Snail Grading is a grading system designed for use in our snail tournaments. It has proved difficult to do Swiss pairings while all our players are ungraded.
Snail Gradings use a similar algorithm to the English Chess Federation’s grading system. As in our club we have no players with current over the board (OTB) standard ratings I have made no attempt to align snailgrades with any other system.
In snailgrades, and the ECF grades each game is scored. For a win the score is 50 + opponents score, and for a loss the score is 50 - opponents score. There is a proviso that if the difference in scores is more than 40 it is taken to be 40.
The code is shown here:
if result == "1-0": n = 1 elif result == "1/2-1/2": n = 0.5 elif result == "0-1": n = 0 else: sys.exit("Invalid Result") new_grade = opponent_grade - 50 + 100*n if new_grade - current > 40: new_grade = current + 40 elif new_grade - current < -40: new_grade = current - 40 if new_grade < 0: new_grade = 0
Snailgrades take the mean of the new grade calculated for the last 20 games.
I intend snailgrades to be available for standard games in organised events within the club. We can use the Welsh Chess Union’s notion a standard game. Effectively games within an OTB tournament with a minimum time of one hour and normal play. (This excludes Bronstein and other exotic clock settings or adjourned games)
Snailgrades have very little signifigance while less than 5 games have been recorded. If more than 5 games and less than 20 games the grade will be provisional. Until most of us have recorded 20 games we can expect a noticable increase in average snailgrade because of ‘leakage’ in the algorithm. (Because new grades of less than zero are taken to be zero)