Candidates 2024 - What Statistics say?
Say you are Hikaru Nakamura and you finish second place in the Candidates Tournament. Is it bad or not? Say you are Abasov and you finish last. Is it bad or not? Let's see what Statistics say.Before the start: Probability of Winning for each player
Each player started the tournament with an Elo, thus since Elo is nothing else than the % of Winning or Not Winning you can guess who has the highest percentage of winning.
The average % of each player for each match is the following:
| Names | Avg % |
|---|---|
| Gukesh D | 50% |
| Hikaru Nakamura | 57% |
| Ian Nepomniachtchi | 50% |
| Fabiano Caruana | 59% |
| R Praggnanandhaa | 50% |
| Vidit Gujarthi | 47% |
| Alireza Firouzja | 52% |
| Nijat Abasov | 33% |
We see that the average Elo of each player is 2745, so Gukesh, Nepo and Pragg are just in the mean. This means that players like Caruana has 9% more chances of them and Abasov 17% less.
Points Classification and Percentage of points to the max:
The games are over and now we see the final classification as follows:
| Names | Pts. |
|---|---|
| Gukesh D | 9 |
| Hikaru Nakamura | 8,5 |
| Ian Nepomniachtchi | 8,5 |
| Fabiano Caruana | 8,5 |
| R Praggnanandhaa | 7 |
| Vidit Gujarthi | 6 |
| Alireza Firouzja | 5 |
| Nijat Abasov | 3,5 |
Gukesh won.. but if we compare the points out of the 14 matches we can calculate the % of their results compared to the max points (14 games aka 14 points).
| Names | Pts. | Avg % |
|---|---|---|
| Gukesh D | 9 | 64% |
| Hikaru Nakamura | 8,5 | 61% |
| Ian Nepomniachtchi | 8,5 | 61% |
| Fabiano Caruana | 8,5 | 61% |
| R Praggnanandhaa | 7 | 50% |
| Vidit Gujarthi | 6 | 43% |
| Alireza Firouzja | 5 | 36% |
| Nijat Abasov | 3,5 | 25% |
Difference between Probability of Winning to games Performance:
And here we are! We can match the starting Elo % with the Results % to see who performed better that the previsions!
| Names | Elo % | Pt % | Diff. |
|---|---|---|---|
| Gukesh D | 50% | 64% | 14% |
| Hikaru Nakamura | 57% | 61% | 4% |
| Ian Nepomniachtchi | 50% | 61% | 11% |
| Fabiano Caruana | 59% | 61% | 2% |
| R Praggnanandhaa | 50% | 50% | 0% |
| Vidit Gujarthi | 47% | 43% | -4% |
| Alireza Firouzja | 52% | 36% | -16% |
| Nijat Abasov | 33% | 25% | -8% |
So here it is what Statistics says on who should be mad and who should be grateful!
And we can see that Gukesh winning exceeded the expectations by a great 14%... but that's not the most shocking thing! The highest difference is that of Firouzja with a loss of 16%!
Other things arises, for example Caruana that even though he was the most rated, he still had a 2% improvement
Performance VS Starting Elo:
What about Performance VS Starting Elo? Let's see who played better than "himself"!
| Names | Elo | Perf. | Diff |
|---|---|---|---|
| Gukesh D | 2743 | 2859 | 104% |
| Hikaru Nakamura | 2789 | 2824 | 101% |
| Ian Nepomniachtchi | 2758 | 2829 | 103% |
| Fabiano Caruana | 2803 | 2822 | 101% |
| R Praggnanandhaa | 2747 | 2745 | 100% |
| Vidit Gujarthi | 2727 | 2690 | 99% |
| Alireza Firouzja | 2760 | 2628 | 95% |
| Nijat Abasov | 2632 | 2561 | 97% |
It's interesting to see that both Gukesh and Nepo achieved a much higher games performance compared to their starting points, and as we stated before, the most shocking result was that of Firouzja!
But nonetheless we can say that every one has done monstrous performances, even compared to themselves.
As a comparison i post some of my otb tournament performances and the difference with my Elo.
| Perf. | Diff |
|---|---|
| 1277 | 93% |
| 1414 | 105% |
| 1190 | 88% |
| 1245 | 96% |
| 1159 | 89% |
As an amateur maybe i lack consistency in results but to see that i can go way below myself only means that the Candidates games where at high level!
Maybe other analysis will follow..
What's your idea about this?
Please let me know in the comments!
