Computer Cheating in Chess Tournaments

[samsloan]

According to a report by Kenneth W. Regan cited above, the moves by a
2200 player will match the computer 48% of the time, by a 2300 player
50% of the time and moves by a 2700 player matches the computer 56% of
the time.

Thus, if a player previously known to be a 1900 player suddenly starts
making moves that match the computer 90% of the time we can be sure
that he is cheating.

Sam Sloan

[abc]

Post by samsloan
According to a report by Kenneth W. Regan cited above, the moves by a
2200 player will match the computer 48% of the time, by a 2300 player
50% of the time and moves by a 2700 player matches the computer 56% of
the time.

That's only an 8% difference, over a range of 500 rating points.
Rather surprising imho, and it would seem to make the cheat-detecting
task more difficult.

Post by samsloan
Thus, if a player previously known to be a 1900 player suddenly starts
making moves that match the computer 90% of the time we can be sure
that he is cheating.

Wouldn't a cunning cheater be able to avoid detection by making
moves ranging from the second to fifth-best computer move or so,
except for clear blunders, and still end up at mostly GM strength?

abc