Angel416,

I had a bit of a play with this using my database and, sorry to say, the results are not great. I tested with all the data I have for last calender year, using NSWTAB divs. One thing to note is that I ignored any races where either of the top 2 by weight were scratched so you may get a slightly different result if you substituted the next eligible runner instead of ignoring the race.

I also checked out the "2nd topweight" as suggested by Big Louie and this seems to have more potential. Anyway the results I got were as follows:

No Filters
1258 selections, TopW=27% Loss, 2ndW = 5% Loss

I tried all sorts of filters on the Top Weight selections but nothing made a great improvement.

Barrier seems pretty much irrelevant - 80% of them start from barrier 8 or less anyway. Barrier 1-8 = 26% LOT, Barrier 9+ = 30% LOT.

Track condition seems to show that Good or Fast tracks are the best bets but again there is only a few percent in it and everything shows a loss.

Race Distance seems to indicate that longer distances are more favourable. 1400m and over showed a loss of 13% whilst under 1400m showed a loss of 40%.

Metro v's NonMetro - Metro (all days) was significantly better than NonMetro for the Top Weight showing a loss of around 18% compared to 33% for country.

Price - Removing any runners over $10 improves POT a bit as only 1 winner over this price.

Even combining all the positive filters together I could not come up with a profit for the years worth of races I was working with.

However the 2nd Top Weights is a different story. Again with no filters there was around a 5% loss.

Barrier was interesting with a higher POT for barriers greater than 8 but again only 20% of runners fall into this category.

Track condition seems to be generally irrelevant (again HEAVY shows a bigger loss but quite a small number of bets in this category so it may be irrelevant, Fast has the best results).

Distance is interesting with a profit showing on races less than 1300m and again over 1600m (up to 2200m) but a bit of a black hole in between.

The Metro/Non-Metro distinction was also interesting with significantly better results for non-metro meetings - a contrast to the TopWeights.

Breakdown by State seems to give very interesting results with ACT/NSW and Qld all showing a good profit, SA & TAS breaking even, WA a loss of around 8% and Vic a loss of around 33%.

I also analysed by State and Metro grouping and found that :
NSW Non-Metro (includes ACT) gave good profit
NSW Metro gave a loss
QLD Metro gave a profit
QLD Non-Metro gave a loss
SA Metro small profit
SA Non-Metro small loss
VIC both gave signicant loss

So for the year in question (2002) you would have made a significant profit by backing the number 2 horse in NSW/ACT non-metro and QLD metro meetings where the top weight was carrying at least 2.5 kg more than the number 2. Based on these rules there were 324 selections, 63 winners (average price $8.11) for profit of $187 or 58%.

The problem is when I ran the same rules over the 2001 data I got a significant LOSS so I doubt if it will hold up.

Anyway best of luck to you.

__________________
"Computers can do that????" - Homer Simpson