|
< Earlier Kibitzing · PAGE 50 OF 57 ·
Later Kibitzing> |
Aug-02-21
 | | beatgiant: Old Guard:
1. Anand 2753
2. Topalov 2730
3. Adams 2716
4. Svidler 2714
5. Sadler 2694
6. Almasi 2687
7. Ivanchuk 2678
8. Gelfand 2675
9. Nisipeanu 2671
10. Malakhov 2666 |
|
Aug-02-21
| | beatgiant: Novelty: The top player for each letter, but this time using "first name." Yes, I know, name conventions vary by culture, so Viswanathan (Anand) or Quang (Le) don't play the same role in the structure of the name as a "first name" in a European name. I simply define "the part of the name after the comma in the FIDE profile" as the "first name." And without further ado, here's the list. 1. M - Magnus 2847
2. F - Fabiano 2806
3. L - Liren 2799
4. I - Ian 2792
5. S - Shakhriyar 2782
6. A - Alexander 2778
7. W - Wesley 2772
8. T - Teimour 2763
9. R - Richard 2763
10. V - Viswanathan 2753
11. H - Hao 2744
12. J - Jan-Krysztof 2738
13. P - Pentala 2730
14. Y - Yi 2725
15-16. D - Dmitry 2724
15-16. N - Nikita 2724
17. Q - Quang 2709
18. E - Evgeny 2706
19. X - Xiangzhi 2705
20-21. B - Bassem 2703
20-21. K - Kirill 2703
22. Z - Zoltan 2687
23. G - Gawain 2684
24. C - Chao 2683
25. O - Olexandr 2604
26. U - Ulf 2516
I was surprised by the huge dropoff at the letter U, which also came in last in surnames. The top player with first name beginning with U is none other than Ulf Andersson! He must have held this rank for most of his career. In former days for the surnames we had the two Wolfgangs - Unzicker and Uhlmann. |
|
Aug-02-21
 | | OhioChessFan: An amusing list. |
|
Aug-02-21
| | beatgiant: And here's the list of players who are top in both the surname letter and "first name" letter. 1. Carlsen, Magnus (C,M) 2847
2. Ding, Liren (D, L) 2799
3. Nepomniachtchi, Ian (N, I) 2792
4. Mamedyarov, Shakhriyar (M, S) 2782
5. Grischuk, Alexander (G, A) 2778
6. So, Wesley (S, W) 2772
7. Radjabov, Teimour (R,T) 2763
8. Wang, Hao (W, H) 2744
9. Harikrishna, Pentala (H, P) 2730
10. Le, Quang Liem (L, Q) 2709
11. Bu, Xiangzhi (B, X) 2705
As we can see, it's harder to accomplish both simultaneously. But if you aspire to be the top player in the world, you must be the top player for your surname's initial and also your first name's initial. |
|
| Aug-02-21 | | nok: Guys I thought of a quick and dirty rating system that might interest you. A sort of simple Elo, or Noko if you will. Matchups are grouped in 6 difficulty levels. An opponent within 40 rating points is difficulty 0. B/w 40 and 100, it's 1; less than 200, it's 2; less than 300 it's 3; b/w 300-500 it's 4, and more than 500 is difficulty 5. That's for the underdog, the favorite has corresponding negative difficulty. You score +5 for a win, -5 for a loss, 0 for draw, and add in the difficulty. That's all. Computing your new rating becomes trivial, and there's no rounding of decimals. There's no need for the complexity of Elo. *Example: you're a 2580 playing a 2725. He outrates you by more than 100 but less than 200 points, so difficulty 2. A win nets you 5+2 = 7 pts, a loss -5+2 = -3. |
|
Aug-03-21
| | beatgiant: <nok>
And how do we interpret the resulting score? As a kind of event performance metric? It might be interesting to see a few examples from a recent event. |
|
| Aug-03-21 | | nok: Difficulty is a proxy for Elo's expected score, but coarser grained and more intuitive. I think the result will be very close to Elo with significantly less effort. |
|
Aug-03-21
| | beatgiant: <nok>
First off, to make sure we understand, let's do an example. Suppose we start with 5 unrated players A, B, C, D and E. They play a round robin. A wins all games, B loses all games, and C, D and E all lose to A, beat B, and draw with each other.A scores +25, B scores -25, and C, D, E score 0. The rating list is now: A: 25
C,D,E: 0
B: -25
Now they play another event. This time, the game between A and B features a 50-point gap and so it's difficulty -1 for A, 1 for B. Suppose the results are the same as before, with A winning all games, B losing all games, and C, D, E losing to A, beating B and drawing each other. This time A scores 24 points, B scores -24, and the others score 0 and the new rating list is: A: 49
C, D, E: 0
B: -49
Did I get that right? |
|
Aug-03-21
| | beatgiant: Oh, duh, of course my example doesn't work because A only scores +20 in the first event and +19 in the second event, and B scores resp. -20 and -19. The final rating list would be: A: 39
B, C, D: 0
E: -39
Other than that, did I get it right? |
|
Aug-03-21
| | beatgiant: Uh, the final rating list would be:
A: 39
C,D,E: 0
B: -39
Did I finally get it right? |
|
| Aug-03-21 | | nok: Yes indeed |
|
Aug-03-21
| | beatgiant: <nok>
Then the main side effect I see with this proposed new system is, it's easy for a stronger player to rack up rating points against a much weaker player. For example in Elo, if a 2800 plays hundreds of games with a 2200 player, the stronger player won't be able to gain hundreds of rating points. That's because a draw causes the stronger player to lose points, and a win brings only a fractional gain. The new system doesn't have those brakes. And if rankings are worth money (for example, are used to determine invitations to major events), professional players will be forced to pursue quantity and play hundreds of games per year with much weaker players to keep their numbers up. |
|
Aug-03-21
 | | alexmagnus: <beatgiant> The current system has the same incentive due to the (mathematically stupid) 400 point rule. With the effect that if you play someone rated more than 400 points below you, you are expected to consistently gain rating. Which is how we saw those obscure Ukrainians on some Blitz rating lists. |
|
| Aug-03-21 | | nok: <That's because a draw causes the stronger player to lose points, and a win brings only a fractional gain. The new system doesn't have those brakes.> The stronger player still loses points with a draw. There's a bigger brake in the new system: when the gap gets as big as 500 you can't win any rating, as the difficulty of -5 sucks out all your points. To game the system you'd have to only play people in a narrow rating strip ~450 points below you, obv harder to arrange than today. |
|
Sep-01-21
| | alexmagnus: September:
European Union:
1. Giri 2777
2. Vachier-Lagrave 2763
3. Rapport 2760
4. Duda 2756
5. Firouzja 2754
6. Topalov 2730
7. Vallejo 2701
8. Wojtaszek 2698
9. Van Foreest 2691
10. Almasi 2687
Former Soviet Union:
1. Nepomniachtchi 2792
2. Aronian 2782
3. Grischuk 2775
4. Radjabov 2763
5. Mamedyarov 2762
6. Karjakin 2758
7. Andreikin 2728
8. Vitiugov 2727
9. Esipenko 2720
10. Dubov 2714
Former British Empire:
1. Caruana 2800
2. So 2778
3. Dominguez 2760
4. Anand 2753
5. Nakamura 2736
6. Vidit 2727
7. Harikrishna 2725
8. Shankland 2720
9. Adams 2716
10. Amin 2704
Asia:
1. Ding 2799
2. Anand 2753
3. Wang 2744
4. Vidit 2727
5-6. Harikrishna 2725
5-6. Wei 2725
7. Le 2709
8. Bu 2705
9. Yu 2704
10. Maghsoodloo 2695
Born later than the world champion:
1. Caruana 2800
2. Ding 2799
3. So 2778
4. Giri 2777
5. Rapport 2760
6. Duda 2756
7. Firouzja 2754
8. Vidit 2727
9. Wei 2725
10-11. Shankland 2720
10-11. Esipenko 2720
Nuclear powers:
1. Caruana 2800
2. Ding 2799
3. Nepomniachtchi 2792
4. So 2778
5. Grischuk 2775
6. Vachier-Lagrave 2763
7. Dominguez 2760
8. Karjakin 2758
9. Firouzja 2754
10. Anand 2753 |
|
Sep-25-21
| | beatgiant: I have been too busy this month to publish my usual lists, but here's one that's special to me: the top 10 players with names natively in a non-Latin script. The names won't display currently. When we repair non-Latin scripts in kibitzing (which WILL happen), the names will appear. 1. ? 2799
2. ?????????? 2792
3. ??????? 2782
4. ???? 2777
5. ?????? 2775
6. ??????? 2758
7. ??????? 2754
8. ?????? 2753
9. ? 2744
10. ?? 2736 |
|
Sep-25-21
| | beatgiant: And out of curiosity, I list below the top 10 scripts by highest ranked player whose name natively appears in that script. 1. Latin 2855
2. Han ideograph 2799
3. Cyrillic 2792
4. Armenian 2782
5. Devanagiri 2777
6. Arabic 2754
7. Tamil 2753
8. Hebrew 2680
9. Malayalam 2652
10. Greek 2623 |
|
Sep-25-21
| | beatgiant: Correction to my list above:
10. Bengali 2630
11. Telegu 2626
12. Greek 2623 |
|
Sep-25-21
| | beatgiant: Another correction to my list above:
1. Latin 2855
2. Han ideograph 2799
3. Cyrillic 2792
4. Armenian 2782
5. Devanagari 2777
6. Arabic 2754
7. Tamil 2753
8. Telugu 2725
9. Hebrew 2680
10. Malayalam 2652
11. Bengali 2630
12. Greek 2623 |
|
Sep-25-21
| | beatgiant: And here are some more rankings, but I've realized it's a lot of work to get this right, so I won't claim ranking numbers. Georgian 2582
Mongolian 2536
Kannada 2522
Burmese 2508
I didn't find any others over 2500. |
|
| Sep-25-21 | | nok: <7. Tamil 2753
8. Telugu 2725...
10. Malayalam 2652>
Aren't these like Czech and Swedish basically? |
|
Sep-25-21
| | beatgiant: <nok>
I'm focusing on <writing systems>, and those are three different ones. those differences go much farther than just different sets of diacritical marks on the Latin alphabet. https://en.wikipedia.org/wiki/Tamil...
https://en.wikipedia.org/wiki/Malay...
https://en.wikipedia.org/wiki/Telug...
And before anyone asks, yes I did try to check whether Vidit Santosh Gujrathi is Gujrati, and concluded that he isn't. |
|
| Sep-25-21 | | nok: Fair enough. Vidit's "surname" is misleading at best, and I don't think he uses it either: Vidit Santosh Gujrathi (kibitz #28) His player page should really be <S Vidit>. |
|
Sep-26-21
| | beatgiant: I found the following chain in the rating list. Gujrathi (Vidit) (2727) is Marathi.
Marathee (1669) is French.
French (1952) is English.
Englisch (1868) is Austrian.
Viennois (1981) is French.
French (1714) is American.
Amerika (2227) is Latvian.
Latvenas (1141) is Lithuanian.
Lith (1683) is Dutch.
Duch (2206) is Polish.
Polich (2149) is Ukrainian.
Ukrainskaya (1059) is Russian.
Rossi (2235) is Italian.
The cycle ends with Italiano (1462) who is actually Italian. |
|
Sep-27-21
| | beatgiant: <I didn't find any others over 2500> In fact, Megaranto (2548) should probably count toward one of the Indonesian scripts. My first guess is Javanese, based on looking up his place of birth on a linguistic map of Indonesia. |
|
 |
|
|
< Earlier Kibitzing · PAGE 50 OF 57 ·
Later Kibitzing> |
Kibitzing
