Re: Who are the Greats?

Carlsbad 1913, I think. And it was Nunn. Richard Forster had a
response at Chess Cafe debunking much, if not all, of Nunn's
conclusions. For example, Nunn bashed the tournament book author
(Vidmar?) for missing things, and then went on to discuss how he used
Fritz to analyze the games. Forster commented "Why didn't Vidmar think
of that?"

Didn't a current GM (I think it might have been Nunn) make such a
systematic survey, but at a tournament level, taking one famous
top-class event early in the twentieth century and running a Fritz
blunder check on all the games,  then comparing the ratio of blunders
to games with the results of the same operation on a contemporary
event ?

It should be possible to perform a similar operation on the mature
Rubenstein or most other well-known players using one of huge
databases you get with Fritz or Aquarium.  In fact, you could rank
players by the ratio of blunders per game.  

For our purposes, how would you define "tactical error" -- one that
drops 2 pawns in the evaluation?  Or some higher number?  

Some factors that might screw up the results:  a player might blunder
more in weak events, or in games subsequent to a well-played loss.
Some players might get more accurate in "money" games.  And, if the
blunder ratio increases with age, you'd have to define a cut-off point
so as not to unfairly penalize some players for simply living longer
and competing while doing it.

 He certainly had great success circa 1922-1928, though he was rather
erratic.

 Rubinstein seems a strange case to me. Certainly he played great
games, won major tournaments, and had a deep understanding of chess.
Still, in playing over his games, I have found an unusually high
number of tactical errors that somehow went unpunished by his
opponents, and unnoticed by later annotators.
 Admittedly this was an unsystematic survey, but my impression was
that his frequency of error was signficantly higher than for other
comparable GMs; he just somehow lucked out and nobody noticed. A
careful, extensive, time-consuming comparative analysis of his and
others' games would be required to determine if this impression was
based on fact or just a skewed, too-small sample.

Bogo was at one point, according to chessmetrics, world number one:
http://db.chessmetrics.com/CM2/PlayerProfile.asp?Params=199510SSSSS3S01387900000
0111000000000000010100
I can't believe I left out Morphy. Also Fine - he must definitely go
in although he is a long way from being my favourite player.
Bronstein and Geller also. Carlsen and that old git Topalov; Ivanchuk.
Out go Short, Mason, Junge, Schlechter, Marshall, Larsen, Zukertort.
I would not include Khalifman, Ponomariov and Kasimdzhanov.
In MGP 1 Kasparov rated Rubinstein very highly.

Who are the 30 best chess players of all time?
f.ck knows.
I would have the World Champions plus:
Reshevsky
Larsen
Rubinstein
Schlechter
Short
Chigorin
Zuckertort
Mason
Pillsbury
Marshall
Junge
Keres
Kortschnoi
Tarrasch
and
Bogoljuboff