Another question for Stanley

[AussieVamp2]

or Liker of Hoch

anyway, a Course Compatability Question

obviously perusing your stats again this week for the US event

What stats are you using for each year to look at these courses -

The yearly stats for the tour, or the stats you had at the time of that event itself, or the current stats for the players and then looking back at the tournament? (some guys would not be playing?)

Sorry for being a bit slow, but it is not quite clear?

Thanks,

AV

 

[Stanley]

AV,

I take the players' yearly stats for that and the previous year and use a 12-month rolling average approach. So if the event is the 24th of forty-two in the 1998 season, then for each player the relevent retrospective stat is 23/42 * 1998 year-long stats + 19/42 * 1997 year-long stats.

I use this 12-month rolling average approach for all the core stats [driving distance, driving accuracy, greens in regulation & putting average] on the site. Eliminates short-term volatility due to the form cycle and yearly statistics are available quite widely across most Tours so comparative stats for players from different Tours are possible.

I then use a weighted average of the top-20 [where the weight is n/sum and n=finishing position and n=1,2,...,20] to provide an estimate of the relative weights of the core stats in determining finishing position on a particular course in a particular year. This is done for the past 3 years to provide average 'course compatibility' stats and show how different parts of the game may become important after course changes.

I'm sure it is not quite clear, but with an air of mysticism there is the presumption of knowledge

 

[AussieVamp2]

cool, thanks

I think knowing which tournament is what number for what year and what tour might require a bit of mysticism