A day will not go by without you hearing one of the pundits talk about this horse or that horse has good course winning form and with those pointers easily readable in most form publications, punters no doubt rush off to back them. Is this really a sensible ploy, should we be backing course winners because ‘they like the place’ or is course form just an illusion ?.
I decided to check both course winning form and the more grainier good runs at a course. First course winning form taken from any races on the flat but then forward applied to only handicap races. I decided to check only course winners going off at less than 8/1. Here are the results for 1999 to 2004 with first two years as a data lead in period.

Bets 8724 PL -1221 ROI% -13.99%

Now if we check horse with no course winning form we have

Bets 24176 PL -3033 ROI% -12.54%

If we switch to good runs rather than just winning ones we get following corresponding figures

Bets 18687 PL -2538 ROI% -13.58%
Bets 14213 PL -1715 ROI% -12.06

I checked with a horse age restriction to eliminate any age bias but this had little effect. It would appear that punters would be better off backing horses that have no course form what so ever. When was the last time you heard a pundit say that ?.
Taking it one step further I checked when a horse had more good runs at a course than bad runs and vice versa. The results were as follows

Bets 17526 PL -245 ROI% -14%

Bets 1210 PL -135 ROI% -11.1%

Fewer bets but still better off with negative net course form. Some might feel that certain courses offer more mileage but I would be suspicious of such small data sizes.