn34
Hi WW

Yes but it's the same basic scam the current top lot are pulling
by organising and playing in these 'closed shop' tournaments.

I'm now looking for twelve 1400 players to join [b]Bates Motel
I'll make one of you famous.

Don't PM me, I'll PM you.

Yours

The newly crowned King of the Golden Scam. (The Bloodgood award).

Greenpawn 😏[/b]

It's not the same scam as, in reality, there is not one player who wins every game, instead there are draws and losses by which his rating will tend to a certain value (in this case, perhaps 2900), and plateau at it. This rating will reflect his ability in comparison to others.

You can claim though, since top players, basically only play themselves, that this rating will only truthfully compare the ratings within that selected group of players. In this case it might not be able to use this rating to compare one of these top-GM with, say, a 1600 player.

This prison dude Bloodgood who committed rating incest till he qualified for the US Championship is a perfect example of how the useless FIDE rating system now has people approaching 2900, next they will be approaching 3000.