I tried asking for a top 50, and that was even worse for poor Keres. The top 25 didn’t change, but Keres wasn’t even included in the top 50. Topalov was as low as 39th, well below for example Mamedyarov, Svidler, Radjabov and Grischuk. Other surprise inclusions in the top 50 are Bologan, Beliavsky and Adams. Maybe the absent Bogo, Zukertort, Pillsbury, Fine or Leko might have deserved top 50 if Bologan, who at best was number 18 on a rating list, is placed in the top 50.
The 25 greatest ever according to ChatGPT
Sort:
Wonder if there was actually an algorithm to this or it was just a aggregation of popular opinion
sawdof wrote:
Wonder if there was actually an algorithm to this or it was just a aggregation of popular opinion
The strange thing is that so much of it is quite reasonable, and then a few really weird things pop up. I mean, Radjabov well ahead of Topalov? The latter was #1 for quite a while, won numerous super tournaments, lost two title matches very narrowly etc. Or Bologan ahead of Leko, who won all the biggest super tournaments including Candidates and drew a title match, while Bologan briefly reached #18… The top 10 is quite good though, only the Kramnik and Lasker positions stand out as difficult to agree with for me.
I asked ChatGPT what it based its ranking on, and it answered ”tournament and match victories, longevity at the top of world rankings, influence on the game, opinions from experts, polls and publications” etc.
fabelhaft wrote:
..., Radjabov well ahead of Topalov? ... Or Bologan ahead of Leko ...
Radjabov was famous as the up and coming youngster while bologan was famous for his kid books. Looks like popularity won
My opinions: Botvinnk looks too high on this list - unless influence in the Soviet school counts heavily. I'd definitely put him below Capablanca.
Korchnoi is too low.
Keres should be in the top 25, maybe ahead of Tal even.
Bronstein is missing. I was happy to see Euwe on it even though his peak was short lived.
Reti looks too high to me as well- top 100 not top 25. I dont understand Aronian and So being top 25.
Just my opinion of course-- Bill
The funny thing with ChatGPT is that it mixes very believable facts and assessements with quite insane stuff. I recall Aronian asking it for a biography of himself, and getting an extremely detailed and correct, long text. But it had mysteriously added a couple of things that were totally made up.
I asked ChatGPT which World Champions that have scored the comparatively worst results. It listed Euwe, Smyslov, Tal, Petrosian, Spassky, Anand and Kramnik as seven of the top eight. This is quite reasonable if you do not count Ding. But first on the list it placed Steinitz, claiming the rank was based on comparative strength vs contemporary opposition, title defenses and dominance.
Steinitz won four title matches, 25 games in a row against contemporary top opposition, won every match he played (and he played many) for more than 30 years, etc. How ChatGPT can get all other things so right and then mean that Steinitz did worst with regards to title defenses and dominance is strange.
I tried various ways of asking for this sort of ranking, but the one with a top 25 ended up like this:
1. Kasparov
2. Carlsen
3. Fischer
4. Karpov
5. Kramnik
6. Botvinnik
7. Capablanca
8. Lasker
9. Alekhine
10. Anand
11. Morphy
12. Smyslov
13. Petrosian
14. Spassky
15. Tal
16. Steinitz
17. Ding Liren
18. Caruana
19. Aronian
20. So
21. Korchnoi
22. Euwe
23. Reti
24. Rubinstein
25. Larsen
Biggest surprises to me: Kramnik gets very high no matter how the question is asked. Top 5 here, always ahead of Lasker and always several places ahead of Anand. So and Aronian ahead of Korchnoi?! Reti and Larsen ahead of Keres?! No Topalov?! I think Steinitz is too low. But the high rankings of So, Reti and Kramnik are the things that stand out most to me, and also the lack of appreciation for Topalov and Keres.