I'm looking for the wisdom of experience on the question of how we select what races to punt on and how we determine which races to avoid. Obviously, some races are tougher to find a winner in than others. I have used the rough n ready guide of no more than 14 runners to hone in on the easier races but I'm interested in other methods. (And then there's the problem that the easier the race the less the dividends...)

Here's a thought on rating races by degree of difficulty:

Refer to the Zip form scores in the Sportsman Chartform.What is the difference between the top Zip and the bottom Zip?

Eg. If the top scored Zip in the race = 56 and the weakest roughie is rated as a Zip = 48, the difference = 8.

Now divide that number into the number of runners.

If the Zip diff. = 8 and there are 16 runners the result = 2.0.

Anything higher than two is a bad race to pick (or set your own cut-off)

I figure that if the Zip form analysts and all their computer power can't tell 'em apart, how can I?

Some examples from Sat. 8th June 2002.

Race 2 Eagle Farm.

Top zip = 55
Bottom zip = 48
Diff = 7
Runners = 20
7/20 = 2.85

NO BET

Race 3 Eagle Farm

Top zip = 58
Bottom Zip = 49
Diff = 9
Runners = 16
9/16 = 1.77

WORTH A BET

Any thoughts or other methods?

Cheers

I start by looking at the top half dozen horses in a race, if I can't narrow those 6 down to 2 or 3 horses after putting them thru my selection criteria, I avoid the race.

In other words if I consider that there are more than 3 horses with a strong winning chance, I usually don't bet.

CLASS

Bet on Blacktype races.
Listed /Grp 1/ Grp 2/ Grp 3.

Avoid Maidens and lesser class races.


Then you know that whatever you back it has some ability and you can get good odds about class horses in bad form which surprise!

__________________
Treat your selection and staking methods not with optimism, nor with pessimism, but with realism.

thanks.

I have been doing something of this already, but in a haphazard way. I haven't considered class enough, though. I just tend to avoid races where too many horses have too few career races to give meaningful stats.

What about the problem of easy race = poor dividends. Should I be looking for the easy races? OK, if I'm using a system that backs favourites, but there's no value for money in it.

Does anyone deliberately seek out races that are hard to pick on the theory they'll pay better if you do get a winner? To what extent should one look at the money anyway?

I have a few rules for selecting races and horses.
* You must have a reliable rating system to rate the fields. I use a personal spreadsheet based on times - subscription services or Techform are OK as long as they are consistent.
* Stick to sprint races only (<1600m). Times are almost meaningless in staying races.
* Stay away from wet tracks - slow or worse.
* Eliminate 2YO races. Too young, too green and not enough exposed form.
* Convert your ratings to prices, as a bookie does - total market = 120%.
* Back the value bets, where there is a substantial margin between your price and the market. One bet, straight out, per race - no each way or place bets.
* Best tip of all - record all your bets in a spreadsheet and analyse the results - continuously.

Couldn't agree more Rain Lover, if you take anything away with you from this forum - this may be the single most important thing!

Thanks for the backup, EI. Most punters haven't got a clue about whether they're in front or behind over time. Or whether novelty bets are good value or net.
If they had the true figures in front of them they would be amazed.

Thanks for the backup, EI. Most punters haven't got a clue about whether they're in front or behind over time. Or whether novelty bets are good value or not.
If they had the true figures in front of them they would be amazed.

Thanks for contributions. And especially Brumby for the info on how they calculate the Zip. Always wondered. But I have found the Zip a good guide. Often you can spend hours rating horses and come up with much the same results as the Zip analysts. I read of an analyst, looking at 2 years of results, who found the Zip star the second most reliable guide to a winner (after both best av. prize/av.places horses). Useful. As explained above I've been using it to select races. I also avoid races where two horses get the zip star. Again, if the zip analysts can't separate them, how can I?

I don't have a "system" as such but the "rules" for me are looking like:

*Only metro races, weekends, public holidays etc.
*No sprints or hurdles.Prefer 1200-2000m races.
*No more than 14 runners
*No 2yo, maidens etc.
*Zip diff. (see above) of less than 2.0.
*No races with two Zip starred horses

This still leaves lots of races but hopefully weeds out the tricky ones. Any more filters and the number of eligible races starts to become a problem. Sometimes I eliminate races with more than two last start winners as well but this cuts out some good races. The problem is to find highly specific filters that don't impact too severely on volume of races.