The Ratings Committee has been requested by the Executive Board to perform an analysis to evaluate the recent DDDE ratings proposal, which was communicated to the Committee as consisting of the following three components:
As we describe below, the Committee strongly recommends against this proposal's adoption. We conclude that, after one year beyond its implementation, a significant number of ratings would become nearly meaningless, and the rating system, one of the USCF's most important promotional tools, would cease to be functional.
Committee member Thomas Doan has recomputed all USCF ratings based on games played from January 1998 through September 1998 from archival data under two systems. First, he recomputed ratings using the new system that will be implemented at the USCF in the next month. In the discussion below, this will be termed the ``ordinary'' system. Secondly, he recomputed ratings using the new system with the addition of the DDDE proposal. A comparison of the two sets of resulting ratings provides information about the effect of the DDDE proposal. The following is a summary of the difference between the two sets of ratings.
While the above analysis is based on actual tournament data, and only 9 months of game outcomes rather than 12 months, we need to stress that the effect of the DDDE proposal is likely to be much worse than the above analysis suggests. The DDDE proposal is an invitation to abuse. The preceding analysis demonstrates the effect of the proposal when players are not aware of the frequency-based rating additions (players in 1998 had no particular incentive to compete more often). If the DDDE rating proposal is prominently publicized, it is easy to imagine that some players will actively exploit the system, leading to enormous rating gains (worse than the type demonstrated in XXXXXXXXXXXXX's case).
The goal of the rating system is to provide an accurate measure of playing ability for the purpose of ranking, sectioning, prize-eligibility, and qualification to invited events. Any rating proposal that is at odds with these goals undermines its functionality. In particular, any proposal, such as the one generated by the DDDE committee, that awards rating points as a consequence of frequency of play, is in direct opposition to the goals of the rating system. We therefore strongly recommend against the specific proposal, and against similar proposals that award rating points as a consequence of playing frequency.
The new rating system, which has been tested extensively, has solid anti-deflationary mechanisms that will prevent ratings from declining over time. Under the new system, players who have exceptional performances will have their ratings increase quickly, so that improving scholastic players will reach appropriate rating levels faster than under the old system. For example, a player with an established pre-tournament rating of 1300 (based on 45 games) competing against players rated 1250, 1400, 1500 and 1550, winning 3 and drawing 1, will have a post-tournament rating under the new system of 1434. Under the old system, the post-tournament rating would have been only 1368. The Ratings Committee therefore suggests immediately marketing the anti-deflationary aspects of the new system.
Mark Glickman, Committee co-chair
Frank Camaratta, Committee co-chair, ANTD, NM, ICCF IM
Christopher Avery, NM
Harry Cohen,
Thomas Doan, STD
Bill Goichberg, NTD, FM
Albyn Jones
Larry Kaufman, IM
David Kuhns, NTD
Alan Losoff, NTD
Andrew Metrick, NM
Kenneth Sloan, STD
Christopher Yaure
Ordinary # of DDDE Rating
ID Player Rating Games Rating Diff
--------------------------------------------------------------------
XXXXXXXX XXXXXXXXXXXXXXXX 1868 286 2366 498
XXXXXXXX XXXXXXXXXXXXXXXX 1915 272 2257 342
XXXXXXXX XXXXXXXXXXXXXXXX 1866 276 2201 335
XXXXXXXX XXXXXXXXXXXXXXXX 1766 437 2096 330
XXXXXXXX XXXXXXXXXXXXXXXX 1981 295 2301 320
XXXXXXXX XXXXXXXXXXXXXXXX 1915 223 2210 295
XXXXXXXX XXXXXXXXXXXXXXXX 2047 267 2330 283
XXXXXXXX XXXXXXXXXXXXXXXX 1615 152 1897 282
XXXXXXXX XXXXXXXXXXXXXXXX 2106 232 2381 275
XXXXXXXX XXXXXXXXXXXXXXXX 1961 189 2235 274
XXXXXXXX XXXXXXXXXXXXXXXX 2115 169 2388 273
XXXXXXXX XXXXXXXXXXXXXXXX 1739 175 2011 272
XXXXXXXX XXXXXXXXXXXXXXXX 2040 121 2306 266
XXXXXXXX XXXXXXXXXXXXXXXX 1918 117 2176 258
XXXXXXXX XXXXXXXXXXXXXXXX 1452 155 1700 248
XXXXXXXX XXXXXXXXXXXXXXXX 1940 129 2184 244
XXXXXXXX XXXXXXXXXXXXXXXX 1912 125 2156 244
XXXXXXXX XXXXXXXXXXXXXXXX 1730 194 1974 244
XXXXXXXX XXXXXXXXXXXXXXXX 1812 205 2056 244
XXXXXXXX XXXXXXXXXXXXXXXX 2132 166 2375 243
XXXXXXXX XXXXXXXXXXXXXXXX 1663 393 1905 242
XXXXXXXX XXXXXXXXXXXXXXXX 1696 185 1930 234
XXXXXXXX XXXXXXXXXXXXXXXX 1961 138 2191 230
XXXXXXXX XXXXXXXXXXXXXXXX 2038 185 2266 228
XXXXXXXX XXXXXXXXXXXXXXXX 1805 186 2032 227
XXXXXXXX XXXXXXXXXXXXXXXX 1859 159 2084 225
XXXXXXXX XXXXXXXXXXXXXXXX 2102 191 2326 224
XXXXXXXX XXXXXXXXXXXXXXXX 2097 134 2319 222
XXXXXXXX XXXXXXXXXXXXXXXX 1671 118 1893 222
XXXXXXXX XXXXXXXXXXXXXXXX 1849 105 2069 220
XXXXXXXX XXXXXXXXXXXXXXXX 1762 159 1981 219
XXXXXXXX XXXXXXXXXXXXXXXX 1971 147 2190 219
XXXXXXXX XXXXXXXXXXXXXXXX 2215 183 2433 218
XXXXXXXX XXXXXXXXXXXXXXXX 2066 141 2283 217
XXXXXXXX XXXXXXXXXXXXXXXX 1992 174 2206 214
XXXXXXXX XXXXXXXXXXXXXXXX 1947 249 2160 213
XXXXXXXX XXXXXXXXXXXXXXXX 2054 135 2266 212
XXXXXXXX XXXXXXXXXXXXXXXX 1761 200 1972 211
XXXXXXXX XXXXXXXXXXXXXXXX 1971 133 2182 211
XXXXXXXX XXXXXXXXXXXXXXXX 1827 144 2035 208
XXXXXXXX XXXXXXXXXXXXXXXX 1459 207 1666 207
XXXXXXXX XXXXXXXXXXXXXXXX 1929 128 2135 206
XXXXXXXX XXXXXXXXXXXXXXXX 2175 152 2381 206
XXXXXXXX XXXXXXXXXXXXXXXX 2094 237 2300 206
XXXXXXXX XXXXXXXXXXXXXXXX 1865 140 2070 205
XXXXXXXX XXXXXXXXXXXXXXXX 1942 131 2146 204
XXXXXXXX XXXXXXXXXXXXXXXX 2191 284 2394 203
XXXXXXXX XXXXXXXXXXXXXXXX 1913 186 2114 201
XXXXXXXX XXXXXXXXXXXXXXXX 2089 94 2290 201