August 2016 ratings: Monster Maxime reaches 2819 FIDE!

by Albert Silver
7/31/2016 – What can one say except 'Wow'? 25-year-old Frenchman Maxime Vachier Lagrave continues his run after a win at Dortmund and crushing win over Svidler to reach 2819 FIDE and rocket into the world no.2 spot. He will bring great momentum into the approaching Sinquefield Cup. Be sure to read about 15-year-old Iranian phenom Parham Maghsoodloo who gained 75 Elo and is now the top rated in his country. Report, stats, and more.

ChessBase 14 Download ChessBase 14 Download

Everyone uses ChessBase, from the World Champion to the amateur next door. Start your personal success story with ChessBase 14 and enjoy your chess even more!


Along with the ChessBase 14 program you can access the Live Database of 8 million games, and receive three months of free ChesssBase Account Premium membership and all of our online apps! Have a look today!

More...

FIDE August 2016 – Top 100 Players

Can there be any doubt who the name of the month is? Frenchman Maxime Vachier-Lagrave has been on a run that has been breathtaking to watch. After winning Dortmund he now followed up with a crushing win over Peter Svidler by 3.0/4 in Biel. His rating of 2819 is right up there as one of the highest ever registered, and the 25-year-old will arrive at the Sinquefield Cup with huge momentum.

A special congratulations to Paraguayan GM Axel Bachmann who broke into the Top 100 for the first time after several near misses. The South American has been incredibly busy this year, racking up no fewer than 133 rated classical games! This month he had 33, from three opens in the US, and one in Spain.

Rk
Name
Ti.
Fed
Rtg
Gms
B-Year
1 Carlsen, Magnus g NOR 2857 10 1990
2 Vachier-Lagrave, Maxime g FRA 2819 11 1990
3 Kramnik, Vladimir g RUS 2808 7 1975
4 Caruana, Fabiano g USA 2807 7 1992
5 Aronian, Levon g ARM 2792 0 1982
6 Nakamura, Hikaru g USA 2791 10 1987
7 So, Wesley g USA 2771 10 1993
8 Anand, Viswanathan g IND 2770 0 1969
9 Giri, Anish g NED 2769 10 1994
10 Karjakin, Sergey g RUS 2769 10 1990
11 Mamedyarov, Shakhriyar g AZE 2764 0 1985
12 Topalov, Veselin g BUL 2761 0 1975
13 Ding, Liren g CHN 2755 16 1992
14 Grischuk, Alexander g RUS 2754 4 1983
15 Li, Chao b g CHN 2753 9 1989
16 Harikrishna, P. g IND 2752 13 1986
17 Rapport, Richard g HUN 2752 0 1996
18 Svidler, Peter g RUS 2751 4 1976
19 Gelfand, Boris g ISR 2743 6 1968
20 Navara, David g CZE 2742 7 1985
21 Nepomniachtchi, Ian g RUS 2740 9 1990
22 Wang, Yue g CHN 2737 13 1987
23 Eljanov, Pavel g UKR 2737 0 1983
24 Andreikin, Dmitry g RUS 2733 0 1990
25 Wojtaszek, Radoslaw g POL 2733 0 1987
26 Tomashevsky, Evgeny g RUS 2731 0 1987
27 Vitiugov, Nikita g RUS 2728 0 1987
28 Adams, Michael g ENG 2727 0 1971
29 Yu, Yangyi g CHN 2725 13 1994
30 Ivanchuk, Vassily g UKR 2723 9 1969
31 Le, Quang Liem g VIE 2723 0 1991
32 Radjabov, Teimour g AZE 2722 0 1987
33 Inarkiev, Ernesto g RUS 2721 6 1985
34 Dominguez Perez, Leinier g CUB 2720 7 1983
35 Bu, Xiangzhi g CHN 2715 13 1985
36 Vallejo Pons, Francisco g ESP 2713 0 1982
37 Wang, Hao g CHN 2712 13 1989
38 Jakovenko, Dmitry g RUS 2712 0 1983
39 Wei, Yi g CHN 2709 10 1999
40 Leko, Peter g HUN 2709 9 1979
41 Ponomariov, Ruslan g UKR 2709 7 1983
42 Rodshtein, Maxim g ISR 2698 0 1989
43 Malakhov, Vladimir g RUS 2697 0 1980
44 Naiditsch, Arkadij g AZE 2696 16 1985
45 Kasimdzhanov, Rustam g UZB 2696 0 1979
46 Bacrot, Etienne g FRA 2692 9 1983
47 Kryvoruchko, Yuriy g UKR 2691 0 1986
48 Rublevsky, Sergei g RUS 2689 0 1974
49 Cheparinov, Ivan g BUL 2688 0 1986
50 Nisipeanu, Liviu-Dieter g GER 2687 16 1976
51 Smirin, Ilia g ISR 2687 9 1968
52 Matlakov, Maxim g RUS 2686 20 1991
53 Ragger, Markus g AUT 2686 0 1988
54 Almasi, Zoltan g HUN 2684 0 1976
55 Morozevich, Alexander g RUS 2683 0 1977
56 Piorun, Kacper g POL 2681 9 1991
57 Safarli, Eltaj g AZE 2678 0 1992
58 Najer, Evgeniy g RUS 2677 7 1977
59 Fressinet, Laurent g FRA 2677 0 1981
60 Ganguly, Surya Shekhar g IND 2676 0 1983
61 Duda, Jan-Krzysztof g POL 2675 9 1998
62 Akopian, Vladimir g ARM 2675 0 1971
63 Bareev, Evgeny g CAN 2675 0 1966
64 Robson, Ray g USA 2674 0 1994
65 Shirov, Alexei g LAT 2674 0 1972
66 Zvjaginsev, Vadim g RUS 2673 13 1976
67 Granda Zuniga, Julio E g PER 2672 9 1967
68 Adhiban, B. g IND 2671 0 1992
69 McShane, Luke J g ENG 2671 0 1984
70 Sargissian, Gabriel g ARM 2670 10 1983
71 Fedoseev, Vladimir g RUS 2670 9 1995
72 Sadler, Matthew D g ENG 2670 0 1974
73 Vidit, Santosh Gujrathi g IND 2669 9 1994
74 Jobava, Baadur g GEO 2669 0 1983
75 Onischuk, Alexander g USA 2668 0 1975
76 Saric, Ivan g CRO 2667 10 1990
77 Short, Nigel D g ENG 2666 10 1965
78 Mamedov, Rauf g AZE 2666 0 1988
79 Movsesian, Sergei g ARM 2666 0 1978
80 Areshchenko, Alexander g UKR 2665 18 1986
81 Artemiev, Vladislav g RUS 2665 9 1998
82 Kovalenko, Igor g LAT 2664 9 1988
83 Howell, David W L g ENG 2663 0 1990
84 Markus, Robert g SRB 2662 0 1983
85 Efimenko, Zahar g UKR 2661 9 1985
86 Shankland, Samuel L g USA 2661 0 1991
87 Dreev, Aleksey g RUS 2660 16 1969
88 Motylev, Alexander g RUS 2660 0 1979
89 Negi, Parimarjan g IND 2660 0 1993
90 Tkachiev, Vladislav g FRA 2660 0 1973
91 Bologan, Victor g MDA 2659 9 1971
92 Sasikiran, Krishnan g IND 2658 10 1981
93 Hou, Yifan g CHN 2658 9 1994
94 Bukavshin, Ivan g RUS 2658 0 1995
95 Van Wely, Loek g NED 2657 9 1972
96 Ipatov, Alexander g TUR 2657 0 1993
97 Meier, Georg g GER 2657 0 1987
98 Bachmann, Axel g PAR 2656 33 1989
99 Zhigalko, Sergei g BLR 2656 18 1989
100 Korobov, Anton g UKR 2656 0 1985

Top climbers and descenders 

The biggest gain of the month was not the frenchman, even if it was the most noteworthy, it was by Azeri GM Arkadij Naiditsch, who earned 29 Elo and is nearly back to his former 2700+ self.

Rk
Old
Name
Fed
Ti.
Rating
Old
Gms
 2 4 Vachier-Lagrave, Maxime FRA GM 2819 +21 2798 11
 9 7 Giri, Anish NED GM 2769 -16 2785 10
 13 8 Ding, Liren CHN GM 2755 -23 2778 16
 21 32 Nepomniachtchi, Ian RUS GM 2740 +15 2725 9
 37 22 Wang, Hao CHN GM 2712 -22 2734 13
39 44 Wei, Yi CHN GM 2709 +13 2696 10
44 74 Naiditsch, Arkadij AZE GM 2696 +29 2667 16
50 64 Nisipeanu, Liviu-Dieter GER GM 2687 +13 2674 16
51 61 Smirin, Ilia ISR GM 2687 +11 2676 9
58 50 Najer, Evgeniy RUS GM 2677 -10 2687 7
67 41 Granda Zuniga, Julio E PER GM 2672 -27 2699 9
73 92 Vidit, Santosh Gujrathi IND GM 2669 +11 2658 9
77 Short, Nigel D ENG GM 2666 +14 10
80 101 Areshchenko, Alexander UKR GM 2665 +11 2654 18

FIDE Top 100 Women

The biggest gain was by 22-year-old Daria Pustovoitova, who moved up 115 Elo to 2386 FIDE. This was the result of three excellent results, not least of which was her third place in the Russian Women Higher League, where she gained 54 Elo no less.

Rk
Name
Ti.
Fed
Rtg
Gms
B-Year
1 Hou, Yifan g CHN 2658 9 1994
2 Koneru, Humpy g IND 2580 11 1987
3 Ju, Wenjun g CHN 2575 15 1991
4 Muzychuk, Anna g UKR 2544 11 1990
5 Harika, Dronavalli g IND 2542 15 1991
6 Muzychuk, Mariya g UKR 2539 11 1992
7 Kosteniuk, Alexandra g RUS 2538 9 1984
8 Cmilyte, Viktorija g LTU 2536 0 1983
9 Lagno, Kateryna g RUS 2530 0 1989
10 Dzagnidze, Nana g GEO 2529 0 1987
11 Gunina, Valentina g RUS 2520 0 1989
12 Zhao, Xue g CHN 2518 15 1985
13 Stefanova, Antoaneta g BUL 2515 15 1979
14 Sebag, Marie g FRA 2488 0 1986
15 Javakhishvili, Lela m GEO 2486 11 1984
16 Pogonina, Natalija wg RUS 2484 0 1985
17 Kosintseva, Nadezhda g RUS 2483 0 1985
18 Lei, Tingjie wg CHN 2480 4 1997
19 Paehtz, Elisabeth m GER 2476 9 1985
20 Tan, Zhongyi wg CHN 2475 15 1991
21 Goryachkina, Aleksandra wg RUS 2475 9 1998
22 Zhukova, Natalia g UKR 2475 0 1979
23 Batsiashvili, Nino m GEO 2474 10 1987
24 Shen, Yang m CHN 2474 3 1989
25 Hoang, Thanh Trang g HUN 2467 0 1980
26 Bodnaruk, Anastasia m RUS 2465 0 1992
27 Khotenashvili, Bela g GEO 2463 11 1988
28 Kashlinskaya, Alina m RUS 2462 9 1993
29 Socko, Monika g POL 2454 9 1978
30 Girya, Olga wg RUS 2452 15 1991
31 Ushenina, Anna g UKR 2452 7 1985
32 Skripchenko, Almira m FRA 2452 0 1976
33 Galliamova, Alisa m RUS 2450 0 1972
34 Huang, Qian wg CHN 2450 0 1986
35 Zatonskih, Anna m USA 2449 0 1978
36 Krush, Irina g USA 2444 9 1983
37 Danielian, Elina g ARM 2444 0 1978
38 Cramling, Pia g SWE 2441 20 1963
39 Hunt, Harriet V m ENG 2440 0 1978
40 Mkrtchian, Lilit m ARM 2439 11 1982
41 Zawadzka, Jolanta wg POL 2439 10 1987
42 Kovalevskaya, Ekaterina m RUS 2437 19 1974
43 Nomin-Erdene, Davaademberel m MGL 2437 7 2000
44 Cori T., Deysi wg PER 2433 19 1993
45 Munguntuul, Batkhuyag m MGL 2428 0 1987
46 Atalik, Ekaterina m TUR 2426 0 1982
47 Khademalsharieh, Sarasadat m IRI 2423 10 1997
48 Saduakassova, Dinara wg KAZ 2423 7 1996
49 Melia, Salome m GEO 2419 10 1987
50 Sukandar, Irine Kharisma m INA 2418 9 1992
51 Ding, Yixin wg CHN 2418 4 1991
52 Guo, Qi m CHN 2417 4 1995
53 Gaponenko, Inna m UKR 2416 0 1976
54 Shvayger, Yuliya wm ISR 2415 10 1994
55 Karavade, Eesha m IND 2415 0 1987
56 Padmini, Rout m IND 2415 0 1994
57 Daulyte, Deimante m LTU 2412 9 1989
58 Nechaeva, Marina m RUS 2410 9 1986
59 Szczepkowska-Horowska, Karina wg POL 2409 9 1987
60 Vega Gutierrez, Sabrina m ESP 2408 0 1987
61 Bojkovic, Natasa m SRB 2406 0 1971
62 Derakhshani, Dorsa wm IRI 2405 0 1998
63 Shuvalova, Polina wf RUS 2398 19 2001
64 Khurtsidze, Nino m GEO 2398 9 1975
65 Ni, Shiqun wg CHN 2398 4 1997
66 Tania, Sachdev m IND 2396 9 1986
67 Peptan, Corina-Isabela m ROU 2395 0 1978
68 Batchimeg, Tuvshintugs m MGL 2391 0 1986
69 Bulmaga, Irina m ROU 2389 20 1993
70 Abdumalik, Zhansaya wg KAZ 2389 9 2000
71 Zimina, Olga m ITA 2389 7 1982
72 Rajlich, Iweta m POL 2389 0 1981
73 Pustovoitova, Daria f RUS 2386 25 1994
74 Houska, Jovanka m ENG 2386 0 1980
75 Videnova, Iva m BUL 2386 0 1987
76 Ziaziulkina, Nastassia m BLR 2386 0 1995
77 Vijayalakshmi, Subbaraman m IND 2384 0 1979
78 Kovanova, Baira wg RUS 2382 9 1987
79 Abrahamyan, Tatev wg USA 2382 6 1988
80 Aulia, Medina Warda wg INA 2382 0 1997
81 Arakhamia-Grant, Ketevan g SCO 2381 9 1968
82 Gara, Ticia wg HUN 2379 6 1984
83 Ryjanova, Julia wg RUS 2379 0 1974
84 Lujan, Carolina m ARG 2378 10 1985
85 Brunello, Marina f ITA 2376 13 1994
86 Savina, Anastasia m RUS 2375 9 1992
87 Michna, Marta wg GER 2375 0 1978
88 Arabidze, Meri m GEO 2374 10 1994
89 Vasilevich, Tatjana m UKR 2372 9 1977
90 Berend, Elvira wg LUX 2372 0 1965
91 Lazarne Vajda, Szidonia m HUN 2372 0 1979
92 Hoolt, Sarah wg GER 2371 9 1988
93 Guramishvili, Sopiko m GEO 2370 7 1991
94 Soumya, Swaminathan wg IND 2370 0 1989
95 Wang, Jue wg CHN 2369 15 1995
96 Fierro Baquero, Martha L. m ECU 2369 10 1977
97 Khukhashvili, Sopiko m GEO 2368 9 1985
98 Peng, Zhaoqin g NED 2368 9 1968
99 Zhai, Mo wg CHN 2368 4 1996
100 Stockova, Zuzana m SVK 2368 0 1977

FIDE Top 100 Juniors

Wei Yi has bounced back after a period in which he dropped under 2700. After winning the Chinese Championship in May (above), he scored 50% at the elite Bilbao Masters in July and is now back at 2709.

The biggest news was that of 15-year-old Parham Mahhsoodloo though. Rated 2501 already, his K factor is the strict minimum, yet in spite of this, the Iranian prodigy leapt an incredible 75 Elo (!!) to 2576 and is now the top Iranian player as well. Bear in mind, this great talent was also the youngest qualikfier for last year's World Cup, and qualified for the Iranian team at the Baku Olympiad. Iran's team, all from qualification, will field three players rated 15 or less. Astonishing.

A special heads up to both Dutch teen Jorden van Foreest, American Samuel Sevian, and Grigoriy Oparin from Russian, all of whom broke 2600 this month for the first time.

Rk
Name
Ti.
Fed
Rtg
Gms
B-Year
1 Rapport, Richard g HUN 2752 0 1996
2 Wei, Yi g CHN 2709 10 1999
3 Duda, Jan-Krzysztof g POL 2675 9 1998
4 Artemiev, Vladislav g RUS 2665 9 1998
5 Dubov, Daniil g RUS 2648 9 1996
6 Xiong, Jeffery g USA 2633 13 2000
7 Nyzhnyk, Illya g UKR 2621 15 1996
8 Bluebaum, Matthias g GER 2618 0 1997
9 Oparin, Grigoriy g RUS 2617 9 1997
10 Eliseev, Urii g RUS 2606 9 1996
11 Van Foreest, Jorden g NED 2601 9 1999
12 Sevian, Samuel g USA 2600 18 2000
13 Bortnyk, Olexandr g UKR 2588 7 1996
14 Alekseenko, Kirill g RUS 2582 16 1997
15 Donchenko, Alexander g GER 2581 9 1998
16 Gledura, Benjamin m HUN 2581 5 1999
17 Maghsoodloo, Parham IRI 2576 22 2000
18 Wagner, Dennis g GER 2572 0 1997
19 Tari, Aryan g NOR 2571 0 1999
20 Ghosh, Diptayan g IND 2570 0 1998
21 Antipov, Mikhail Al. g RUS 2565 16 1997
22 Vavulin, Maksim m RUS 2559 16 1998
23 Vaibhav, Suri g IND 2556 0 1997
24 Li, Ruifeng m USA 2555 33 2001
25 Gordievsky, Dmitry m RUS 2555 9 1996
26 Rambaldi, Francesco g ITA 2553 6 1999
27 Pichot, Alan g ARG 2550 23 1998
28 Chigaev, Maksim m RUS 2550 9 1996
29 Svane, Rasmus m GER 2546 9 1997
30 Santos Latasa, Jaime m ESP 2545 9 1996
31 Aravindh,Chithambaram VR. g IND 2543 0 1999
32 Burke, John M m USA 2537 18 2001
33 Bai, Jinshi g CHN 2537 15 1999
34 Boruchovsky, Avital g ISR 2531 7 1997
35 Paravyan, David m RUS 2528 18 1998
36 Schroeder, Jan-Christian g GER 2523 15 1998
37 Troff, Kayden W g USA 2522 15 1998
38 Henriquez Villagra, Cristobal m CHI 2520 33 1996
39 Deac, Bogdan-Daniel m ROU 2520 0 2001
40 Xu, Yinglun CHN 2516 25 1996
41 Sunilduth Lyna, Narayanan g IND 2515 0 1998
42 Karthikeyan, Murali g IND 2514 27 1999
43 Petrosyan, Manuel m ARM 2512 6 1998
44 Yuffa, Daniil m RUS 2507 16 1997
45 Stefanov, Emil BUL 2505 27 1999
46 Supi, Luis Paulo m BRA 2505 0 1996
47 Fang, Yuxiang m CHN 2503 36 1996
48 Drozdowski, Kacper m POL 2503 18 1996
49 Lampert, Jonas m GER 2500 0 1997
50 Dastan, Muhammed Batuhan m TUR 2490 19 1997
51 Esipenko, Andrey f RUS 2490 17 2002
52 Petrov, Nikita m RUS 2490 0 1996
53 Kobo, Ori m ISR 2489 12 1997
54 Chandra, Akshat m USA 2489 9 1999
55 Tabatabaei, M.amin m IRI 2489 9 2001
56 Ali Marandi, Cemil Can m TUR 2488 17 1998
57 Yang, Darwin m USA 2488 16 1996
58 Smirnov, Anton m AUS 2486 9 2001
59 Gagare, Shardul g IND 2486 0 1997
60 Sanal, Vahap m TUR 2485 17 1998
61 Aryan Chopra m IND 2485 9 2001
62 Repka, Christopher m SVK 2485 9 1998
63 Harutyunian, Tigran K. m ARM 2485 0 1997
64 Kelires, Andreas m GRE 2483 17 1999
65 Kollars, Dmitrij m GER 2483 16 1999
66 Zajic, Milan m SRB 2483 9 1999
67 Theodorou, Nikolas GRE 2481 25 2000
68 Golubov, Saveliy m RUS 2481 18 2000
69 Lei, Tingjie wg CHN 2480 4 1997
70 Lorparizangeneh, Shahin m IRI 2478 10 1999
71 Stremavicius, Titas m LTU 2477 28 1998
72 Studer, Noel m SUI 2477 9 1996
73 Goryachkina, Aleksandra wg RUS 2475 9 1998
74 Georgiadis, Nico m SUI 2474 9 1996
75 Izzat, Kanan m AZE 2474 0 1996
76 Tran, Tuan Minh m VIE 2473 0 1997
77 Albornoz Cabrera, Carlos Daniel f CUB 2471 9 2000
78 Salomon, Johan m NOR 2470 9 1997
79 Preotu, Razvan m CAN 2469 9 1999
80 Puranik, Abhimanyu m IND 2469 9 2000
81 Codenotti, Marco m ITA 2468 7 1997
82 Korpa, Bence m HUN 2468 7 1998
83 Ebeling, Daniel m FIN 2468 0 1996
84 Bellahcene, Bilel m FRA 2467 7 1998
85 Triapishko, Olexandr RUS 2466 9 2000
86 Steinberg, Nitzan m ISR 2466 7 1998
87 Igonin, Temur m UZB 2466 0 2000
88 Liang, Awonder m USA 2465 18 2003
89 Moroni, Luca Jr f ITA 2465 9 2000
90 Firouzja, Alireza IRI 2464 22 2003
91 Zenzera, Alexey m RUS 2464 0 1997
92 Gholami, Aryan f IRI 2461 17 2001
93 Gurevich, Daniel m USA 2461 17 1997
94 Xu, Xiangyu CHN 2461 15 1999
95 Martirosyan, Haik M. m ARM 2461 9 2000
96 Enkhnar, Enkhbaatar f MGL 2460 20 1997
97 Khegay, Dmitriy f RUS 2460 9 1997
98 Martinez Alcantara, Jose Eduardo f PER 2458 0 1999
99 Basso, Pier Luigi m ITA 2455 9 1997
100 Harmon-Vellotti, Luke m USA 2454 20 1998
101 Le, Tuan Minh m VIE 2454 0 1996

FIDE Top 100 Girls

The single biggest gain was Canadian talent FM Qiyu Zhou, 15 years old, who leapt 123 Elo to reach 2308 FIDE. This is not her first time at 2300, having attained it before, but her youth and high K factor mean swings are her bread and butter at the moment.

Rk
Name
Ti.
Fed
Rtg
Gms
B-Year
1 Lei, Tingjie wg CHN 2480 4 1997
2 Goryachkina, Aleksandra wg RUS 2475 9 1998
3 Nomin-Erdene, Davaademberel m MGL 2437 7 2000
4 Khademalsharieh, Sarasadat m IRI 2423 10 1997
5 Saduakassova, Dinara wg KAZ 2423 7 1996
6 Derakhshani, Dorsa wm IRI 2405 0 1998
7 Shuvalova, Polina wf RUS 2398 19 2001
8 Ni, Shiqun wg CHN 2398 4 1997
9 Abdumalik, Zhansaya wg KAZ 2389 9 2000
10 Aulia, Medina Warda wg INA 2382 0 1997
11 Zhai, Mo wg CHN 2368 4 1996
12 Tsolakidou, Stavroula wg GRE 2363 18 2000
13 Bivol, Alina wm RUS 2362 9 1996
14 Mammadzada, Gunay wg AZE 2361 8 2000
15 Marjanovics, Annamaria wf HUN 2349 16 2001
16 Osmak, Iulija wf UKR 2344 7 1998
17 Maltsevskaya, Aleksandra RUS 2341 9 2002
18 Obolentseva, Alexandra wf RUS 2339 5 2001
19 Hojjatova, Aydan wf AZE 2339 0 1999
20 Khomeriki, Nino wm GEO 2338 0 1998
21 Pratyusha, Bodda wm IND 2329 0 1997
22 Tokhirjonova, Gulrukhbegim wm UZB 2328 0 1999
23 Rodriguez Rueda, Paula Andrea m COL 2321 0 1996
24 Kiolbasa, Oliwia wm POL 2319 7 2000
25 Styazhkina, Anna wm RUS 2315 18 1997
26 Tejaswini, Sagar wf IND 2310 0 2000
27 Zhou, Qiyu f CAN 2307 19 2000
28 Dordzhieva, Dinara wm RUS 2304 9 1999
29 Navrotescu, Andreea-Cristiana wm FRA 2297 18 1996
30 Osmanodja, Filiz wm GER 2297 9 1996
31 Xiao, Yiyi wf CHN 2292 4 1996
32 Frayna, Janelle Mae wm PHI 2292 0 1997
33 Kalaiyalahan, Akshaya f ENG 2288 0 2001
34 Assaubayeva, Bibisara wf KAZ 2287 9 2004
35 Martynkova, Olena wf UKR 2287 7 2000
36 Vaishali R wm IND 2284 6 2001
37 Rodionova, Daria wf RUS 2279 0 1998
38 Mammadova, Narmin wm AZE 2277 9 1999
39 Solozhenkina, Elizaveta wf RUS 2275 18 2003
40 Harazinska, Ewa wf POL 2272 9 1998
41 Blagojevic, Tijana wm MNE 2267 15 1997
42 Chernyak, Viktoria wf RUS 2264 18 1997
43 Imnadze, Nato wm GEO 2261 9 1996
44 Buksa, Nataliya wg UKR 2261 0 1996
45 Dimitrova, Aleksandra wf RUS 2259 9 2000
46 Velikic, Adela SRB 2258 7 1997
47 Fataliyeva, Ulviyya wm AZE 2258 0 1996
48 Amina, Battsooj f MGL 2255 36 2001
49 Kazarian, Anna-Maja f NED 2254 9 2000
50 Injac, Teodora wf SRB 2252 18 2000
51 Nicolas Zapata, Irene wm ESP 2252 0 1997
52 Maslova, Polina wf RUS 2248 9 1999
53 Gorti, Akshita f USA 2246 8 2002
54 Movileanu, Daniela wf ITA 2245 0 1996
55 Avramidou, Anastasia wf GRE 2244 7 2000
56 Gueci, Tea wf ITA 2243 0 1999
57 Antova, Gabriela f BUL 2241 15 2002
58 Unuk, Laura wm SLO 2240 7 1999
59 Gu, Tianlu wm CHN 2239 3 1997
60 Bluhm, Sonja Maria wf GER 2235 9 1998
61 Khokhlova, Daria RUS 2235 9 1999
62 Bykova, Anastasia wf RUS 2235 7 1997
63 Priyanka Nutakki IND 2234 8 2002
64 Heinemann, Josefine wm GER 2231 9 1998
65 Garcia Martin, Marta wf ESP 2229 19 2000
66 Eswaran, Ashritha wm USA 2226 18 2000
67 Badelka, Olga BLR 2222 16 2002
68 Makarenko, Alexandra wf RUS 2219 9 1996
69 Ren, Xiaoyi CHN 2216 15 1996
70 Marikova, Jana CZE 2214 4 1996
71 Garcia-Castany Musellas, Gal.la wf ESP 2213 10 1997
72 Goltseva, Ekaterina wf RUS 2213 9 2002
73 Bykovtsev, Agata wm USA 2209 18 1999
74 Kubicka, Anna POL 2209 16 1999
75 Zayas Gonzalez, Laura Amalia wf CUB 2208 0 1998
76 Gazikova, Veronika wf SVK 2207 0 1999
77 Michelle Catherina, P wm IND 2205 0 1996
78 Di Benedetto, Desiree wf ITA 2204 9 2000
79 Berdnyk, Mariia UKR 2204 0 2003
80 Parnali, S Dharia wm IND 2203 26 1997
81 Narva, Mai wm EST 2202 8 1999
82 Alinasab, Mobina wf IRI 2199 9 2000
83 Uuriintuya, Uurtsaikh wm MGL 2199 0 1998
84 Mendoza, Shania Mae wf PHI 2191 0 1998
85 Sanchez Ones, Yeny wm CUB 2188 0 1997
86 Terbe, Julianna wf HUN 2187 9 1997
87 Roca Rojas, Ana Flavia wf CUB 2183 0 1996
88 Aakanksha Hagawane IND 2181 0 2000
89 Harshita Guddanti IND 2180 18 2001
90 Kanakova, Natalie CZE 2179 14 1999
91 Zarkovic, Mila wm SRB 2178 0 1996
92 Abdusattorova, Bakhora wf UZB 2177 0 1999
93 Georgescu, Lena wf SUI 2174 9 1999
94 Zhang, Lanlin CHN 2173 22 1999
95 Sieber, Fiona wf GER 2173 6 2000
96 Diakonova, Ekaterina RUS 2172 9 1999
97 Paramzina, Anastasya wf RUS 2172 9 1998
98 Yao, Lan CHN 2167 32 2000
99 Richterova, Natasa wf CZE 2167 18 1996
100 Kovacs, Judit HUN 2167 17 2002

Top 100 Rapid

It was a quiet month in the Rapid ratings with no significant changes.

Rk
Name
Ti.
Fed
Rtg
Gms
B-Year
1 Carlsen, Magnus g NOR 2894 0 1990
2 Nakamura, Hikaru g USA 2839 0 1987
3 Karjakin, Sergey g RUS 2818 0 1990
4 Nepomniachtchi, Ian g RUS 2812 0 1990
5 Dominguez Perez, Leinier g CUB 2803 0 1983
6 Anand, Viswanathan g IND 2802 0 1969
7 Vachier-Lagrave, Maxime g FRA 2795 4 1990
8 Mamedyarov, Shakhriyar g AZE 2791 0 1985
9 Radjabov, Teimour g AZE 2788 0 1987
10 Kramnik, Vladimir g RUS 2778 0 1975
11 Ivanchuk, Vassily g UKR 2771 0 1969
12 Aronian, Levon g ARM 2770 0 1982
13 Grischuk, Alexander g RUS 2767 0 1983
14 Sjugirov, Sanan g RUS 2765 0 1993
15 Wang, Hao g CHN 2763 0 1989
16 Le, Quang Liem g VIE 2761 0 1991
17 So, Wesley g USA 2759 0 1993
18 Gelfand, Boris g ISR 2753 6 1968
19 Caruana, Fabiano g USA 2752 0 1992
20 Giri, Anish g NED 2750 0 1994
21 Kamsky, Gata g USA 2749 8 1974
22 Andreikin, Dmitry g RUS 2743 0 1990
23 Adams, Michael g ENG 2741 0 1971
24 Kryvoruchko, Yuriy g UKR 2740 0 1986
25 Melkumyan, Hrant g ARM 2736 0 1989
26 Wang, Yue g CHN 2733 0 1987
27 Svidler, Peter g RUS 2729 4 1976
28 Rapport, Richard g HUN 2729 0 1996
29 Zhigalko, Sergei g BLR 2728 10 1989
30 Morozevich, Alexander g RUS 2725 0 1977
31 Short, Nigel D g ENG 2723 0 1965
32 Wojtaszek, Radoslaw g POL 2721 0 1987
33 Fedoseev, Vladimir g RUS 2720 0 1995
34 Onischuk, Vladimir g UKR 2720 0 1991
35 Bogdanovich, Stanislav m UKR 2719 0 1993
36 Rublevsky, Sergei g RUS 2718 11 1974
37 Yu, Yangyi g CHN 2716 0 1994
38 Petrosian, Tigran L. g ARM 2715 7 1984
39 Topalov, Veselin g BUL 2715 0 1975
40 Guseinov, Gadir g AZE 2714 0 1986
41 Korobov, Anton g UKR 2714 0 1985
42 Ponomariov, Ruslan g UKR 2712 0 1983
43 Jobava, Baadur g GEO 2710 0 1983
44 Tomashevsky, Evgeny g RUS 2710 0 1987
45 McShane, Luke J g ENG 2709 6 1984
46 Kovalenko, Igor g LAT 2708 0 1988
47 Inarkiev, Ernesto g RUS 2707 6 1985
48 Harikrishna, P. g IND 2706 0 1986
49 Popov, Ivan g RUS 2705 0 1990
50 Ding, Liren g CHN 2704 0 1992
51 Socko, Bartosz g POL 2702 0 1978
52 Artemiev, Vladislav g RUS 2701 0 1998
53 Malakhov, Vladimir g RUS 2701 0 1980
54 Bacrot, Etienne g FRA 2700 0 1983
55 Dubov, Daniil g RUS 2700 0 1996
56 Jakovenko, Dmitry g RUS 2699 0 1983
57 Akopian, Vladimir g ARM 2698 0 1971
58 Savchenko, Boris g RUS 2698 0 1986
59 Navara, David g CZE 2696 9 1985
60 Salem, A.R. Saleh g UAE 2694 0 1993
61 Amonatov, Farrukh g TJK 2693 0 1978
62 Bukavshin, Ivan g RUS 2692 0 1995
63 Leko, Peter g HUN 2692 0 1979
64 Amin, Bassem g EGY 2691 0 1988
65 Dreev, Aleksey g RUS 2690 8 1969
66 Meier, Georg g GER 2689 0 1987
67 Riazantsev, Alexander g RUS 2685 11 1985
68 Nguyen, Ngoc Truong Son g VIE 2685 0 1990
69 Berkes, Ferenc g HUN 2684 0 1985
70 Khairullin, Ildar g RUS 2683 0 1990
71 Georgiev, Kiril g BUL 2681 9 1965
72 Onischuk, Alexander g USA 2681 0 1975
73 Ragger, Markus g AUT 2680 0 1988
74 Bortnyk, Olexandr g UKR 2678 0 1996
75 Salgado Lopez, Ivan g ESP 2678 0 1991
76 Vitiugov, Nikita g RUS 2677 0 1987
77 Fressinet, Laurent g FRA 2676 0 1981
78 Movsesian, Sergei g ARM 2674 9 1978
79 Bartel, Mateusz g POL 2673 0 1985
80 Kasimdzhanov, Rustam g UZB 2673 0 1979
81 Matlakov, Maxim g RUS 2673 0 1991
82 Sargissian, Gabriel g ARM 2673 0 1983
83 Naiditsch, Arkadij g AZE 2672 0 1985
84 Howell, David W L g ENG 2671 6 1990
85 Mamedov, Rauf g AZE 2670 0 1988
86 Motylev, Alexander g RUS 2669 0 1979
87 Khismatullin, Denis g RUS 2668 0 1984
88 Bu, Xiangzhi g CHN 2667 0 1985
89 Najer, Evgeniy g RUS 2667 0 1977
90 Granda Zuniga, Julio E g PER 2665 0 1967
91 Delgado Ramirez, Neuris g PAR 2664 9 1981
92 Ponkratov, Pavel g RUS 2664 9 1988
93 Shirov, Alexei g LAT 2664 0 1972
94 Bauer, Christian g FRA 2662 15 1977
95 Laznicka, Viktor g CZE 2662 0 1988
96 Sadler, Matthew D g ENG 2661 0 1974
97 Nisipeanu, Liviu-Dieter g GER 2659 0 1976
98 Rodshtein, Maxim g ISR 2658 5 1989
99 Swiercz, Dariusz g POL 2658 0 1994
100 Istratescu, Andrei g FRA 2657 9 1975

Top 50 Women Rapid

There were no big changes in the top ranks.

Rk
Name
Ti.
Fed
Rtg
Gms
B-Year
1 Hou, Yifan g CHN 2631 0 1994
2 Lagno, Kateryna g RUS 2594 0 1989
3 Stefanova, Antoaneta g BUL 2565 0 1979
4 Dzagnidze, Nana g GEO 2549 0 1987
5 Ju, Wenjun g CHN 2542 0 1991
6 Kosteniuk, Alexandra g RUS 2519 0 1984
7 Gunina, Valentina g RUS 2512 0 1989
8 Tan, Zhongyi wg CHN 2501 0 1991
9 Ushenina, Anna g UKR 2498 0 1985
10 Khotenashvili, Bela g GEO 2487 0 1988
11 Koneru, Humpy g IND 2486 0 1987
12 Paehtz, Elisabeth m GER 2483 0 1985
13 Zhao, Xue g CHN 2479 0 1985
14 Krush, Irina g USA 2470 0 1983
15 Harika, Dronavalli g IND 2464 0 1991
16 Bodnaruk, Anastasia m RUS 2460 0 1992
17 Pogonina, Natalija wg RUS 2458 0 1985
18 Javakhishvili, Lela m GEO 2457 0 1984
19 Munguntuul, Batkhuyag m MGL 2455 0 1987
20 Hoang, Thanh Trang g HUN 2450 7 1980
21 Lei, Tingjie wg CHN 2446 0 1997
22 Zhu, Chen g QAT 2446 0 1976
23 Huang, Qian wg CHN 2445 0 1986
24 Turova, Irina m RUS 2444 0 1979
25 Guo, Qi m CHN 2442 0 1995
26 Wang, Jue wg CHN 2442 0 1995
27 Shen, Yang m CHN 2439 0 1989
28 Arakhamia-Grant, Ketevan g SCO 2437 0 1968
29 Matnadze, Ana m ESP 2427 0 1983
30 Mkrtchian, Lilit m ARM 2425 0 1982
31 Goryachkina, Aleksandra wg RUS 2424 0 1998
32 Pustovoitova, Daria f RUS 2421 0 1994
33 Batsiashvili, Nino m GEO 2417 0 1987
34 Nechaeva, Marina m RUS 2413 18 1986
35 Socko, Monika g POL 2411 0 1978
36 Gaponenko, Inna m UKR 2406 0 1976
37 Arabidze, Meri m GEO 2404 9 1994
38 Cramling, Pia g SWE 2401 0 1963
39 Houska, Jovanka m ENG 2400 6 1980
40 Peptan, Corina-Isabela m ROU 2397 9 1978
41 Kovanova, Baira wg RUS 2393 9 1987
42 Zhang, Xiaowen wg CHN 2390 0 1989
43 L'Ami, Alina m ROU 2388 0 1985
44 Lomineishvili, Maia m GEO 2388 0 1977
45 Girya, Olga wg RUS 2386 0 1991
46 Khademalsharieh, Sarasadat m IRI 2386 0 1997
47 Zimina, Olga m ITA 2382 0 1982
48 Rajlich, Iweta m POL 2381 0 1981
49 Ambartsumova, Karina wg RUS 2376 0 1989
50 Lujan, Carolina m ARG 2375 9 1985

Top 100 Blitz

Ding Liren maintains his top spot by a hair after Carlsen lost Blitz rating in both Paris and Leuven. Unless a rated blitz event is held in Sinquefield this coming week, possibly for the pairings (hint! hint!), it seems unlikely there will be any real changes until the World Blitz Championship at the end of the year in Qatar.

Rk
Name
Ti.
Fed
Rtg
Gms
B-Year
1 Ding, Liren g CHN 2875 0 1992
2 Carlsen, Magnus g NOR 2873 0 1990
3 Nakamura, Hikaru g USA 2842 0 1987
4 Nepomniachtchi, Ian g RUS 2840 0 1990
5 Aronian, Levon g ARM 2826 0 1982
6 Vachier-Lagrave, Maxime g FRA 2823 0 1990
7 Shkuro, Iuri g UKR 2814 0 1982
8 Caruana, Fabiano g USA 2800 0 1992
9 Karjakin, Sergey g RUS 2800 0 1990
10 Radjabov, Teimour g AZE 2800 0 1987
11 Svidler, Peter g RUS 2795 0 1976
12 Tomashevsky, Evgeny g RUS 2793 0 1987
13 So, Wesley g USA 2791 0 1993
14 Anand, Viswanathan g IND 2790 0 1969
15 Bortnyk, Olexandr g UKR 2784 0 1996
16 Dominguez Perez, Leinier g CUB 2783 0 1983
17 Artemiev, Vladislav g RUS 2781 0 1998
18 Mamedov, Rauf g AZE 2770 0 1988
19 Adams, Michael g ENG 2768 0 1971
20 Wang, Hao g CHN 2768 0 1989
21 Giri, Anish g NED 2766 0 1994
22 Gelfand, Boris g ISR 2765 0 1968
23 Amonatov, Farrukh g TJK 2764 0 1978
24 Grischuk, Alexander g RUS 2761 0 1983
25 Harikrishna, P. g IND 2759 0 1986
26 Ivanchuk, Vassily g UKR 2754 0 1969
27 Navara, David g CZE 2754 0 1985
28 Jobava, Baadur g GEO 2752 0 1983
29 Mamedyarov, Shakhriyar g AZE 2748 0 1985
30 Le, Quang Liem g VIE 2747 0 1991
31 Ponomariov, Ruslan g UKR 2743 0 1983
32 Onischuk, Vladimir g UKR 2740 0 1991
33 Kasimdzhanov, Rustam g UZB 2737 0 1979
34 Dubov, Daniil g RUS 2734 0 1996
35 Kamsky, Gata g USA 2732 14 1974
36 Fedoseev, Vladimir g RUS 2732 0 1995
37 Melkumyan, Hrant g ARM 2727 0 1989
38 Bogdanovich, Stanislav m UKR 2721 0 1993
39 Petrosian, Tigran L. g ARM 2717 14 1984
40 Vitiugov, Nikita g RUS 2717 0 1987
41 Zubov, Alexander g UKR 2715 9 1983
42 Lu, Shanglei g CHN 2713 10 1995
43 Kramnik, Vladimir g RUS 2713 0 1975
44 Yu, Yangyi g CHN 2712 0 1994
45 Topalov, Veselin g BUL 2710 0 1975
46 Andriasian, Zaven g ARM 2709 9 1989
47 Kravtsiv, Martyn g UKR 2708 0 1990
48 Ponkratov, Pavel g RUS 2705 9 1988
49 Swiercz, Dariusz g POL 2705 0 1994
50 Berkes, Ferenc g HUN 2701 0 1985
51 Zhigalko, Sergei g BLR 2699 0 1989
52 Iturrizaga Bonelli, Eduardo g VEN 2698 0 1989
53 Matlakov, Maxim g RUS 2698 0 1991
54 Kovalenko, Igor g LAT 2697 0 1988
55 Leko, Peter g HUN 2697 0 1979
56 Megaranto, Susanto g INA 2697 0 1987
57 Bachmann, Axel g PAR 2696 14 1989
58 Adly, Ahmed g EGY 2696 9 1987
59 Bruzon Batista, Lazaro g CUB 2692 0 1982
60 Jones, Gawain C B g ENG 2689 0 1987
61 Riazantsev, Alexander g RUS 2686 0 1985
62 Laznicka, Viktor g CZE 2684 0 1988
63 Wang, Yue g CHN 2684 0 1987
64 Duda, Jan-Krzysztof g POL 2680 0 1998
65 Guseinov, Gadir g AZE 2676 0 1986
66 Hou, Yifan g CHN 2676 0 1994
67 Andreikin, Dmitry g RUS 2675 0 1990
68 Granda Zuniga, Julio E g PER 2673 0 1967
69 Safarli, Eltaj g AZE 2673 0 1992
70 Malakhov, Vladimir g RUS 2672 0 1980
71 Moiseenko, Alexander g UKR 2672 0 1980
72 Morozevich, Alexander g RUS 2671 0 1977
73 Stevic, Hrvoje g CRO 2671 0 1980
74 Amin, Bassem g EGY 2670 0 1988
75 Sjugirov, Sanan g RUS 2669 0 1993
76 Ragger, Markus g AUT 2667 0 1988
77 Korobov, Anton g UKR 2666 0 1985
78 Predojevic, Borki g BIH 2666 0 1987
79 Wojtaszek, Radoslaw g POL 2666 0 1987
80 Meier, Georg g GER 2664 0 1987
81 Saric, Ivan g CRO 2664 0 1990
82 Khairullin, Ildar g RUS 2663 0 1990
83 Vallejo Pons, Francisco g ESP 2663 0 1982
84 Alekseev, Evgeny g RUS 2662 0 1985
85 Grachev, Boris g RUS 2662 0 1986
86 Cheparinov, Ivan g BUL 2661 0 1986
87 Bocharov, Dmitry g RUS 2660 0 1982
88 Bosiocic, Marin g CRO 2659 0 1988
89 Eljanov, Pavel g UKR 2659 0 1983
90 Socko, Bartosz g POL 2658 0 1978
91 Ramirez, Alejandro g USA 2657 0 1988
92 Akobian, Varuzhan g USA 2655 14 1983
93 Akopian, Vladimir g ARM 2654 0 1971
94 Brunello, Sabino g ITA 2652 0 1989
95 Gasanov, Eldar g UKR 2652 0 1982
96 Lenderman, Aleksandr g USA 2650 0 1989
97 Onischuk, Alexander g USA 2650 0 1975
98 Nguyen, Ngoc Truong Son g VIE 2648 0 1990
99 Cvitan, Ognjen g CRO 2647 0 1961
100 Naiditsch, Arkadij g AZE 2647 0 1985

Top 50 Women Blitz

German IM Elizabeth Paehtz had a gain of 82 Elo this month after crushing the German Women Blitz Championship with the monster score of 37.5/38. Fantastisch!

Rk
Name
Ti.
Fed
Rtg
Gms
B-Year
1 Hou, Yifan g CHN 2676 0 1994
2 Lagno, Kateryna g RUS 2641 0 1989
3 Gunina, Valentina g RUS 2611 0 1989
4 Stefanova, Antoaneta g BUL 2582 0 1979
5 Kosteniuk, Alexandra g RUS 2576 0 1984
6 Ju, Wenjun g CHN 2571 0 1991
7 Tan, Zhongyi wg CHN 2552 0 1991
8 Paehtz, Elisabeth m GER 2538 38 1985
9 Zhao, Xue g CHN 2526 0 1985
10 Koneru, Humpy g IND 2502 0 1987
11 Harika, Dronavalli g IND 2501 0 1991
12 Dzagnidze, Nana g GEO 2500 0 1987
13 Ushenina, Anna g UKR 2485 0 1985
14 Krush, Irina g USA 2464 0 1983
15 Wang, Jue wg CHN 2457 0 1995
16 Matnadze, Ana m ESP 2452 0 1983
17 Arabidze, Meri m GEO 2442 9 1994
18 Socko, Monika g POL 2438 0 1978
19 Lei, Tingjie wg CHN 2434 0 1997
20 Rogule, Laura wg LAT 2426 11 1988
21 Gaponenko, Inna m UKR 2414 21 1976
22 Bodnaruk, Anastasia m RUS 2407 0 1992
23 Sukandar, Irine Kharisma m INA 2406 0 1992
24 Pogonina, Natalija wg RUS 2402 0 1985
25 Zhai, Mo wg CHN 2398 0 1996
26 Peptan, Corina-Isabela m ROU 2397 0 1978
27 Li, Ruofan m SIN 2394 0 1978
28 Lomineishvili, Maia m GEO 2394 0 1977
29 Zhukova, Natalia g UKR 2394 0 1979
30 Khotenashvili, Bela g GEO 2387 0 1988
31 Nemcova, Katerina wg USA 2384 0 1990
32 Pham, Le Thao Nguyen m VIE 2382 0 1987
33 Javakhishvili, Lela m GEO 2380 0 1984
34 Zhu, Chen g QAT 2379 0 1976
35 Mkrtchian, Lilit m ARM 2374 0 1982
36 Charkhalashvili, Inga wg GEO 2372 0 1983
37 Huang, Qian wg CHN 2372 0 1986
38 Batsiashvili, Nino m GEO 2370 0 1987
39 Pustovoitova, Daria f RUS 2369 0 1994
40 Linares Napoles, Oleiny wg CUB 2366 0 1983
41 Lubbe, Melanie wg GER 2364 38 1990
42 Khukhashvili, Sopiko m GEO 2364 0 1985
43 Ubiennykh, Ekaterina wm RUS 2364 0 1983
44 Padmini, Rout m IND 2362 0 1994
45 Abdumalik, Zhansaya wg KAZ 2361 0 2000
46 Cori T., Deysi wg PER 2358 0 1993
47 Babiy, Olga wg UKR 2354 0 1989
48 Kochetkova, Julia wg SVK 2354 0 1981
49 Heinemann, Josefine wm GER 2349 38 1998
50 Khademalsharieh, Sarasadat m IRI 2349 9 1997

Source: FIDE



Born in the US, he grew up in Paris, France, where he completed his Baccalaureat, and after college moved to Rio de Janeiro, Brazil. He had a peak rating of 2240 FIDE, and was a key designer of Chess Assistant 6. In 2010 he joined the ChessBase family as an editor and writer at ChessBase News. He is also a passionate photographer with work appearing in numerous publications.
Discussion and Feedback Join the public discussion or submit your feedback to the editors


Discuss

Rules for reader comments

 
 

Not registered yet? Register

imdvb_8793 imdvb_8793 8/5/2016 06:06
"As I mentioned before we have been dealing in the last decades with the changes of globalization. In big countries like India and China chess has grown enormously and that is something which goes slowly so more or less in a trendy way."

Yeah, that's true, China and India did, indeed, start to develop more seriously already at the beginning of the 1980's... China already had a solid result in the 1985 World Cup in Luzern, and India had Anand rising to fame precisely around 1985-1987... I mean, it still doesn't seem like enough to immediately affect the top of the list (though, if Asian players entering the list for the first time were, in fact, as they are today, on average, stronger than the mostly European/American roster of newcomers up until that point, that could, indeed, add some points to the list), but it's definitely an interesting idea I hadn't considered.

"2) With such small numbers we have to look at the consequences of deaths or retirements compared with new blood. Only Fischers retirement already had a huge influence on the average rating of the top 40."

This is something I, too, have considered, at times. The early 1970's had the deaths of Stein and Keres as well, both rated around 2600 (and other pretty good players, too, like, for example, Alexander Zaitsev), as well as Mecking's retirement, later... This is the strongest argument to be made, I think.

"From 1984 the start of an enormous battle between 2 giants started. They didn't only bring out the best in each other but also other topplayers learned from them. So in that sense it is well possible that some randomness was important."

Again, I doubt the effect of this could have been felt within less than two years, but, yeah, who knows?! Still, why was there not the same effect after the Fischer-Spassky Match of the Century boom? (Which is certain, and well documented.) I guess that goes back to the '70's deaths/retirements thing, perhaps... like I said, that still sounds like the strongest argument to me.

"I also want to make sure we don't overestimate the value in the pre-computertime of chess evolution."

OK, but you have to admit the Informants (which also brought the first opening encyclopedias) were a pretty serious development!... Also, if not for the deaths/retirements of the 1970's taking so many points out of the list at the top, it stands to reason that the average would, in fact, have grown even then, since even WITH them the average didn't drop, so the chess boom of the early 1970's still looks like it's a thing...
(You can probably calculate the top 40 average adjusted for deaths and such, to see how much they really impacted things - I should do that some time!...)

Yup, you definitely make some good points! In particular, the Asian growth and '70's deaths arguments. Food for thought...

Cheers! I'm out...
brabo_hf brabo_hf 8/5/2016 05:08
"So why does that 40-player average rating start to increase at all, and why in 1985, and not sooner - or later? And why at such a steady pace? Is there a simple explanation for this that I'm missing? Or can this really be COMPLETELY RANDOM?"
Well honestly I have never investigated this stuff but there are some possible reasons which I can just state out of my head.
1) As I mentioned before we have been dealing in the last decades with the changes of globalization. In big countries like India and China chess has grown enormously and that is something which goes slowly so more or less in a trendy way.
2) With such small numbers we have to look at the consequences of deaths or retirements compared with new blood. Only Fischers retirement already had a huge influence on the average rating of the top 40. From 1984 the start of an enormous battle between 2 giants started. They didn't only bring out the best in each other but also other topplayers learned from them. So in that sense it is well possible that some randomness was important.
3) I also want to make sure we don't overestimate the value in the pre-computertime of chess evolution. I really have my doubts if champions like Euwe, Smyslov, Tal were better than Lasker, Capablanca and Alekhine. That is also more or less what the IPR files from the professors tell us. If you assume top players keep learning then we really should already see a few decades later averagely much higher IPRs which isn't the case.
4) Finally it is a fact that ratings linked with playing strength (not actual/ absolute ones) is something flexible . As mentioned in one of my previous posts, for the highest rated players there is no way to exclude an existence of inflation/ deflation.

Anyway above are opinions based on some arguments so definitely of a lower level of seriousness than the inflation.
imdvb_8793 imdvb_8793 8/5/2016 04:59
"I won't deny there are differences but there are however much more similarities than you think there are.
1) We never know the speed of any fish even after doing thousands of measurements.
2) We never know the absolute length of the cable even after doing thousands of measurements
3) We are relying on a watch which brings its own level of inaccuracy. If there is very little difference in length between the cables then the given method will not work to define which one is longer."

Of course, I realize all of that :) - I wasn't saying there was NO margin for error in that example, far from it! Just that it was (at least in my opinion) smaller, and that the parameters used there were better defined and more accurately/confidently measurable than in the case of trying to determine playing/move strength using ELO ratings and chess engines (again, at this time - later, with technological advances, that will very likely change.)
imdvb_8793 imdvb_8793 8/5/2016 04:54
"Also here you are talking about actual playing strength. I think better is absolute playing strength as definition. Again we use delta values instead of absolute values. It is just a matter of a different definition of the playing strength which allows us to measure it."

I understand - which is why I acknowledge that the study is valid and 100% sound. But, just as you like to warn against drawing conclusions about inflation based on limited samples and questionable statistics, so, too, I think it is important for people to realize that the Regan/Haworth study only actually draws conclusions about the inflation (or lack thereof) of the ratings themselves, and cannot actually help us compare the absolute playing strength (you're right, of course - this sounds better) of the players in question, but only about their strength as perceived by us at this moment, through the use of the aforementioned two clearly imperfect (and potentially, for all we know, even just incorrect, though I'm not saying that's also likely, nor do I believe it) measuring systems (ELO, engine eval.)
imdvb_8793 imdvb_8793 8/5/2016 04:43
"However we also bypass that problem by looking at a bunch of players rated at the same level. Some will be rated too low and some too high but the average will already be much less arbitrary."

Of course. But probably still arbitrary enough that we can't make definitive pronouncements about the existence or not of inflation...

"How is the rating of the average rated player evolved?"

Yeah, I've also been thinking about that avenue, of course, but that's complicated too much, in my opinion, by the lowering of the threshold under 2200 and the presence of woman players with much lower average ratings and who, at least in the beginning, since few, if any, played in men's tournaments, kind of made up a rating pool of their own, mostly independent of the male rating pool... If you take care of those problems, somehow, then you might be able to draw some conclusions for 1971-1999, or whenever it is the threshold was lowered (I don't know the year, off the top of my head), but, even then, I have doubts about how relevant it would be - for one, too few players were rated in the beginning, so the average initial rating for new players on the first few lists is probably considerably higher than it is for later lists, which should affect the average player's rating in unpredictable ways. There are other things, too... I don't know if such an analysis can be performed without significant bias, and I don't know if there'd be a way to calculate an adjustment.

"I find it wrong to make any general statements of inflation/ deflation by just looking at the topplayers as ratings are much more than just them."

Agreed - that's not what I was asking, though. :) I was just asking for your opinion on that particular phenomenon, of the consistency, and then sudden and steady inflation beginning in, for some reason, 1985, of the rating average of that particular section of the rating list (11th-50th.) No opinion at all? (I'm not asking for an actual verdict, based on a sample size of 40, obviously - just asking if you think this phenomenon is normal, consistent with your view that there is no inflation, or not, or just completely irrelevant to the discussion. I'm guessing, based on your last post, that your answer is the latter, but I just want to clarify if that's the case.)
brabo_hf brabo_hf 8/5/2016 04:41
"difference between this and the example with the cables and the fish"
I won't deny there are differences but there are however much more similarities than you think there are.
1) We never know the speed of any fish even after doing thousands of measurements.
2) We never know the absolute length of the cable even after doing thousands of measurements
3) We are relying on a watch which brings its own level of inaccuracy. If there is very little difference in length between the cables then the given method will not work to define which one is longer.

At the end of the measurements we will have a time average for cable 1 and a time average of cable 2. With some calculations we can define how much wrong the time averages can be for getting 99,99% certainty.
Suppose you get 56 seconds for cable 1 and 58 seconds for cable 2 and the deviation is 2 seconds then we can't make a conclusive verdict of which cable is the longest. Most likely cable 2 but we have a low probability.
brabo_hf brabo_hf 8/5/2016 04:33
"the ACTUAL playing strength of any players, or the ACTUAL strength of any moves."
I understand you mean how it is compared with the absolute truth of the position. I think here you go a bit off track. Today we have no idea which rating is linked to playing the best possible chess. Take note that I don't talk about playing perfect chess as there are very often many moves which are not changing the truth of a position. Still ratings are used to compare the playing strength between players without knowing what the absolute truth is of the players or the strength of their moves.
Inflation is still linked with comparing players without knowing what the absolute truth is but with the difference that we are dealing with players from a different time.

"measuring systems that, themselves, have no scientifically proven correlation to playing strength (in the case of ELO ratings, because they use some fairly arbitrary constants and variables"
Also here you are talking about actual playing strength. I think better is absolute playing strength as definition. Again we use delta values instead of absolute values. It is just a matter of a different definition of the playing strength which allows us to measure it.
brabo_hf brabo_hf 8/5/2016 04:12
"Oh, believe me, I understand that the ratings system itself is fairly arbitrary itself."
Today you get already a rating based on 20 games. The results in the study were based on 150.000 games. It can be proven scientifically but I am sure that the study was much more accurate than a rating of some individual.

However we also bypass that problem by looking at a bunch of players rated at the same level. Some will be rated too low and some too high but the average will already be much less arbitrary.

It also explains why we can't make any strong verdict of an inflation compared with 2700 players before 1990. There weren't enough to get sufficient accuracy in our results. So we can't state for sure there was no inflation at the toplevels before 1990.

There are however 2 remarks I want to give on that.
1) Ratings are communicating vessels so if there was no inflation at the level of 2600 then it is really unlikely there is inflation at a higher level. However I do remember that Kasparov asked for exceptions on the ratingcalculations. When he won a tournament but anyway underperformed then he could not lose rating. That was later abolished but it could have created for some top-players some period of very local inflation.
2) Inflation for me has always been something much broader than just the topplayers. My rating was more or less stable the last 20 years. What about my playing strength? Can I assume there was no inflation/deflation? How is the rating of the average rated player evolved? I find it wrong to make any general statements of inflation/ deflation by just looking at the topplayers as ratings are much more than just them.
imdvb_8793 imdvb_8793 8/5/2016 02:57
However, I do have another quick inflation-related question for you, if you still have any interest (and I DO realize there's a small risk this will, nevertheless, lead to another drawn out debate, though I will certainly do my best to keep that from happening, as far as it's within my power to do so - but I'm just too curious what you think about it to not ask): given the analysis Rod Edwards performed, however small his sample (40 players per list), how do you explain the fact that, between 1971 and 1985, the average of said 40 players (11th through 50th place in the world) remained almost constant, and only in 1985 did it start to increase - and this, in a linear, mathematically describable fashion? (And I know that, since 2006, when Edwards published his first inflation-related article online, this "inflation", or whatever it is, has greatly deviated from that linear equation - it still hasn't gone down, but it's grown in less predictable fashion -, but let's ignore that and imagine we're in 2006, but knowing all that we know today about Regan and Haworth's study, and everything else, except for the evolution of rating lists between then and now!)

Since the Regan/Haworth study suggests this should simply be the result of an increase in average playing strength for that (definitely not random but, also, I think, not biased, either)

subset of players (or what would be the right way to put it... 'rating subgroup'? 'zone'? - I have no idea), then why was there no such increase, at all, between 1971 and 1985? Did chess really not evolve at all during that decade and a half? I seriously doubt that. A study of the numbers of players on the rating lists also seems to suggest no particular answer to this question: the percentage of new players per list is, it seems to me, even greater for that period than for, say, 1982-1989, on average. (Naturally.) The rating threshold wasn't moved below 2200 until some time late in the 1990's, I believe, so that's no answer. So why does that 40-player average rating start to increase at all, and why in 1985, and not sooner - or later? And why at such a steady pace? Is there a simple explanation for this that I'm missing? Or can this really be COMPLETELY RANDOM? I doubt it, because I bet if you analyze the top, let's say, 200 players on each list (either excluding the top 10, or not), instead, you'll still come up with similar findings - though, based on some quick studies I've performed during our little debate, I believe the "inflation" rate you would come up with, while still fairly consistent, would be somewhat smaller than the one obtained by Edwards, based on his 40-player sample. (Or is that also irrelevant?) I've not done this, though, obviously, so my question to you remains only in relation to Edwards' 40-player sample.

In 1985 computers were still basically useless for preparation, and were definitely not in wide use for such until, again, the late 1990's, so that can't be it. I know of no particular spike in theoretical growth in the game of chess around 1980-1985, either, compared to the early 1970's (if anything, going by what Kasparov said in his first Modern Chess book, you'd think there was a bigger such "revolution" closer to the early part of the 1970's, due to Fischer's exploits, the appearance of Chess Informant, and so on.) Maybe Kasparov's appearance on the scene, but could that really be felt quite so quickly - the very next year after he became champion? I don't know... So what is the explanation, then? I can't find a satisfactory one that doesn't involve the word 'inflation'... :)
imdvb_8793 imdvb_8793 8/5/2016 02:56
Part 2:

Therefore, while the study does prove that there is no inflation in the sense that players of the same strength - AS WE UNDERSTAND IT RIGHT NOW, by using the two unproven measuring units I

just talked about - as 2600's (or 2500's, or whatever) from 1973, or whatever, do NOT have more points now than they did then, on average (and, again, the study does not include the very top of the rating lists of the early 1970's and 1980's, so that part remains unproven even in this sense) - which is, of course, still a big deal, and, unless I'm being selective with my memory here (I could check, but I'm too lazy), I believe I acknowledged this from the very beginning -, it still doesn't prove that there is no inflation in the sense of players of the same ACTUAL playing strength between the two time periods in question making equally strong moves, on average, because we simply have no scientifically sound way of measuring (and being confident that we're right) either the ACTUAL playing strength of any players, or the ACTUAL strength of any moves. (Again, except for tablebase situations.)

But I agree that this is nitpicking (actually, I think I stated this at an earlier point as well, even before we went into details) and, overall, a minor problem, as the study (if I remember correctly) never claims to be able to define real playing strength. I think we've reached a point where we can both be satisfied that we're right about the part we wanted to be right about - you, about the theory involved being 100% sound, and about the study proving there is no inflation in the only sense in which it COULD prove that (see above), given the current state of technology and the parameters being used, and myself, about the fact that, while the study is very valuable/useful and even telling, it is still not ABSOLUTE proof of the relation between the ACTUAL playing strengths of the players in question, past and present, because the parameters it compares (the ELO ratings themselves and the engine's output) are nothing more than (mostly) empirical measuring units of properties (playing strength / move strength) which we do not yet fully understand, because we don't yet fully understand the game they relate to, chess.

Does that sound fair to you, too? Or am I being naive/vain in thinking I was right about any of the above? (Specifically, the part I claim to be right about.) I never like to be anything but

100% as objective as I can be, about anything, so this is actually quite important to me - which is why I'd appreciate another honest, well considered answer (which is exactly what you've

been giving me for the past few days - I was wrong to assume you were a troll, in the heat of the moment, and I do apologize for saying/suggesting that!) Was I rambling on, or did I have a point about what I said in the last paragraph, however abstract and, in your opinion, at least, nitpicky, that point may be? (If we can agree on this, or if you can convince me that I AM, in fact, wrong about that part as well, if you think that's the case, then I will have no further questions in this matter, and we can each go back completely to our lives :), at last...)
imdvb_8793 imdvb_8793 8/5/2016 02:56
Part 1:

Beautifully explained! I think I understand the whole thing now... There's still an asterisk I'd like to attach to the whole thing (see below), but everything you explained makes perfect sense to me now.

"If the input would have been totally randomized and lacking any sort of accuracy then I would never have been able to reproduce the tests with the same results over and over again and I

would never have seen any correlation between the result and the rating."

So you're saying this proves this correlation is statistically unlikely to be random. Fair enough. Sounds solid to me. :)

"B.t.w. for building the ratings as a an indication of the playing strength there is neither any knowledge available of the ultimate truth of the positions. Still ratings are everywhere

accepted to categorize people conform playing strength instead of age/ name/ ..."

Oh, believe me, I understand that the ratings system itself is fairly arbitrary itself. :) (More on that below) I accept it as is, for the purposes of this discussion, and other things. I'm not saying it's fair or that it has any real correlation to playing strength itself. It's a fun thing to speculate on and follow, that's all...

"As we have proven there is a link between engine evaluations and a playing strength"

I see how you got here, and I agree that there's proof there is a link between the two, due to of the consistency of the results and the laws of probability, and such. Which helps me

understand better what my problem was in the first place - the fact that both of these parameters, ELO ratings AND engine evaluations (however you choose to use them, with the 'top 3 choices'

method or the numerical evaluation method, or whatever), are, in fact, measuring systems that, themselves, have no scientifically proven correlation to playing strength (in the case of ELO

ratings, because they use some fairly arbitrary constants and variables and whatnot, and, in the case of engine evaluations, because of the reasons I've already given far too often to repeat).

Now, don't get mad - I understand now the reason why this is, in a way, beside the point - at least for the purposes of the study itself. Like I said, you've explained everything perfectly (Unless, of course, I'm still misunderstanding something and am not aware of it.) You're not trying to prove that there is a correlation between engine evaluations and ACTUAL playing strength (which, as I was assuming, unless my logic still fails me, somehow, remains impossible to prove at the moment), but rather that there is a correlation between engine evaluations and playing strength AS EXPRESSED BY ELO RATINGS. Which you then use in your study about inflation. Like I said: it makes perfect sense. :) So, yes, in this sense, you're right, the study DOES prove what it sets out to prove, within the parameters it has at its disposal to begin with.

But, then, you have to at least admit that there is a clear (and, I believe, crucial) difference between this and the example with the cables and the fish - because there the measuring units involved in the study were scientifically indisputable, or whatever you want to call it, being the standard units for measuring time/distance/speed, which are physical properties of the objects under scrutiny we fully understand (well, more or less - at least enough to be able to measure them properly), whereas here, in this study, the property being measured (playing strength) is not something we fully understand (if at all), and, as a consequence, so are the measuring units (ELO ratings / engine evaluations.)
brabo_hf brabo_hf 8/5/2016 12:52
"I'm asking how you know that ANY of the engine's evaluations, IN ANY POSITION, have any relation to the ultimate reality of the position"
The evaluations of an engine are not about who plays the most beautiful chess, who is playing the most aggressive moves, who is playing the fastest,... The evaluations of the engine have a direct link with the moves an engine plays which can be verified/ proven by checking the algorithm. We can define very accurately the playing strength of the engine within a group by a huge number of produced moves. As the engine evaluations are directly linked to the produced moves and the produced moves are directly linked to a playing strength, we do have proof that the engine evaluations do relate with a playing strength.

Our study was about defining the inflation of ratings to different playing strengths. As we have proven there is a link between engine evaluations and a playing strength there is no need to find another link between engine evaluations and ultimate truth of a position.

B.t.w. for building the ratings as a an indication of the playing strength there is neither any knowledge available of the ultimate truth of the positions. Still ratings are everywhere accepted to categorize people conform playing strength instead of age/ name/ ...
brabo_hf brabo_hf 8/5/2016 10:09
Last night I was thinking of an example which could explain the principle in an easy way.

You have 2 cables laying in the sea and for whatever reason you want to know which one is the longest.
You see can see the beginning and the end point of the cables but you can't see which one is the longest. Neither is there any equipment you can use to measure with sufficient accuracy the length.
However you do see that a family of fishes is attracted to the cables and likes to swim along it. You have a watch and can measure the time one random selected fish needs to swim along a cable. Only problem is that all fishes swim with a different speed so you can't compare 1 measurement with another one.
Still it is possible to find out based on measurements which cable is very likely the longest.
What you do is to make many hundred/ thousands of measurements of the time a random selected fish needs to swim along cable 1 and you do this over again for cable 2. You define the average of the results of cable 1 and compare this with the average of the results of cable 2.

We bypassed that a measurement of 1 fish can't say us anything about the length of a cable and defined a parameter linked to the family which is along both cables active. The characteristics of a family is something fixed and can be defined by measuring a sufficient amount of times the speed of randomly chosen fishes.
brabo_hf brabo_hf 8/5/2016 09:46
Once you have chosen the parameter on which you want to use for comparison, you must first find out if this parameter has a strong correlation with the ratings or not. That is something easy to find out as you need to check 2 things.
1) you first check if for a specific rating the result of the parameter based on x amount of random selected moves is always the same (within the boundaries of the standard deviation).
E.g. you take 100 random selected games of unknown 2600 rated players and the result of the parameter gives 0,5 with standard deviation 0,01 and rule of the deviation is 99,999999999%. Then if we redo the test 10 times so taking each time another 100 random selected games of unknown 2600 rated players, the result of the parameter should each time be within 0,49 and 0,51.
2) However point 1 alone doesn't make the parameter yet useful for proving a correlation with a rating. Assume you use a very stupid engine which gives only the evaluation 0,00 then you will be compliant for point 1 but each rating will give you that same result. What we need to see, is that there is a direct link between the rating and the result of the parameter.
Example: 2000 rating gives 2,5 ; 2100 rating gives 1,9 ; 2200 rating gives 1,4; 2300 rating gives 1,0; 2400 rating gives 0,7; 2500 rating gives 0,5; 2600 rating gives 0,4
With a standard deviation of 0,01 on top of each result, we know with a certainty extremely close to 100% that the parameter chosen is directly linked to a rating.

So I had not any proof from the beginning if there was any correlation between the parameter and the ratings. Neither did I have any proof at the start that the accuracy of the evaluations would be sufficient. If the input would have been totally randomized and lacking any sort of accuracy then I would never have been able to reproduce the tests with the same results over and over again and I would never have seen any correlation between the result and the rating.
brabo_hf brabo_hf 8/5/2016 09:11
"am I understanding your explanation right?"
Yes you do.
But it still has to be proven that those deviation values mean anything to begin with, no?
Correct.
Are we just ASSUMING the engine's evaluations are at least somewhat accurate some of the time?
No, the results must proof that the engine evaluations are giving some accuracy

Before going to the heart of the topic, I first want to point out that you concentrate too much on the values I gave. The average evaluation deviation is defined in pawn units but another study used a very different parameter. There they defined a parameter which was linked to the % of mistakes in a game. A move was considered a mistake if it was not popping up in the 3 top choices of an engine. So instead of values like 0,5 they got 20% for a 2500 rating and 25% for a 2400 rating. Deviation in pawnunits, % of mistakes, % of exact matches,.... are all parameters which are interesting for a study.
imdvb_8793 imdvb_8793 8/5/2016 06:56
I realize I probably shouldn't have called them "deviations" - that's a little confusing in a discussion that also includes standard deviation. But that's still what those values are, and that's the term I thought best described them. They're deviations from the score of the engine's top choice, as opposed to deviations from the mean, so the distinction is clear. I just wish there was a less confusing term I could use, but I still can't think of one, even now, after having slept for a little bit. But I think it won't be a problem for you - you get what I was talking about...
imdvb_8793 imdvb_8793 8/4/2016 11:04
OK - be back tomorrow! Gotta get some sleep...
imdvb_8793 imdvb_8793 8/4/2016 09:58
Part 3:

This is the part I can't get past. How can you use something that can be anywhere from 0% to 100% accurate, given what we actually KNOW about it at the moment, to "prove" (statistically) anything about anything else? So, to quote myself, from the paragraph above:

"Are we just ASSUMING the engine's evaluations are at least somewhat accurate some of the time?"

I refuse to believe you're saying we don't need to know this at all, somehow! Because, otherwise, why can't we just assign random numerical values to each move we're "analyzing", and draw our conclusions based on that? Wouldn't that be just as likely to be relevant, if the engine's evaluation accuracy is completely irrelevant to the results of the study? Why perform the engine analysis at all, then? No, clearly you either have to assume a certain level of accuracy for the engine's evaluations, or have some kind of proof of it to begin with. That's the part I want you to clarify for me: are you making an assumption, or do you have any scientific evidence that if, say, the engine evaluates a move as -0.35 instead of +3.42, that means anything at all? (Other than that that's what the engine thinks of the position.)

Having read all of your posts so far, I think you ARE just assuming that this means something (because the engine is strong, which, again, in my opinion, is completely irrelevant, as long as it's not perfect - see above for details), and, due to your background and years of working with engines, it's so hard for you to wrap your head around the notion that it might not mean anything, that whenever I say or suggest this, you just figure either that I'm talking nonsense, or that I'm missing some more important point; if so, then THAT'S the point I want you to explain to me. Not the statistical stuff. I want you to explain to me how you KNOW - scientifically speaking, not empirically, through having played chess and used engines yourself - that there's any REAL qualitative difference between a move evaluated as, let's be a bit less extreme, +0.54, and a move evaluated as -0.31, or, alternatively, if this doesn't matter AT ALL, then THAT is what I want you to explain to me - WHY it doesn't matter if there's a difference or not. And remember, these are just examples. I'm asking how you know that ANY of the engine's evaluations, IN ANY POSITION, have any relation to the ultimate reality of the position...
imdvb_8793 imdvb_8793 8/4/2016 09:58
Part 2:

For example, say a move is 0.4 points away from that engine's top choice in one position, and another move is also 0.4 points away in a completely different position; let's say the top choice of the engine is a perpetual in both positions, for simplicity's sake. Now, since the two moves that are 0.4 points away from the top choice lead to, in the engine's opinion, positions 0.4 points below equality, but positions that the engine (obviously, given the necessity for a score other than 0.00 or #X - mate in X) can't calculate to the end, and is merely evaluating based on the resulting positions at the end of however far it CAN calculate (a.k.a. its current depth), there's every chance that, once an engine comes along that CAN calculate far enough to draw definitive, mathematical conclusions about the strength of both those moves, in those given positions, it might conclude that one of them is, in fact, winning for white, by force, whereas the other is losing, also by force (both, in X/Y amount of moves), despite the earlier, less advanced engine thinking they were equally inferior to its top choice. Therefore, the fact that, right now - when our engines are not advanced enough yet to calculate those two different moves (that don't even have to be in different positions, as I decided to formulate the example - they can even be in the same position) to either mate or a forced draw (with perfect play by both sides), whichever may be the case -, this or that (or even EVERY) engine gives a -0.40 evaluation for both of those moves, and a 0.00 evaluation for its top choice in both cases, really does mean nothing, and proves nothing about the strength of those moves.
imdvb_8793 imdvb_8793 8/4/2016 09:57
Part 1:

I won't lie, that's a good explanation, and I think I understand your point of view better now. I'm not sure, but I think so. (And thank you for taking the time to explain! I do appreciate it, belive me, despite what you might think after reading the rest of what I have to say, below.) Nevertheless, I still don't see how any of this solves my main complaint/question:

"That means that we know for 99,7% sure that moves played by a bunch of 2600 elo players will deviate averagely between 0,39 and 0,41 points from what the top choice of the engine tells us."

So, in short, you're saying that we conclude from tests that 2600's (in, say, 2016) will deviate this much, on average (0.4, or whatever), and then we compare this to how other rating groups (in this time period, as well as earlier time periods) deviate compared to the same engine's evaluations, and we draw conclusions about, say, inflation, based on that, without needing to know how accurate the engine's evaluations actually are - am I understanding your explanation right? But it still has to be proven that those deviation values mean anything to begin with, no? Are we just ASSUMING the engine's evaluations are at least somewhat accurate some of the time? Because we really don't know that. At all. We think we do, because engines are better than us at chess, and we tend to think whoever is smarter than us is usually right about stuff, but that's a fallacy.

I mean, so what if 2600's average an 0.4 deviation and 2500's average an 0.5 deviation (random examples)? Those are still arbitrary, pawn unit values, that, for all we know, could very well mean nothing at all. So what if, say, 2600's in the 1970's were deviating, on average, the same number as 2600's from today? That just means their move choices, compared to the evaluating engine's move choices, were, on average, at an equal numerical distance, in pawn units, from said engine's choices, as those of present day 2600's. That doesn't ACTUALLY MEAN their PLAYING STRENGTH, or average move quality, or whatever, is the same. Because, as I've been saying, those pawn units (which make up the, for example, 0.4 deviation) are abstract and, really, could well mean absolutely nothing. (Especially since the games analyzed contain two completely different sets of positions being evaluated, but that's just a sidenote, and not the main issue.) There's zero evidence (as far as I know, and as far as you've explained up until now) that the two groups having the same deviation means in any way that they play at roughly the same strength - or even that they play at EVEN REMOTELY SIMILAR strength. Because (again I come to this), since chess is not solved, we just have no way of knowing that those pawn-value measuring units have anything to do with the actual quality of moves. Because we don't, in fact, know that the engine we used (or, indeed, any other engine) can properly evaluate any position, ever, or assess the strength of any move, ever, with any degree of accuracy. We can't even be 50% confident of that. Because this has never been proven scientifically.
brabo_hf brabo_hf 8/4/2016 08:16
About the certainty calculation I also still want to say something without going into details about which formulas exactly are used.
If you have defined an average figure based on x amount of figures then it is statistical perfectly possible to define what the average figure would be within the standard deviation based on an unlimited amount of figures. For that we need to look at how the x amount of figures are behaving compared with the average figure. Weaker engines will most likely show in the x amount of figures compared to the average much more randomness.
brabo_hf brabo_hf 8/4/2016 07:59
I guess after my last post still some questions will remain so I try to explain it with a simple example.

Assume you do a test based on 10 games. You get as result 0,5 with 0,1 standard deviation and you used as rule 90%.
Now if we would redo the test an unlimited amount of times then our result should be between 0,4 and 0,6 in 9/10 cases averagely.

So in the end we were aware that we are not able to know the accuracy of the evaluation from the engine of one move and replaced this by the accuracy of the average deviation over many (hundred-) thousand moves. So we bypassed our unknown element and defined a new parameter which we are able to define the accuracy of. As this parameter is popping up for each ratingsegment and each period of time , we were able to make comparisons.
brabo_hf brabo_hf 8/4/2016 07:58
Before explaining the principle of a standard deviation, I want to make a disclaimer that the values I use in my examples are not the ones from the study. So they are purely used for instructive purpose. The real values should be looked up in the paper.

When I state something is 0,5 with a standard deviation of 0,01 , we first must need to explain which rule we use for the standard deviation. There are different rules for that see https://en.wikipedia.org/wiki/68%E2%80%9395%E2%80%9399.7_rule

Assume we use rule 3. That means that we know for 99,7% sure that moves played by a bunch of 2600 elo players will deviate averagely between 0,39 and 0,41 points from what the top choice of the engine tells us.

You are not satisfied with 99,7% then we need to use a larger deviation. How large, depends on what you consider a sufficient degree of certainty. We can take a % which makes even winning euro millions easier or even random selecting 1 specific atom in the universe easier.

Anyway the certainty you want to achieve has a direct link with the standard deviation. For many domains there exists consensus of what sufficient certainty is but not every domain uses the same kind of certainty level. For medicines the certainty must be much higher than in construction to be accepted.

Above explains the principle of the standard deviation but not how the standard deviation actually is calculated. When you have a fixed set of data then it is straight forward to define the % of the data which must be within the boundaries. However if you have only a calculated value and you need to know the standard deviation of that, then often some very complex calculations are needed. The paper includes those calculations and can be verified.
imdvb_8793 imdvb_8793 8/4/2016 05:18
Also, you're calculating your standard deviations using values corresponding to pawn units, which, again, is, ultimately arbitrary, because it's based on human players' impressions of what this or that positional or material advantage is worth in general, which could be quite a bit off their REAL worth, not to mention that they could also vary WILDLY from position to position, and not be fixed, the way the engines use them; there's absolutely no guarantee the numerical evaluation of a position is a linear function.
imdvb_8793 imdvb_8793 8/4/2016 05:12
(I use caps NOT to shout, but to emphasize/accentuate certain words - instead of bold font, for instance. Don't read it as shouting in my case, please! It's not.)

OK, but how is ANY of that truly relevant if you're calculating your standard deviations using an arbitrary tool (an engine - weak or strong, it doesn't matter) that assigns arbitrary values that none of us know for sure have any real significance whatsoever to the question of how strong a move is? (And won't know until chess is solved.) This has been the question I've been trying to get you to answer for a while now. (And, yes, if I'm too dumb to understand some part of this, then, please, by all means, explain it to me assuming I know NOTHING about ANYTHING. I won't be insulted. BUT, if you DON'T explain it properly, then don't get angry if I still don't understand and, therefore, still don't believe you're right!...)
brabo_hf brabo_hf 8/4/2016 04:47
"WE CAN'T KNOW (AT ALL) what the current accuracy of the engines is, BECAUSE chess is not solved, or anywhere near being solved."
Don't use capitals please. No need to shout.

There is no need to know the exact accuracy of the engines. What we need to see is if there can be made different profiles for some well selected eloranges. I try to explain.

Test A with a weak engine gives following results:
A bunch of 2500 rated players makes moves of which the evaluation deviates averagely 0,5 points from what the top choice of the engine tells us. The standard deviation is 0,2
A bunch of 2600 rated players makes moves of which the evaluation deviates averagely 0,4 points from what the top choice of the engine tells us. The standard deviation is 0,2
It is clear from the results that the weak engine gives insufficient accurate evaluations to make a clear distinction between both ratings.

Test 2 with a strong engine gives following results
A bunch of 2500 rated players makes moves of which the evaluation deviates averagely 0,5 points from what the top choice of the engine tells us. The standard deviation is 0,02
A bunch of 2600 rated players makes moves of which the evaluation deviates averagely 0,4 points from what the top choice of the engine tells us. The standard deviation is 0,02
Here the accuracy was sufficient as there is no overlap between both profiles.

The next step is to check how the profile evolves over time. If it changes beyond the boundaries of the standard deviation then we must agree there is very likely a serious inflation/ deflation. If it is not then it means inflation/ deflation is minimal.

The study showed that
1) the accuracy was sufficient to get clear profiles
2) over time the changes stayed within the standard deviation

Therefore the final conclusion was that inflation/ deflation has been minimal during the last decades. Definitely a 2700 rated player of today is averagely playing better moves than a 2600 rated player of 20 years ago.
imdvb_8793 imdvb_8793 8/4/2016 03:59
"Of course chess is not solved but that is not an argument to say that we can't make a verdict based on the current accuracy of the engines."

See, NOW who's being dumb? (Probably on purpose.) WE CAN'T KNOW (AT ALL) what the current accuracy of the engines is, BECAUSE chess is not solved, or anywhere near being solved. We can't know if they're EVEN CLOSE to the right evaluation for ANY position. I guess that's the part you weren't getting, for some reason. (Or maybe you just disagree, but, if you disagree with obviously true statements... that I can't help you with...)

"I gave this example only to explain how the principle of a standard deviation works. I don't get the impression you understand this which I deduct from your never ending rumbling about perfection/ complete solving of chess."

No, see, you're the one rambling (yes, it's 'rambling', not 'rumbling') on about standard deviation and other statistical principles that simply CANNOT BE APPLIED (without making incorrect assumptions, which you keep making) to this particular problem, because you have no objective means of measuring your predictor variables. SEE ABOVE!

"Many courts will decide to put people into prison because of a DNA match although everybody knows a DNA match doesn't give a perfect answer. Such kind of high statistical values are in a lot of domains used to make a verdict which goes far beyond an opinion."

1. That doesn't make it right. I strongly disagree with that, and other, similar, practices.
2. This is not the same thing, because you can't actually calculate the probability that an engine's evaluation of a position is right or wrong (or how right or wrong it is.) Not objectively/scientifically. Maybe I'm wrong about this - I'm open to that possibility. So, explain it to me! How do you do that? (And don't tell me you use other engines, LOL!... And don't link me to other people's articles - if you actually understand how to do this, and aren't just blowing smoke, then YOU explain it to me!...)

"How should we summarize your posts?"

Summarize them however you want. I don't care.
brabo_hf brabo_hf 8/4/2016 01:01
"YOU started trolling ME by just repeating yourself with every new post and refusing to answer simple questions."
Dumb questions you mean as I pointed out in my last comment.

I summarize the content of my answers just to proof how little repetition there was:
http://www.cse.buffalo.edu/~regan/papers/pdf/ReHa11c.pdf
Based on 150.000 games
1971 first fide rating
Very few players rated before 1976
15 years for +2700
30 years for players below
http://www.olimpbase.org/Elo/Elo197501e.html and the current top 100 list.
0 players of India and China in top 100. Today we have 7 players from India and 8 from China
Rating follows Gauss curve
Greatest player is not always the strongest player.
Each champion is a product of its time
A collection of games is not a good presentation of somebodies strength.
http://www.cse.buffalo.edu/~regan/papers/pdf/Reg12IPRs.pdf
2 examples of Tal in which the performances fluctuate
Make your own study and let it review by experts.
Fischer did make errors in his analysis.
http://chess-brabo.blogspot.com/2016/06/fischer.html
Kasparov Predecessors includes a number of correction which you can check yourself.
I have no interest in a discussion about some individuals.
I have no personal link to the study of Kenneth Regan and Guy Haword.
Their study is based on profiles not individuals
http://schaken-brabo.blogspot.ro/2012/12/elo-inflatie.html
I've been working 20 years extensively with engines.
Many articles on my blog are about new methods, techniques.
http://chess-brabo.blogspot.com/2015/10/mistakes.html
Correlation between mistakes and playing strength.
Data , C++ and Perl code can be found on www.cse.buffalo.edu/~regan/chess/
Processing was done fully automated.
https://en.wikipedia.org/wiki/Uncertainty_principle
My profile which proofs that I am 40 years old: https://ratings.fide.com/card.phtml?event=203602
Other study: https://kar.kent.ac.uk/44719/1/1-s2.0-S0167404814001485-main.pdf
Classical against alternative doctors
What is an opinion based on wikipedia?
Explaining the term standard deviation wth an example of http://www.computerchess.org.uk/ccrl/4040/
Example of the inaccuracy of a watch
Example of the inaccuracy of DNA to be used in courts

Probably I miss still a few things. How should we summarize your posts?
brabo_hf brabo_hf 8/4/2016 10:32
"Do you, or do you not, agree with me on this point? "
That chess is not solved? You really expect me to give answers on such stupid questions. Applaus for the troll. Of course chess is not solved but that is not an argument to say that we can't make a verdict based on the current accuracy of the engines.

Many courts will decide to put people into prison because of a DNA match although everybody knows a DNA match doesn't give a perfect answer. Such kind of high statistical values are in a lot of domains used to make a verdict which goes far beyond an opinion.

"It doesn't matter which is best."
My point was if you want to buy/ download the best publicly available engine then today it is impossible to know which program plays the best chess. I gave this example only to explain how the principle of a standard deviation works. I don't get the impression you understand this which I deduct from your never ending rumbling about perfection/ complete solving of chess.
imdvb_8793 imdvb_8793 8/4/2016 09:19
"Today engines have proven many times that they are now playing strong enough to give us sufficient accuracy to give us a pretty good verdict about the quality of the moves."

Actually, they haven't - because we don't know what perfect play is like. For all we know, if an engine that calculated all variations to the end (mate or forced draw) from the starting position was released today, it might beat the latest Komodo, Stockfish, Houdini, etc., 100 out of 100 games. We just don't know. We assume it wouldn't (I don't, but I imagine most people do, yourself included), but we don't know. There's zero evidence about whether this is or is not the case. Because the game of chess, in its entirety, is still an unsolved problem, mathematically speaking.

Do you, or do you not, agree with me on this point?

"Besides the standard deviation in a statistical analysis also warns us if the accuracy is too low for making any verdict. A simple example of too low accuracy to be sure who is best can be found on http://www.computerchess.org.uk/ccrl/4040/. The standard deviation is too big to make any strong verdict between Komodo and Stockfish"

It doesn't matter which is best, as long as neither plays perfect, unbeatable (by, let's say, future versions of the same engines) chess. Which we can be pretty sure they don't.
As long as neither plays perfect chess, they're an unfit tool to make pronouncements about the quality of anybody's moves. To say they have any worth for that purpose is no more than an act of faith, and most certainly not scientific fact. Tablebases are the only such tool we have today (because they calculate certain positions to mate or to a mathematical draw), but they can only help us with, on average, maybe 1-2% of all moves played in a chess game. Maybe 5%?! I don't know - but it's definitely not more than that.

"I think there is also somehow a misunderstanding on your side as you always talk about specific cases: Fischer,Tal ... I have stated in the very beginning that we were talking about profiles so not mapping to any specific individual."

It's been many messages since I made any mention of either Fischer, Tal or anybody else...
brabo_hf brabo_hf 8/3/2016 05:18
"whatever conclusions anybody draws about inflation (or anything else) based on chess engine analysis, no matter their qualifications, IS nothing more than an opinion."
I don't agree with that. An opinion is something subjective see wikipedia. An evaluation of an engine follows a very well defined algoritme which can be verified at any time. There is nothing subjective about it. The only thing you can argue if the accuracy of the evaluations is sufficient to give a verdict about the quality of the moves.
You could compare by using a watch to measure the time. You know in advance that it will never give you an absolute measurement but everybody will accept that the watch gives a sufficient accuracy to make a verdict of the passed time and comparisons between old and new measurements are possible. Ok maybe not the best comparison but I hope you get my point.
Today engines have proven many times that they are now playing strong enough to give us sufficient accuracy to give us a pretty good verdict about the quality of the moves. Besides the standard deviation in a statistical analysis also warns us if the accuracy is too low for making any verdict. A simple example of too low accuracy to be sure who is best can be found on http://www.computerchess.org.uk/ccrl/4040/. The standard deviation is too big to make any strong verdict between Komodo and Stockfish

Please also note that I have never talked about giving a verdict of somebodies rating. We never talk about specific individuals with its own characteristics. I think there is also somehow a misunderstanding on your side as you always talk about specific cases: Fischer,Tal ... I have stated in the very beginning that we were talking about profiles so not mapping to any specific individual.
imdvb_8793 imdvb_8793 8/3/2016 03:45
"I am open for freedom of speech but a debate about inflation should not be steered by opinions."

Again, until chess is solved - which it's currently not -, whatever conclusions anybody draws about inflation (or anything else) based on chess engine analysis, no matter their qualifications, IS nothing more than an opinion. All you're doing is presenting opinions as facts. At least the rest of us aren't that pretentious...
brabo_hf brabo_hf 8/3/2016 03:32
"I don't even claim inflation is a fact. I claim that, in my (and other people's) opinion, there is inflation. "
"There is no way to prove that either side, mine or yours, is wrong."

Well I am convinced that it is possible via a scientific model (the earlier mentioned professors developed one but I know of at least one other pair of scientists which did a similar kind of research using a different method see https://kar.kent.ac.uk/44719/1/1-s2.0-S0167404814001485-main.pdf) to define the correlation of strength of moves with the elo. I am an engineer so obviously I very much rely on science. However I do realize that some people will keep going to alternative doctors instead of a classical doctor despite whatever study tells them.

Still I want to add that people are shouting too often their opinion on the internet. I am open for freedom of speech but sometimes opinions only harm.
imdvb_8793 imdvb_8793 8/3/2016 02:51
I don't even claim inflation is a fact. I claim that, in my (and other people's) opinion, there is inflation. There is no way to prove that either side, mine or yours, is wrong, however, for the reasons I stated, and all your comments and links have done nothing to prove otherwise.

(You're the only one claiming things are facts - and using anything but to prove it.)

As for what your impressions about my intentions were/are, I couldn't care less... Also, you're a bit naive if you think anybody followed our argument as far as the two of us did... Anyway, good luck with your little obsession, as well as your superiority complex, in the future!
brabo_hf brabo_hf 8/3/2016 01:47
"Yes, because, like I said, I have no further interest in it."
From the very beginning of this discussion there was no balance. So even if you ever had interest, I never had the impression you were willing to seriously discuss inflation. Probably it is the wrong place too but where else?

You made your point about inflation and I did my best to counter it with numerous comments. At least the readers will now think twice when somebody claims inflation is a fact. More I didn't want to achieve.
imdvb_8793 imdvb_8793 8/3/2016 01:10
"So I provided you the paragraph where you could find the link and also mentioned that the study is about more than 150.000 games. Maybe I misunderstood the request but I just tried to give an answer"

So that's what that was. OK, fair enough. Link's broken, by the way ('The requested URL /∼regan/chess/ was not found on this server'), but don't bother correcting it, because, if you read through our conversation since then carefully, you'll notice that the "anybody knows" part of the discussion took a different turn, which makes the need for this link obsolete. Also, for the millionth time, I don't care anymore, so stop bombarding me with links, broken or otherwise!

"There is no balance in this discussion."

Yes, because, like I said, I have no further interest in it. (At some point, if you still refuse to understand this, I'll just have to leave this page for good - and break my promise, which I'd rather not have to do, but, if you keep being a dick about it, I can live with it -, if you insist on wasting my time further. It's not far off, so, if that's the escape hatch you're looking for, go ahead, take it! I don't care.)

"I am 40 year old and my fidecard can be found on https://ratings.fide.com/card.phtml?event=203602"

I didn't ask for extra details. Zero interest.
I'm 30.

"Well yes you trolled me as you are not really interested in this"

I WAS interested - until YOU started trolling ME by just repeating yourself with every new post and refusing to answer simple questions. But I don't care about that anymore - you can delude yourself into thinking I'm the one trolling all you want. Have at it! :)

"I can see on the statistics of my blog that you never did a serious effort reading anything as all information is there available and much more."

Actually, your statistics are wrong, because I DID read quite a bit of the stuff on your site - before I realized you were just an amateur with delusions of grandeur, at which point, again, I lost interest. (I assume you're not going to take this as a personal attack, since you call people amateurs all the time yourself.)
brabo_hf brabo_hf 8/3/2016 12:14
"Or maybe I have no interest in that part of the problem. Where did I say that I did? THAT is what you are clearly failing to grasp."
I quote you "I want a link to a study of how they correlate with, let's say, 1000 players' strengths. And I want it done by the book, with mathematical proof for every step, no random assumptions, "
So I provided you the paragraph where you could find the link and also mentioned that the study is about more than 150.000 games. Maybe I misunderstood the request but I just tried to give an answer (something which you complain that I don't do enough but I don't want to write about every sentence of your very long replies always something which often has nothing to do anymore with the original subject and is often nothing more than personal attacks).

"The uncertainty principle is highly debatable in both my and other people's opinions. You're trying to sound like a savant again. You're not fooling anyone - just give it up!... "
Your opinion, others, my friends,... I give references, have spent countless hours on research, ... There is no balance in this discussion.

"And how old are YOU? (I'm not dumb enough to answer that before you do.)"
I am 40 year old and my fidecard can be found on https://ratings.fide.com/card.phtml?event=203602
Well yes you trolled me as you are not really interested in this and just wanted to see my reaction. I can see on the statistics of my blog that you never did a serious effort reading anything as all information is there available and much more.
imdvb_8793 imdvb_8793 8/3/2016 11:14
"I only copied the full paragraph out of the paper because you could find back where I found the links. You are clearly too lazy to do it yourself."

Or maybe I have no interest in that part of the problem. Where did I say that I did? THAT is what you are clearly failing to grasp.

"That is not necessary either as we can work around that by using mathematics based on the uncertainty principle"

The uncertainty principle is highly debatable in both my and other people's opinions. You're trying to sound like a savant again. You're not fooling anyone - just give it up!...

"Nobody forces you to read the comments here and reply. However if you write comments then you should also expect people looking at it critically."

So? You did. I countered. You repeated yourself. I got bored. So why do YOU want to keep up the conversation? What's there to gain? I know I have nothing left to hear from you that you haven't said already, in one form or another, but what about you? What's your reason for not stopping?

"B.t.w. I notice that you already had quite some quarrels on the internet. Seems you like to shock people. How old are you?"

You have an unhealthy obsession with me. Grow up! And how old are YOU? (I'm not dumb enough to answer that before you do.)
brabo_hf brabo_hf 8/3/2016 09:21
"I'm not talking about gamescore errors - where did you get that notion? "
I only copied the full paragraph out of the paper because you could find back where I found the links. You are clearly too lazy to do it yourself. I had no intention to start up a discussion about the necessity to first clean gamescores from errors which is an obvious thing.

"Until then, what computers "think" about it is about as reliable (from a scientific standpoint) as what any half-decent player thinks about it. Which is not at all. "
Well I never claimed that engine evaluations must give you at any time an absolute verdict about a position. That is not necessary either as we can work around that by using mathematics based on the uncertainty principle (https://en.wikipedia.org/wiki/Uncertainty_principle). The article warns that it is likely too technical for most readers so I guess you will need some help from your friends.

"I have no more interest in this conversation, so please leave me alone! Thank you!"
Nobody forces you to read the comments here and reply. However if you write comments then you should also expect people looking at it critically. B.t.w. I notice that you already had quite some quarrels on the internet. Seems you like to shock people. How old are you?
imdvb_8793 imdvb_8793 8/3/2016 07:23
I'd like that too, pwcca - a (genuine) rivalry between Carlsen and anybody would be nice, really... But, sadly, MVL has just pretty much always been outplayed by Carlsen, maybe even more so than the average elite GM, so I doubt that he'll ever catch up enough to pose him serious problems... No, I think we need to look elsewhere for a worthy rival for Carlsen!