Thats a very good summary of the use of filters.

I think with a lot of data even going back past 2009 you can can a good feel for the strike rate of two variables.

For example - not knowing the stats but if in the last 2 years the favourites are winning at 35% but over the last 50 years favourites win at 33% one might think that with limited data the 35% is the common trend.

Some punters like to look over particular tracks and distances and see if particular running styles have an advantage from particular barriers and rail positions. If you only have 2 years of data you might only get 100 selections. Over 12 years you have 600 selections - giving you more of a idea long-term.

I suppose its a personal thing and what you are trying to achieve.

I think with Barny's approaches and the higher POT approach you would need more data to get a more realistic idea of the long term profit.

Anyhow - off to see if my $1000 propun horse gets me the cash today.