Keep track of corrections to matetrack.epd

[robertnurnberg]

I would like to collect corrections made to the file matetrack.epd and that still need to be reflected in the stored statistics. Once I find the time, I will correct the old statistics.

• correct several wrong mate scores #26 : Affects all SF revisions between aa2368a6878a867fe63247ee2adf2fde3dfe22be and 6deb88728fb141e853243c2873ad0cda4dd19320 (inclusive). The corrections will only affect the best mate counts, by at most +/-17.
• correct three mates, and update some PVs #44 : Affects all SF revisions between aa2368a6878a867fe63247ee2adf2fde3dfe22be and fc41f64dfd8a61d0e275ddbecec292833458b86a (inclusive).
  The corrections will only affect the best mate counts, by at most +/-3.
• remove two likely 50mr draws #48 : Will in all likelihood have no effect on results at all.
• improve two mates #50 : Affects all SF revisions between aa2368a6878a867fe63247ee2adf2fde3dfe22be and d99f89506bd0ed535fb1c55dbb7cc8f7c29444d4 (inclusive). The corrections will only affect the best mate counts, by at most +/-2.
• improve one mate and add new PVs #53 : Affects all SF revisions between and
  aa2368a6878a867fe63247ee2adf2fde3dfe22be and 0716b845fdef8a20102b07eaec074b8da8162523 (inclusive). The corrections will only affect the best mate counts, by at most +/-1.

[robertnurnberg]

Closing this now, as richer data collection for the .csv file means that sooner or later we are likely to recompute the historical data anyway. Probably once the test suite file as settled on a (near) final status.