Expanding the research of "which countries have the most inflated elo"
Follow up of https://en.chessbase.com/post/which-countries-have-the-most-inflated-elo-chess-players (but I am not the original author!)With the help of claude sonnet 4.5 (that shortens the prototyping and testing time by a lot) I collected and analyzed the data of the u18 WCh from 2015 to 2019 as follow up of this article https://en.chessbase.com/post/which-countries-have-the-most-inflated-elo-chess-players. Why pre covid? Because then one cannot argue about possible inflation/deflation due to covid measures (IMO the covid rating lag ended around 2023). Additionally I considered the u18 WCh from 2023 to 2025 and the Grand swisses (2019, 2023, 2025). More tournaments may be added later. Of course in the WCh u18 there are likely the best young players of each country, so underrated if anything, but still it gives an idea which countries gain rating as a whole and which don't.
Result directly from the output
================================================================================
EXTRACTION SUMMARY
================================================================================
Total players extracted: 4032
Federations represented: 117
Tournaments covered: 36
Overall statistics:
Average rating change: +0.18
Players with gains: 1834 (45.5%)
Players with losses: 1998 (49.6%)
Top 20 federations by player count:
IND: 543 players
RUS: 228 players
ESP: 207 players
GER: 181 players
FRA: 152 players
None: 150 players
USA: 149 players
ISL: 128 players
CHN: 98 players
ENG: 95 players
NOR: 86 players
UKR: 80 players
ARM: 75 players
MEX: 70 players
ISR: 67 players
KAZ: 67 players
ITA: 64 players
POL: 62 players
HUN: 62 players
IRI: 58 players
Analyzed 97 federations with 3+ players
Analyzed 7 continents
==========================================================================================
RATING CHANGES BY CONTINENT
==========================================================================================
Rank Continent Players Feds Avg Change Total % Gainers
------------------------------------------------------------------------------------------
1 Oceania 34 2 +10.4 +354 58.8%
2 Africa 68 13 +5.3 +361 45.6%
3 North America 281 8 +1.8 +499 47.0%
4 Other 236 13 +1.1 +255 48.3%
5 Asia 1004 28 +0.9 +900 47.4%
6 Europe 2230 44 -0.7 -1506 44.4%
7 South America 179 10 -0.7 -129 39.7%
==========================================================================================
ALL FEDERATIONS RANKED BY AVERAGE RATING CHANGE (per player)
==========================================================================================
Rank Fed Players Avg Change Total % Gainers Avg Rating
------------------------------------------------------------------------------------------
1 NZL 5 +38.0 +190 80.0% 1790
2 NAM 3 +37.7 +113 100.0% 1324
3 LAT 13 +37.0 +481 69.2% 2239
4 LTU 10 +28.8 +288 40.0% 2324
5 ALB 6 +28.5 +171 66.7% 2118
6 BIH 9 +21.1 +190 66.7% 2310
7 FIN 10 +20.9 +209 70.0% 1983
8 POR 8 +16.0 +128 50.0% 2148
9 PAR 8 +14.1 +113 75.0% 2431
10 LBN 3 +13.7 +41 66.7% 1882
11 TPE 22 +13.0 +285 59.1% 1675
12 WLS 7 +12.9 +90 57.1% 2045
13 RSA 25 +12.8 +319 52.0% 2031
14 EST 14 +12.1 +169 50.0% 2352
15 MNE 18 +11.5 +207 50.0% 1896
16 MEX 70 +10.0 +700 52.9% 1601
17 KOS 5 +9.2 +46 40.0% 1971
18 KGZ 8 +7.5 +60 62.5% 1752
19 BRA 29 +7.3 +213 51.7% 2048
20 BEL 34 +6.5 +222 50.0% 2228
21 MDA 11 +6.2 +68 45.5% 2382
22 VIE 12 +6.0 +72 83.3% 2446
23 FAI 5 +5.8 +29 60.0% 2120
24 AUS 29 +5.7 +164 55.2% 2238
25 ALG 3 +5.3 +16 66.7% 2078
26 BAN 4 +4.5 +18 25.0% 2116
27 IRL 20 +3.8 +75 50.0% 2118
28 EGY 15 +3.7 +55 33.3% 2439
29 SWE 57 +3.6 +207 45.6% 2351
30 COL 24 +3.3 +79 45.8% 2269
31 ISL 128 +3.1 +398 46.9% 2145
32 UZB 56 +3.0 +170 51.8% 2397
33 MAS 13 +3.0 +39 53.8% 2113
34 IND 543 +3.0 +1620 49.9% 2353
35 SVK 19 +2.9 +55 52.6% 2344
36 GRE 36 +2.8 +100 50.0% 2310
37 FID 48 +2.8 +132 52.1% 2450
38 CAN 47 +2.7 +125 51.1% 2132
39 SCO 9 +2.6 +23 66.7% 2193
40 NOR 86 +2.4 +206 51.2% 2295
41 CRO 18 +2.3 +42 50.0% 2508
42 SGP 22 +2.1 +47 36.4% 2327
43 CHI 20 +2.0 +41 30.0% 2315
44 CUB 10 +2.0 +20 40.0% 2407
45 IOM 3 +1.7 +5 33.3% 2210
46 INA 5 +1.6 +8 40.0% 2405
47 SLO 28 +1.2 +35 57.1% 2375
48 ARG 51 +1.2 +63 45.1% 2357
49 DEN 35 +1.2 +41 45.7% 2228
50 KEN 5 +1.0 +5 40.0% 1981
51 UAE 13 +0.1 +1 46.2% 2191
52 SRI 13 +0.1 +1 61.5% 1940
53 ENG 95 -0.0 -3 46.3% 2330
54 None 150 -0.0 -5 46.7% 2613
55 RUS 228 -0.3 -59 43.0% 2543
56 HUN 62 -0.3 -17 43.5% 2453
57 PHI 6 -0.8 -5 66.7% 2374
58 MGL 27 -1.1 -29 40.7% 2352
59 USA 149 -1.6 -241 44.3% 2409
60 CHN 98 -1.7 -165 39.8% 2457
61 POL 62 -1.7 -108 45.2% 2456
62 ITA 64 -1.8 -115 40.6% 2315
63 MKD 7 -2.0 -14 57.1% 2183
64 FRA 152 -2.1 -315 43.4% 2418
65 BLR 11 -2.2 -24 54.5% 2594
66 SRB 33 -2.3 -77 30.3% 2411
67 UKR 80 -2.4 -189 43.8% 2497
68 ESP 207 -2.6 -528 42.0% 2347
69 TUR 35 -2.8 -98 42.9% 2432
70 MAR 4 -3.0 -12 50.0% 2134
71 GEO 50 -3.1 -153 42.0% 2335
72 ARM 75 -3.1 -234 42.7% 2525
73 GER 181 -3.3 -590 45.9% 2369
74 NED 50 -3.4 -171 46.0% 2419
75 SUI 25 -3.8 -94 36.0% 2271
76 IRI 58 -3.9 -224 37.9% 2504
77 ISR 67 -4.8 -322 32.8% 2446
78 CZE 31 -4.9 -151 38.7% 2412
79 VEN 11 -4.9 -54 18.2% 2436
80 TJK 5 -6.0 -30 40.0% 2283
81 AZE 39 -6.7 -261 46.2% 2519
82 KAZ 67 -6.9 -461 37.3% 2295
83 ROU 45 -8.4 -379 40.0% 2415
84 HKG 4 -11.8 -47 50.0% 1962
85 AUT 35 -12.1 -423 28.6% 2369
86 PER 23 -12.6 -290 30.4% 2284
87 BUL 29 -12.8 -371 37.9% 2429
88 MRI 3 -13.3 -40 33.3% 552
89 NGR 3 -14.7 -44 0.0% 2242
90 CYP 5 -15.8 -79 40.0% 1781
91 KUW 3 -16.3 -49 33.3% 1830
92 LUX 4 -17.5 -70 50.0% 1930
93 JPN 4 -20.0 -80 25.0% 1703
94 NEP 4 -27.0 -108 50.0% 1870
95 TKM 7 -27.1 -190 28.6% 2079
96 URU 8 -27.6 -221 0.0% 1407
97 ECU 4 -32.2 -129 0.0% 2070
that is, this add support to my hypothesis that states more or less the following:
FIDE statistics from 2023-2025 indicate that India has between 32,000 and 75,000 active rated players. A massive proportion of these are Under-18 (U18) players. The Elo system has a known lag in tracking rapid improvement. A junior player might improve their skill by 200 Elo points in six months through intensive training. However, if they play only one rated tournament in that period, or—crucially—if they play primarily against other underrated juniors, their rating will not reflect this 200-point gain.
In the Indian ecosystem, underrated juniors largely compete against one another. Consider a tournament in Chennai where Player A (rated 1400, strength 1700) plays Player B (rated 1400, strength 1700). The game ends in a draw. The ratings remain 1400. Both players are essentially "smurfing"—carrying a rating far below their true strength. The "National" rating often diverges from the FIDE rating, creating a "heavy" pool where 1400 Elo requires 1700-level skill to maintain.
[...]
Recent analysis of global rating reliability has quantified this disparity. Research indicates that players from South Asia (dominated by India) are, on average, 150 to 250 Elo points underrated relative to a global skill baseline. Conversely, players from older, established European federations like Denmark are often overrated by ~162 points. (I wanted to double check those claims)
Europe (and to a lesser extent, the USA) serves as the "reserve currency" of the FIDE rating system. It has:
- High Tournament Density: The vast majority of FIDE-rated Open tournaments occur in Europe.
- Amateur Density: A large population of adult hobbyists who maintain ratings established decades ago.
- Inflationary Tendency: Older players tend to lose strength faster than their rating decays (due to low K-factors and floors), effectively "storing" points that should have been lost.
When an underrated Indian junior travels to play in the Grand Swiss, Cappelle-la-Grande, or the Gibraltar Masters, they engage in what is essentially rating arbitrage.
The Transaction:
- The Matchup: An Indian Junior (Rated 2100, Strength 2350) vs. A German Amateur (Rated 2300, Strength 2250).
- The Prediction: The Elo formula predicts the German (2300) will score 76% against the 2100 opponent.
- The Reality: The Indian Junior is stronger. The match ends in a Draw or a Win for the Indian.
- The Consequences:
- For the German: He loses significantly more points than he would losing to a peer. A draw might cost him rating points (since he was expected to win). A loss is catastrophic (-15 to -20 points depending on K-factor).
- For the Indian: He gains massive points.
- For the System: The German player is now rated 2280. He goes back to his local club and plays a peer rated 2280. He is still 2250 strength (one bad game didn't change his skill). He beats his peer. The peer loses points. The deflation spreads.
The points gained by the Indian junior are eventually taken back to India. Once there, they are likely lost to another underrated Indian junior. The points essentially vanish into the "black hole" of the deflationary Asian pool, never returning to the European ecosystem.
The mechanism driving the Asian Deflation (Geographic Isolation) is mathematically identical to the mechanism driving Rapid/Blitz Deflation (Format Isolation/Inactivity).
- Isolation: Just as Chinese players don't play enough international games to normalize their ratings, top Grandmasters often don't play enough FIDE-rated Blitz tournaments to normalize their speed ratings. They play online (Speed Chess Championship, Titled Tuesday) which does not affect FIDE ratings.
- Inertia: The K-factor (rating volatility) drops to 10 for elites. To gain 100 points, a player must massively outperform expectations over a long period. If they play only one Blitz tournament a year, their rating remains "frozen" in the past.
- The "Heavy" Asset: This results in a player having a rating (e.g., 2600) that is far below their current skill level (e.g., 2800).
python code here .
