EGD Go Rating anti-deflation loophole

This post is a follow-up to EGD Go Rating win/loss lookup table.

Winning against an opponent of the same rating gains more rating points (GoR) than losing to that opponent loses. This is by design to avoid deflation. However, this can also be used to artificially increase one’s rating.

The two tables below answer the question: “How does the rating change if, starting from a certain GoR, one plays a series of games against someone of equal rating, alternately winning and losing?” The columns show the results after 10, 20, 50 and 100 games. The first table is for the situation where all games are played in a single tournament, so the starting GoR is always the same. The second table is for the situation where each game is played in its own “tournament”, so the starting GoR is updated after each game.

The values are for class A tournaments without handicaps.

This is hypothetical, of course, and probably wouldn’t be allowed in the EGD. However, a series of ten-game matches (十番碁, “juubango”) would not be that unusual. Go Seigen played a series of such matches, and in 2014 Lee Sedol played such a match against Gu Li.

Table 1: All games are played in one tournament. Therefore, the starting GoR is the same for each game.

GoR 10 games 20 games 50 games 100 games
1000 1032.5 1097.5 1260.0 1585.0
1050 1081.3 1143.8 1300.0 1612.5
1100 1130.0 1190.0 1340.0 1640.0
1150 1178.8 1236.3 1380.0 1667.5
1200 1227.5 1282.5 1420.0 1695.0
1250 1276.3 1328.8 1460.0 1722.5
1300 1325.0 1375.0 1500.0 1750.0
1350 1373.8 1421.3 1540.0 1777.5
1400 1422.5 1467.5 1580.0 1805.0
1450 1471.3 1513.8 1620.0 1832.5
1500 1520.0 1560.0 1660.0 1860.0
1550 1568.8 1606.3 1700.0 1887.5
1600 1617.5 1652.5 1740.0 1915.0
1650 1666.3 1698.8 1780.0 1942.5
1700 1715.0 1745.0 1820.0 1970.0
1750 1763.8 1791.3 1860.0 1997.5
1800 1812.5 1837.5 1900.0 2025.1
1850 1861.3 1883.8 1940.1 2052.6
1900 1910.0 1930.0 1980.1 2080.2
1950 1958.8 1976.3 2020.2 2108.0
2000 2007.5 2022.6 2060.4 2135.8
2050 2056.3 2069.0 2100.7 2164.0
2100 2105.2 2115.5 2141.3 2192.8
2150 2154.0 2162.1 2182.3 2222.6
2200 2203.0 2209.0 2224.0 2254.1
2250 2252.1 2256.3 2266.9 2287.9
2300 2301.4 2304.2 2311.1 2325.0
2350 2350.9 2352.6 2356.9 2365.4
2400 2400.5 2401.5 2404.0 2409.1
2450 2450.3 2450.9 2452.3 2455.1
2500 2500.2 2500.5 2501.3 2502.8
2550 2550.1 2550.3 2550.7 2551.5
2600 2600.0 2600.1 2600.4 2600.8
2650 2650.0 2650.1 2650.2 2650.5
2700 2700.0 2700.0 2700.1 2700.2

Table 2: Each game is played in a separate “tournament”. Therefore, the starting GoR is updated after each game.

GoR 10 games 20 games 50 games 100 games
1000 1034.2 1099.9 1249.4 1494.4
1050 1082.9 1146.0 1289.7 1525.2
1100 1131.5 1192.1 1330.0 1556.1
1150 1180.2 1238.2 1370.3 1587.0
1200 1228.9 1284.3 1410.6 1617.9
1250 1277.5 1330.5 1451.0 1648.8
1300 1326.2 1376.6 1491.3 1679.8
1350 1374.9 1422.7 1531.7 1710.7
1400 1423.6 1468.9 1572.1 1741.7
1450 1472.3 1515.0 1612.5 1772.8
1500 1520.9 1561.2 1653.0 1803.8
1550 1569.6 1607.4 1693.4 1834.9
1600 1618.3 1653.5 1733.9 1866.0
1650 1667.0 1699.7 1774.3 1897.2
1700 1715.7 1745.9 1814.8 1928.4
1750 1764.4 1792.1 1855.4 1959.6
1800 1813.1 1838.3 1895.9 1990.9
1850 1861.8 1884.6 1936.5 2022.3
1900 1910.5 1930.8 1977.1 2053.8
1950 1959.3 1977.1 2017.8 2085.4
2000 2008.0 2023.4 2058.6 2117.4
2050 2056.7 2069.8 2099.6 2149.7
2100 2105.5 2116.2 2140.9 2182.6
2150 2154.4 2162.9 2182.6 2216.4
2200 2203.3 2209.8 2225.0 2251.5
2250 2252.4 2257.2 2268.3 2288.3
2300 2301.7 2305.0 2312.9 2327.4
2350 2351.1 2353.4 2358.8 2369.1
2400 2400.8 2402.3 2406.0 2413.1
2450 2450.5 2451.5 2454.1 2459.0
2500 2500.4 2501.1 2502.9 2506.4
2550 2550.3 2550.8 2552.1 2554.7
2600 2600.2 2600.6 2601.6 2603.6
2650 2650.2 2650.5 2651.3 2652.8
2700 2700.1 2700.4 2701.0 2702.2