Were old masters really worse than present masters?

Nimzoditch
Nimzoditch
Apr 11, 2015, 8:01 AM |
4

Have blogs replaced starting a forum for Premium members at Chess.com?  If so, I guess I will stick to blogging.

I will blog later today on a computer analysis I did to shed light on GM John Nunn's assertion that the masters of old made more blunders than today's masters.  This is asserted in an article published around 1999 and he uses Fritz Blundercheck to compare two tournaments, Karlsbad 1911 (325 games) with Biel 1993 Interzonal (469 games).  I show that with updated hardware this conclusion still holds true, but with a 'lesser quality' tournament (one with less presige than an interzonal--I used Aeroflot 2004 A group and Hastings 2009, in particular Hastings 2009 matched well with the blunder rate of Karlsbad 1911) the modern masters in these lesser tournaments blundered at about the same rate as the Karlsbad 1911 masters.  However we need more data to be collected.

I will also publish not just the results but also the protocol, and invite readers to also run their own experiments, and publish to this blog post as a thread.  It takes about 2 to 20+ hours to analyze each tournament, depending on what settings you pick (longer time is obviously more accurate, but surprisingly even with a few seconds a position, a fast PC can often spot obvious blunders and very often these blunders really are blunders rather than deep positional sacrifices).  It would be great if we had more users running these experiments to collect data, and at longer time controls, so we can arrive at more robust statistical conclusions.  BTW far from being discouraged by Nunn's theory, I found out, replaying the old master games, that they were better than I thought after reading Nunn's article, though even they blundered often.

If this thread grows perhaps later we can even correlate Elo with blunder rate, and perhaps figure out the Elo of past masters and compare them, as a group, to today's masters.  My pet theory is that there is some rating inflation; I speculate about 100 points.

If we had a forum it would be better than a blog but I'll try this format.

Keywords: old GM elo vs new ELO, Elo inflation, ratings inflation, how strong were old GMs, how strong were old masters, masters today versus masters of old, Is Fischer stronger than Capablanca, is Alekine stronger than Kasparov..