Thursday 28 August 2008

About the Overall GP Score

The overall GP (Golf Predictor) score for each golfer in a tournament is the normalised chance that s/he (the occasional woman also plays!) has of winning/playing well in the tournament. The GP score is based on the statistics available for that golfer and is computed as an easy to digest percentage. The higher this percentage, the greater the chance of that golfer winning the event.

In simple terms, a GP score of 100% would represent the top world ranked golfer in the field with perfect statistics in all the metrics used, e.g. season, last five events, in the past on the same course, in the same tournament, at similar events etc. Of course, no golfer ever wins all the relevant historical events, but a golfer who hits a rich vein of form leading up to a tournament on a course he has a great record on can certainly achieve a very high GP Score. Such golfers would obviously be very likely to perform well again in the current week. Conversely, golfers with low GP scores generally are amateurs, debutants or sponsor invites with little or no history on either major tour.

The GP score is not a pure mathematical probability. Tiger Woods playing his best could have a GP score of 90%, but even he would never have a pure 90% chance of winning in a field of 156 golfers. If all the golfers were of equal ability, the pure probability would be 1/156=0.641% for any golfer to win. In a real life tournament of 156 golfers of varying abilities, the true odds for most of the field would be difficult to digest decimals slightly bigger or smaller than 0.641%. That is the main reason the GP Score has been normalised to an easy to understand percentage. It is a valid measure of likely success and, like bookie odds, the relative difference between two GP scores is a measure of how much more likely one golfer is to perform well than the other.

No comments: