Main
Date: 09 Jan 2008 13:56:28
From: raylopez99
Subject: Greatest upsets or unexpected results in chess tournament history
I was inspired by the age rating chart and Lasker's incredible
statistics found here: ( http://db.chessmetrics.com/CM2/AgeLists.asp )
and the fact that in 1918 at age 50, he had nearly the same Elo as
when he was at is peak younger. Then I wanted to see why he didn't
play better against Capa (who in 1920 was at the top of his game), and
fortunately I stumbled upon Arnad Elo's book, which has a section on
"Expected Scores vs Actual Scores", section 5.55 of the original
edition of Elo's book, which shows, statitistically (laying aside the
fact that Elo is slightly skewed when there is a large rating
difference between players, but at the top level I don't think that
matters that much), which contests are the greatest mismatches and
upsets.

Here they are: (Higher rated player first, and negative number meaning
he could have scored even better)

Greatest mismatches/ upsets in chess history below. Where the results
were clearly statistically significant in that they are more than one
standard deviation, I indicate with an asterisk (*).

1866 Steinitz-Bird (-1.2) (that is, Steinitz should have scored 1.2
games more than he actually did)

1866 Steinitz-Anderssen (1.0) (Anderssen should have scored better, an
extra game better, while Steinitz overachieved by one extra game)

1872 Steinitz-Zukertort (1.3) (this match at least shows Steinitz did
better than he should have, or, conversely, Zukertort did worse than
expected).

1878 Chigorin-Schiffers (-1.3) (Chigorin did worse than expected)

1879-80 Chigorin-Schiffers (2.2 Chigorin lucks out; Schiffers chokes)

1882-83 Steinitz-tinez (3.5!! tinez chokes, and disappears from
chess history, though he was a stronger player than results show--who
is this guy? never heard of him)*(this result is statistically
significant, as sigma was 2.2 games)

1886 Burn-Bird (-2.0 Burn chokes, or Bird does better than expected)

1892 Lipscheutz-Showalter (2.1; Showalter chokes)

1895 Janowsky-Mieses (-2.0; Janowsky chokes)(*statistically
significant, as sigma was 1.8)

1897-98 Pillsbury-Showalter (-2.9; Showalter makes up for his poor
1892 performance and does better than expected against Pillsbury)*
(sigma was 2.7)

1905 Tarrasch-shall (2.3; Tarrasch does better than expected)

1916 Janowski-Showalter (1.3; Showalter, the overachiever, meets
Janowsky, the choke artist, and Janowski surprisingly does better than
expected)

1921 Capablanca-Lasker 1.4

1935 Alekhine-Euwe -2.6 (Alekhine chokes, but sigma was 2.7 so this is
not quite statistically signficant, or more than one sigma away from
the mean)

1937 Botvinnik-Levenfish (-2.1, Levenfish does well, better than
expected)* sigma =1.7

(for some reason, in the modern era not so many mismatches or upsets--
perhaps they train better?)

1977 Korchnoi-Polugaevsky (1.6; almost statistically significant, as
sigma=1.8)

1981 Karpov-Korchnoi (1.8; Korchnoi chokes, but sigma= 2.1)

1985 Karpov-Kasparov (-1.7, Karpov chokes, this was not the halted
1985 match but later, but this is not statistically significant since
sigma = 2.4)

RL




 
Date: 14 Jan 2008 02:36:42
From: help bot
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 10, 2:29 pm, raylopez99 <[email protected] > wrote:
> On Jan 10, 6:35 am, Taylor Kingston <[email protected]> wrote:
>
> > Ray, you said "I wanted to see why [Lasker] didn't play better
> > against Capa..." I don't see how statistics can answer such a
> > question. It can only be answered in terms of physical and
> > psychological factors affecting the players involved. AFAIK, Elo-type
> > statistics can only be used to sumize the players' results up to
> > 1921, which become a basis for a mathematical prediction. They don't
> > explain the real "why."
>
> Yes, but statistics are the basis for how atoms and molecules work

But what if they are retired atoms, or unemployed
molecules?


> Taylor, and indeed have a say in free will. If everything was
> deterministic, then there would be no free will. The fact that at any
> time we can fall into a black hole (Hume's scepticism) is the basis
> for free will, the "why", as you say.

So, you are saying that if Mr. Hume had not come
up with this theory, there would be no basis whatever
for free will? Hmmm.


> > > Smyslov, who almost beat Kasparov in the
> > > 1980s during a Candidate match,
>
> > "Almost beat"? Kasparov won +4 -0 =9.
>
> I was thinking that Smyslov played well for his age, and as per the
> statistics he would have won +6 - 0 = 7

You even got this wrong. The guy with four
wins was Gary Kasparov, so had two of their
draws been won instead by GM Smyslov, the
score would then become +4 -2 =7, with GK
up by two games.

Face it: you are punch-drunk; if my hunch
is correct, there is a 10% chance that a
famous boxer has bitten off one or more of
your ears, then landed a solid blow which
has rendered your brain permanently impaired.
The other 80% chance is that you have been
drinking and posting. (I deliberately messed
up the math here, so as not to confuse you
even more.)


-- help bot





 
Date: 14 Jan 2008 01:48:29
From: help bot
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 9, 5:43 pm, Taylor Kingston <[email protected] > wrote:
> On Jan 9, 4:56 pm, raylopez99 <[email protected]> wrote:
>
> > I was inspired by the age rating chart and Lasker's incredible
> > statistics found here: (http://db.chessmetrics.com/CM2/AgeLists.asp)
> > and the fact that in 1918 at age 50, he had nearly the same Elo as
> > when he was at is peak younger. Then I wanted to see why he didn't
> > play better against Capa (who in 1920 was at the top of his game),
>
> The reasons for Lasker's debacle in his 1921 match with Capablanca
> are well known. The most important include:
>
> 1) Lasker was in very poor form, depressed by by Germany's defeat in
> WWI and the hardships the war had inflicted. He had no competitive
> fire. He really didn't want to play the match at all all, in fact he
> resigned the title to Capablanca beforehand, and withdrew from pre-
> match negotiations. But then Havana came through with financial terms
> he could not afford to refuse. He played mainly for the money, with
> great reluctance and little interest.
> 2) The heat in Havana was hard on Lasker. No air-conditioning back
> then. Capablanca, a native Cuban, was unfazed.
> 3) Capablanca was a very good chess player.
>
> Lasker would probably have stayed away from chess indefinitely after
> that, but Germany's hyper-inflation in the early 1920s ate up his
> savings. So he had to go "back to work," which he did quite
> successfully, winning at Maehrisch-Ostrau 1923, and of course at New
> York 1924, ahead of Capablanca. Capa did not finish ahead of him in
> tournament play until 1936.

Anyone who is familiar with writings on chess
from this and earlier times will know that the
poor health excuse was the most common of
all excuses, bar none.

I vaguely recall an effort by one pro-Capa writer
to counter some of what has been written by Em.
Lasker's apologists; for instance, that while Cuba
may be hot, those games were played at night,
etc. Typically, books on JC put things in a bad
light for EL, while books on EL do precisely the
opposite; I think it's unwise to swallow everything
such hack writers might spew, without very careful
consideration.

It is a historical fact that the tactic of ducking out
of a chess match in progress by requesting a time
out, and then getting embroiled in a dispute and
quitting, was far from unknown. In Paul Morphy's
day, this might even be said to be a standard
operating procedure, and claiming poor health
rather than poor chess skill, its main competition!

One question which might shed light on on this
is: just how poorly did EL play in this match? Is
it an established fact that EL played well below
his usual standard, or is it possible that JC
raised his level of play to new heights?

The lack of objectivity is key; for instance, as
TK would have it, poor EL lost his "savings" due
to hyper-inflation; this makes one wonder: what
"savings", and which job gave him the extra
income to "save" up? (Books I've read state
clearly that EL ried into money.) Distortion
of facts -- or pawns -- is best avoided whenever
possible, if one wishes to succeed.


-- help bot




 
Date: 13 Jan 2008 14:32:02
From: Taylor Kingston
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 13, 5:03=A0pm, raylopez99 <[email protected] > wrote:
> On Jan 11, 9:51=A0am, Taylor Kingston <[email protected]> wrote:
>
>
>
> > =A0 If ifs and buts were candy and nuts, what a merry Christmas we would=

> > have. Or, as "Dreamer" Tatum said in the novel Semi-Tough: "Whatever
> > coulda happened, did."
>
> Is that a good book? =A0Sounds like Studs Terkel wrote it.
>
> > =A0 In any event, Lasker losing the title to Capablanca was virtually
> > inevitable. The older champion always must give way to a younger man
> > eventually.
>
> No, that's what makes this match interesting. =A0Lasker at 50 was about
> the same as Lasker at 25. =A0With a little luck, about 33% of the time
> he would have (in an alternate universe) beat Capa.

"In an alternate universe" are the operative words. I have enough
trouble attaining a small understanding of this universe. I leave
"alternate universes" to sci-fi writers.

> > > =A0That's
> > > pretty interesting to me.
>
> > =A0 I find such statistical convolutions only slightly more interesting
> > than a phone book or tax form, and much less useful. In contrast, I
> > find the interplay of historical, environmental, financial and
> > psychological factors quite interesting, besides their being the main
> > reasons the match went as it did.
>
> Ah, you are a determinist--

That is an unwarranted conclusion. I have expressed no opinion on
the clich=E9d issue of "determinism vs. free will."

> cause and effect. =A0World doesn't always
> work like that,

Really? At a quantum level, perhaps yes, but at our macro level,
deterministic Newtonian physics have a very good track record.

> because if it did we'd all be robots. =A0

One must make a distinction here between inanimate objects such as
billiard balls, and complex living human beings.

> Chessically,
> along this line, there's even a chess book by Saidy or somebody that
> basically shows how a number of games where one side had a "horrible"
> looking position, positionally, won the game anyway! =A0

Yep, won games are lost all the time. Done it myself, more often
than I care to remember. This is news of the "dog bites man" variety.

> Steiniz also
> played like this (anti-positionally). =A0Good defense. =A0Some books by
> the Australian master Purdy also emphasize this theme--looking at the
> board afresh.

Your comments seem confused. Anti-positional =3D looking afresh? The
connection is unclear.

> =A0And if you've ever played a "lost" game through Fritz,
> autoannotating, towards the end of the game sometimes the computer
> finds drawing (or even winning) chances for the side that lost, when
> all seemed doomed.

"Dog bites man" again.


  
Date: 13 Jan 2008 23:55:54
From: David Richerby
Subject: Re: Greatest upsets or unexpected results in chess tournament history
Taylor Kingston <[email protected] > wrote:
> "Dog bites man" again.

Somebody really ought to do something about that damned dog.


Dave.

--
David Richerby Evil Pointy-Haired Widget (TM):
www.chiark.greenend.org.uk/~davidr/ it's like a thingy that's completely
clueless but it's genuinely evil!


 
Date: 13 Jan 2008 14:03:28
From: raylopez99
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 11, 9:51=A0am, Taylor Kingston <[email protected] > wrote:
>
> =A0 If ifs and buts were candy and nuts, what a merry Christmas we would
> have. Or, as "Dreamer" Tatum said in the novel Semi-Tough: "Whatever
> coulda happened, did."

Is that a good book? Sounds like Studs Terkel wrote it.

> =A0 In any event, Lasker losing the title to Capablanca was virtually
> inevitable. The older champion always must give way to a younger man
> eventually.
>

No, that's what makes this match interesting. Lasker at 50 was about
the same as Lasker at 25. With a little luck, about 33% of the time
he would have (in an alternate universe) beat Capa.

> > =A0That's
> > pretty interesting to me.
>
> =A0 I find such statistical convolutions only slightly more interesting
> than a phone book or tax form, and much less useful. In contrast, I
> find the interplay of historical, environmental, financial and
> psychological factors quite interesting, besides their being the main
> reasons the match went as it did.

Ah, you are a determinist--cause and effect. World doesn't always
work like that, because if it did we'd all be robots. Chessically,
along this line, there's even a chess book by Saidy or somebody that
basically shows how a number of games where one side had a "horrible"
looking position, positionally, won the game anyway! Steiniz also
played like this (anti-positionally). Good defense. Some books by
the Australian master Purdy also emphasize this theme--looking at the
board afresh. And if you've ever played a "lost" game through Fritz,
autoannotating, towards the end of the game sometimes the computer
finds drawing (or even winning) chances for the side that lost, when
all seemed doomed.

Bye,

RL


 
Date: 11 Jan 2008 13:44:47
From: roadkill
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 10, 10:56 am, raylopez99 <[email protected] > wrote:
> I was inspired by the age rating chart and Lasker's incredible
> statistics found here: (http://db.chessmetrics.com/CM2/AgeLists.asp)
> and the fact that in 1918 at age 50, he had nearly the same Elo as
> when he was at is peak younger. Then I wanted to see why he didn't
> play better against Capa (who in 1920 was at the top of his game), and
> fortunately I stumbled upon Arnad Elo's book, which has a section on
> "Expected Scores vs Actual Scores", section 5.55 of the original
> edition of Elo's book, which shows, statitistically (laying aside the
> fact that Elo is slightly skewed when there is a large rating
> difference between players, but at the top level I don't think that
> matters that much), which contests are the greatest mismatches and
> upsets.
>
> Here they are: (Higher rated player first, and negative number meaning
> he could have scored even better)
>
> Greatest mismatches/ upsets in chess history below. Where the results
> were clearly statistically significant in that they are more than one
> standard deviation, I indicate with an asterisk (*).
>
> 1866 Steinitz-Bird (-1.2) (that is, Steinitz should have scored 1.2
> games more than he actually did)
>
> 1866 Steinitz-Anderssen (1.0) (Anderssen should have scored better, an
> extra game better, while Steinitz overachieved by one extra game)
>
> 1872 Steinitz-Zukertort (1.3) (this match at least shows Steinitz did
> better than he should have, or, conversely, Zukertort did worse than
> expected).
>
> 1878 Chigorin-Schiffers (-1.3) (Chigorin did worse than expected)
>
> 1879-80 Chigorin-Schiffers (2.2 Chigorin lucks out; Schiffers chokes)
>
> 1882-83 Steinitz-tinez (3.5!! tinez chokes, and disappears from
> chess history, though he was a stronger player than results show--who
> is this guy? never heard of him)*(this result is statistically
> significant, as sigma was 2.2 games)
>
> 1886 Burn-Bird (-2.0 Burn chokes, or Bird does better than expected)
>
> 1892 Lipscheutz-Showalter (2.1; Showalter chokes)
>
> 1895 Janowsky-Mieses (-2.0; Janowsky chokes)(*statistically
> significant, as sigma was 1.8)
>
> 1897-98 Pillsbury-Showalter (-2.9; Showalter makes up for his poor
> 1892 performance and does better than expected against Pillsbury)*
> (sigma was 2.7)
>
> 1905 Tarrasch-shall (2.3; Tarrasch does better than expected)
>
> 1916 Janowski-Showalter (1.3; Showalter, the overachiever, meets
> Janowsky, the choke artist, and Janowski surprisingly does better than
> expected)
>
> 1921 Capablanca-Lasker 1.4
>
> 1935 Alekhine-Euwe -2.6 (Alekhine chokes, but sigma was 2.7 so this is
> not quite statistically signficant, or more than one sigma away from
> the mean)
>
> 1937 Botvinnik-Levenfish (-2.1, Levenfish does well, better than
> expected)* sigma =1.7
>
> (for some reason, in the modern era not so many mismatches or upsets--
> perhaps they train better?)
>
> 1977 Korchnoi-Polugaevsky (1.6; almost statistically significant, as
> sigma=1.8)
>
> 1981 Karpov-Korchnoi (1.8; Korchnoi chokes, but sigma= 2.1)
>
> 1985 Karpov-Kasparov (-1.7, Karpov chokes, this was not the halted
> 1985 match but later, but this is not statistically significant since
> sigma = 2.4)
>
> RL


You need to get out more.



 
Date: 11 Jan 2008 09:51:36
From: Taylor Kingston
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 11, 12:22=A0pm, raylopez99 <[email protected] > wrote:
> On Jan 10, 4:06=A0pm, Taylor Kingston <[email protected]> wrote:
>
> > > So, in our counterfactual, Lasker is indeed capable of remaining world=

> > > champion. =A0That's what statistics tell us.
>
> > =A0 History tells us Capa won +4 -0 =3D10.
>
> But counterfactuals are by definition not what actually happened; they
> are relevant however in answering 'why'?

So far in this discussion your statistics have not produced a single
causative factor. They seem to be quite useless in any discussion of
"Why?".

> > =A0 In other words, if events had transpired in a very different way,
> > Lasker would have won. Yes, and if John Kerry had gotten more votes in
> > 2004, he'd be President now. This seems rather tautological, not a
> > fruitful line of inquiry.
>
> Historian Niels Furgenson, the father of counterfactual history, would
> disagree. =A0And if Kerry had won likely the USA would have a smaller
> committment in Iraq (Kerry was not against the war BTW).
>
> > =A0 I'm not trying to be overly argumentative here, Ray; I just don't
> > see much point in these speculations.
>
> It's interesting IMO. Consider than it any 24 game match, depending
> on the strength of the players, there's always a statistical chance of
> the weaker side winning. =A0My quick eyeballing of the figures,
> figureing Capa was 50-75 Elo points greater than Lasker, shows that
> Lasker could have won the 24 game match, if he put his heart into it
> (he did not) about roughly 30% of the time or slightly less.

If ifs and buts were candy and nuts, what a merry Christmas we would
have. Or, as "Dreamer" Tatum said in the novel Semi-Tough: "Whatever
coulda happened, did."
In any event, Lasker losing the title to Capablanca was virtually
inevitable. The older champion always must give way to a younger man
eventually.

> =A0That's
> pretty interesting to me.

I find such statistical convolutions only slightly more interesting
than a phone book or tax form, and much less useful. In contrast, I
find the interplay of historical, environmental, financial and
psychological factors quite interesting, besides their being the main
reasons the match went as it did.


 
Date: 11 Jan 2008 09:22:25
From: raylopez99
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 10, 4:06=A0pm, Taylor Kingston <[email protected] > wrote:

>
> > So, in our counterfactual, Lasker is indeed capable of remaining world
> > champion. =A0That's what statistics tell us.
>
> =A0 History tells us Capa won +4 -0 =3D10.
>

But counterfactuals are by definition not what actually happened; they
are relevant however in answering 'why'?


>
> =A0 In other words, if events had transpired in a very different way,
> Lasker would have won. Yes, and if John Kerry had gotten more votes in
> 2004, he'd be President now. This seems rather tautological, not a
> fruitful line of inquiry.

Historian Niels Furgenson, the father of counterfactual history, would
disagree. And if Kerry had won likely the USA would have a smaller
committment in Iraq (Kerry was not against the war BTW).


> =A0 I'm not trying to be overly argumentative here, Ray; I just don't
> see much point in these speculations.

It's interesting IMO. Consider than it any 24 game match, depending
on the strength of the players, there's always a statistical chance of
the weaker side winning. My quick eyeballing of the figures,
figureing Capa was 50-75 Elo points greater than Lasker, shows that
Lasker could have won the 24 game match, if he put his heart into it
(he did not) about roughly 30% of the time or slightly less. That's
pretty interesting to me.

Goodbye

RL


 
Date: 10 Jan 2008 16:06:55
From: Taylor Kingston
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 10, 5:46=A0pm, raylopez99 <[email protected] > wrote:
> On Jan 10, 11:51=A0am, Taylor Kingston <[email protected]> wrote:
>
> > =A0 Ray, I'm not discussing free will at all; that's an issue for
> > philosophy, not chess history.
>
> But chess is life.

Hmmmm. Then non-chessplayers are dead?

> > You asked "why didn't Lasker play
> > better against Capa," so I gave you the reasons I know of from reading
> > chess history. Add what Wlod pointed out, and I think you have all the
> > important reasons, without bringing statistics into it at all.
>
> But it's incomplete. =A0See here for a fuller discussion:http://www.chessg=
ames.com/perl/chess.pl?tid=3D54111
>
> Note that technically Lasker didn't lose his championship; he resigned
> it, like Wlod says. =A0He did it for the money, and the heat killed him
> (as he pointed out 10 years earlier), plus Capa was a very good
> player.

Yes, all of which I pointed out several posts ago.

> But, all things being equal, Lasker should have won 1.4 more games.

But all things were not equal. They seldom are.

> That's an extra win and a draw. =A0So referring to this link:http://www.ch=
essgames.com/perl/chess.pl?tid=3D54111, note the first four
> games were draws. =A0If Lasker was playing to (statistical) form, he
> *could* have won the fifth game, instead of Capa. =A0Now it would be +1
> for Lasker. =A0Games 6,7,8 and 9 were draws, but, Lasker *could* have
> drew instead of losing the crucial 10th game. =A0This is as far as
> statistics would take Lasker, since Capa was better at that time. =A0But
> let's think about this: =A0after 10 games, the score would have been in
> Lasker's favor: =A0+1 - 0 =3D9. =A0At this point, Lasker *could* have used=

> his superior and legendary "psychological skills" to ask for a change
> in venue, to a cooler clime. =A0Using the equivalent of Campomanes
> (http://en.wikipedia.org/wiki/Florencio_Campomanes) in Fide at that
> time (who that might be I have no idea, since Fide was not around),
> and analogizing to Karpov's first match with Kasparov, Lasker *could*
> have maybe postponed the remainder of the match (recall Lasker
> resigned the match after game 14, with 10 more to go).
>
> But even after 13 games (Capa won game 11), in our counterfactual, the
> score would have been +1 -1 =3D11, so the match would be evenly tied.
> And remember, game 14 saw Lasker in undoubtably ill health--assuming
> that in our counterfactual Lasker asked for a change in venue before
> game 14, it's conceivable that the series would have been tied when it
> moved to a cooler climate, like New York.
>
> Then--and this is key--if Lasker played *above* his potential in games
> 14-24 in our counterfactual, in a cooler clime, he could have "psyched
> out" Capa and won.
>
> So, in our counterfactual, Lasker is indeed capable of remaining world
> champion. =A0That's what statistics tell us.

History tells us Capa won +4 -0 =3D10.

> =A0Lasker beating Capa would
> have been no more than one standard deviation from the mean, which
> happens about 33% of the time, so it's doable. =A0That's the power of
> statistics Taylor.

In other words, if events had transpired in a very different way,
Lasker would have won. Yes, and if John Kerry had gotten more votes in
2004, he'd be President now. This seems rather tautological, not a
fruitful line of inquiry.

> > > > > Smyslov, who almost beat Kasparov in the
> > > > > 1980s during a Candidate match,
>
> > > > =A0 "Almost beat"? Kasparov won +4 -0 =3D9.
>
> > > I was thinking that Smyslov played well for his age, and as per the
> > > statistics he would have won +6 - 0 =3D 7
>
> > =A0 No question that Smyslov did well for his age against GK in 1984,
> > but there's quite a gap between "playing well for his age" and "almost
> > beating."
>
> Not age adjusted there isn't.

Last I checked, FIDE did not make any "age adjustment" in match
scores.

I'm not trying to be overly argumentative here, Ray; I just don't
see much point in these speculations.




 
Date: 10 Jan 2008 14:46:02
From: raylopez99
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 10, 11:51=A0am, Taylor Kingston <[email protected] > wrote:
> =A0 Ray, I'm not discussing free will at all; that's an issue for
> philosophy, not chess history.

But chess is life.

> You asked "why didn't Lasker play
> better against Capa," so I gave you the reasons I know of from reading
> chess history. Add what Wlod pointed out, and I think you have all the
> important reasons, without bringing statistics into it at all.

But it's incomplete. See here for a fuller discussion:
http://www.chessgames.com/perl/chess.pl?tid=3D54111

Note that technically Lasker didn't lose his championship; he resigned
it, like Wlod says. He did it for the money, and the heat killed him
(as he pointed out 10 years earlier), plus Capa was a very good
player.

But, all things being equal, Lasker should have won 1.4 more games.
That's an extra win and a draw. So referring to this link:
http://www.chessgames.com/perl/chess.pl?tid=3D54111, note the first four
games were draws. If Lasker was playing to (statistical) form, he
*could* have won the fifth game, instead of Capa. Now it would be +1
for Lasker. Games 6,7,8 and 9 were draws, but, Lasker *could* have
drew instead of losing the crucial 10th game. This is as far as
statistics would take Lasker, since Capa was better at that time. But
let's think about this: after 10 games, the score would have been in
Lasker's favor: +1 - 0 =3D9. At this point, Lasker *could* have used
his superior and legendary "psychological skills" to ask for a change
in venue, to a cooler clime. Using the equivalent of Campomanes
(http://en.wikipedia.org/wiki/Florencio_Campomanes) in Fide at that
time (who that might be I have no idea, since Fide was not around),
and analogizing to Karpov's first match with Kasparov, Lasker *could*
have maybe postponed the remainder of the match (recall Lasker
resigned the match after game 14, with 10 more to go).

But even after 13 games (Capa won game 11), in our counterfactual, the
score would have been +1 -1 =3D11, so the match would be evenly tied.
And remember, game 14 saw Lasker in undoubtably ill health--assuming
that in our counterfactual Lasker asked for a change in venue before
game 14, it's conceivable that the series would have been tied when it
moved to a cooler climate, like New York.

Then--and this is key--if Lasker played *above* his potential in games
14-24 in our counterfactual, in a cooler clime, he could have "psyched
out" Capa and won.

So, in our counterfactual, Lasker is indeed capable of remaining world
champion. That's what statistics tell us. Lasker beating Capa would
have been no more than one standard deviation from the mean, which
happens about 33% of the time, so it's doable. That's the power of
statistics Taylor.

>
> > > > Smyslov, who almost beat Kasparov in the
> > > > 1980s during a Candidate match,
>
> > > =A0 "Almost beat"? Kasparov won +4 -0 =3D9.
>
> > I was thinking that Smyslov played well for his age, and as per the
> > statistics he would have won +6 - 0 =3D 7
>
> =A0 No question that Smyslov did well for his age against GK in 1984,
> but there's quite a gap between "playing well for his age" and "almost
> beating."

Not age adjusted there isn't.

RL



 
Date: 10 Jan 2008 11:51:38
From: Taylor Kingston
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 10, 2:29=A0pm, raylopez99 <[email protected] > wrote:
> On Jan 10, 6:35=A0am, Taylor Kingston <[email protected]> wrote:
>
> > =A0 Ray, you said "I wanted to see why [Lasker] didn't play better
> > against Capa..." I don't see how statistics can answer such a
> > question. It can only be answered in terms of physical and
> > psychological factors affecting the players involved. AFAIK, Elo-type
> > statistics can only be used to sumize the players' results up to
> > 1921, which become a basis for a mathematical prediction. They don't
> > explain the real "why."
>
> Yes, but statistics are the basis for how atoms and molecules work
> Taylor, and indeed have a say in free will. =A0If everything was
> deterministic, then there would be no free will. =A0The fact that at any
> time we can fall into a black hole (Hume's scepticism) is the basis
> for free will, the "why", as you say.

Ray, I'm not discussing free will at all; that's an issue for
philosophy, not chess history. You asked "why didn't Lasker play
better against Capa," so I gave you the reasons I know of from reading
chess history. Add what Wlod pointed out, and I think you have all the
important reasons, without bringing statistics into it at all.

> > > Smyslov, who almost beat Kasparov in the
> > > 1980s during a Candidate match,
>
> > =A0 "Almost beat"? Kasparov won +4 -0 =3D9.
>
> I was thinking that Smyslov played well for his age, and as per the
> statistics he would have won +6 - 0 =3D 7

No question that Smyslov did well for his age against GK in 1984,
but there's quite a gap between "playing well for his age" and "almost
beating."


 
Date: 10 Jan 2008 11:29:25
From: raylopez99
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 10, 6:35=A0am, Taylor Kingston <[email protected] > wrote:

> =A0 Ray, you said "I wanted to see why [Lasker] didn't play better
> against Capa..." I don't see how statistics can answer such a
> question. It can only be answered in terms of physical and
> psychological factors affecting the players involved. AFAIK, Elo-type
> statistics can only be used to sumize the players' results up to
> 1921, which become a basis for a mathematical prediction. They don't
> explain the real "why."

Yes, but statistics are the basis for how atoms and molecules work
Taylor, and indeed have a say in free will. If everything was
deterministic, then there would be no free will. The fact that at any
time we can fall into a black hole (Hume's scepticism) is the basis
for free will, the "why", as you say.

>
> > Smyslov, who almost beat Kasparov in the
> > 1980s during a Candidate match,
>
> =A0 "Almost beat"? Kasparov won +4 -0 =3D9.

I was thinking that Smyslov played well for his age, and as per the
statistics he would have won +6 - 0 =3D 7

RL


 
Date: 10 Jan 2008 06:35:23
From: Taylor Kingston
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 10, 3:46=A0am, raylopez99 <[email protected] > wrote:

> Thanks to Taylor Kingston and Wlod for this "historian's" view of
> things. =A0I was speaking as a statistician.

Ray, you said "I wanted to see why [Lasker] didn't play better
against Capa..." I don't see how statistics can answer such a
question. It can only be answered in terms of physical and
psychological factors affecting the players involved. AFAIK, Elo-type
statistics can only be used to sumize the players' results up to
1921, which become a basis for a mathematical prediction. They don't
explain the real "why."

> Smyslov, who almost beat Kasparov in the
> 1980s during a Candidate match,

"Almost beat"? Kasparov won +4 -0 =3D9.



 
Date: 10 Jan 2008 00:46:47
From: raylopez99
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 9, 7:32=A0pm, "Wlodzimierz Holsztynski (Wlod)"
<[email protected] > wrote:
> You're listing only upsets in matches.
> There were also famous tournament upsets.
>
> Some were on the natural causes. In particular the
> Akiba Rubinstein's poor result
> in Petersburg, 1914 tournament.
>
> In the case od Leonid Stein, it was politics
> and arbitrary rulings.
>
> =A0 =A0 Wlod

Thanks to Taylor Kingston and Wlod for this "historian's" view of
things. I was speaking as a statistician.

I did notice in the modern era (post 1920s) there seems to be fewer
'upsets', but I also missed one in my original list:

1957 Botvinnik-Smyslov -2.0 (Smyslov, who almost beat Kasparov in the
1980s during a Candidate match, did better than expected, but it's not
more than 1 sigma from the mean, which was 2.4 games, so it's
technically not statistically significant).

RL


 
Date: 09 Jan 2008 19:32:28
From: Wlodzimierz Holsztynski (Wlod)
Subject: Re: Greatest upsets or unexpected results in chess tournament history
You're listing only upsets in matches.
There were also famous tournament upsets.

Some were on the natural causes. In particular the
Akiba Rubinstein's poor result
in Petersburg, 1914 tournament.

In the case od Leonid Stein, it was politics
and arbitrary rulings.

Wlod


 
Date: 09 Jan 2008 19:29:21
From: Wlodzimierz Holsztynski (Wlod)
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 9, 2:43 pm, Taylor Kingston <[email protected] > wrote:
> On Jan 9, 4:56 pm, raylopez99 <[email protected]> wrote:
>

> The reasons for Lasker's debacle in his 1921 match with Capablanca
> are well known. The most important include:

0) Lasker's health was failing him during the match. It' s not a mere
excuse. Lasker was hospitalized for months after the match.
(He asked Capa for a break from playing. Cap refused, and Lasker
resigned).

In general, it's amazing that Lasker, despite his fragile
health could play with incredible energy and endurance.
Tal was like that too. Nevertheless, Tal was trying to
finish his opponents fast, even when he was young,
to save himself for the next game or in general (Saidy(?)
wrote this about Tal). Tal was making his moves fast,
and was aiming at short games, winning fast in combinational
style.

> 1) Lasker was in very poor form, depressed by by Germany's defeat in
> WWI and the hardships the war had inflicted. He had no competitive
> fire. He really didn't want to play the match at all all, in fact he
> resigned the title to Capablanca beforehand, and withdrew from pre-
> match negotiations. But then Havana came through with financial terms
> he could not afford to refuse. He played mainly for the money, with
> great reluctance and little interest.
> 2) The heat in Havana was hard on Lasker. No air-conditioning back
> then. Capablanca, a native Cuban, was unfazed.
> 3) Capablanca was a very good chess player.

Regards,

Wlod


 
Date: 09 Jan 2008 14:43:26
From: Taylor Kingston
Subject: Re: Greatest upsets or unexpected results in chess tournament history
On Jan 9, 4:56=A0pm, raylopez99 <[email protected] > wrote:
> I was inspired by the age rating chart and Lasker's incredible
> statistics found here: (http://db.chessmetrics.com/CM2/AgeLists.asp)
> and the fact that in 1918 at age 50, he had nearly the same Elo as
> when he was at is peak younger. =A0Then I wanted to see why he didn't
> play better against Capa (who in 1920 was at the top of his game),

The reasons for Lasker's debacle in his 1921 match with Capablanca
are well known. The most important include:

1) Lasker was in very poor form, depressed by by Germany's defeat in
WWI and the hardships the war had inflicted. He had no competitive
fire. He really didn't want to play the match at all all, in fact he
resigned the title to Capablanca beforehand, and withdrew from pre-
match negotiations. But then Havana came through with financial terms
he could not afford to refuse. He played mainly for the money, with
great reluctance and little interest.
2) The heat in Havana was hard on Lasker. No air-conditioning back
then. Capablanca, a native Cuban, was unfazed.
3) Capablanca was a very good chess player.

Lasker would probably have stayed away from chess indefinitely after
that, but Germany's hyper-inflation in the early 1920s ate up his
savings. So he had to go "back to work," which he did quite
successfully, winning at Maehrisch-Ostrau 1923, and of course at New
York 1924, ahead of Capablanca. Capa did not finish ahead of him in
tournament play until 1936.