From the Turf Accountant website I mentioned in another post, he lists what he says are the ten most reliable selection methods. He claims that this is the result of working through a large database over many, many years. Here they are:

1. DOUBLE TOPS.
Using the Sportsman Chartform mark the Top Place % and the Top AvPrize. A double top is where the same horse is either top of both or equal top.

2. ZIPFORM
The horse or horses with the "Zipform" * from the Sportsman.

3. COURSE/DISTANCE QUALIFIED.
To assess distance and course qualified horses add win % to Place % at course /distance 100 or above are qualified. e.g. 4 minor places from 4 starts qualified. 1 wins and 2 minor placings from 5 starts (20+60) not qualified but 2 wins and 1 minor placing from 5 starts (40+60) is qualified.

4. WOT.
From Wizard mark the 100% horses in the wrat and trat columns and mark the horse assessed at the lowest or equal lowest odds. A WOT horse is one marked for all three.

5. TURFIES TOP RATED.
Simply multiply the most recent finishing position by 5 - the next most recent finishing position by 3 - the next by 2 and the next by 1 and add them up the horse with the lowest or equal lowest total is the qualifier.

6. OCI'S.
Using the Sportsman Chartform mark the Top Three Place % and the Top three AvPrize including equal thirds. An OCI - Only Co-Incider is where one and only one horse in top three of both columns.

7. NINES QUALIFIED.
Simply total the best three finishing positions of last four starts any with a total of 9 or less qualify.

8. LAST START WINNERS.
Self explanatory.

9. FORM PATTERN.
Last four starts must show constant improvement or at least level ending in either Win or Second, each start other than fourth last must be in first 9.
e.g. 0882 qualifies 0021 does not. 6642 qualifies 6543 does not 0711 qualifies 1121 does not.

10. WIZARD FORM.
From Wizard check best form last 12 months and best recent form - only horses that top both these polls qualify.


These, he says, are the ten most reliable pointers to lots of winners (although not necessarily decent priced winners). Seems to contradict other stats I have seen sometimes, but some useful ideas here I think. Food for thought. Check out this old website:

http://tucows.ecopost.com.au/~turfacts/turfacts.html

Cheers

Hermes

Some of this has merit. One concern though - multiplying nos by this and that.

When you start doing that you end up with some odd nos. Lets say a horse ran 8th last start - 8*5 =40. So what? it may have only been beaten 2 lengths and was unlucky - yet this method has given this horse a poor rating.

What about the form improvement, consistent improvers? Good to find a horse on the improve. What's a quick way to establish that? So you think, Chief, that form numerals for last races eg. 4061 are suspect because that 6th second last race might have only been by a length in a blanket finish? True, but for convenience, don't we have to forgo too much detail and expect that any form of analysis is, at best, only an indicator, not an exact science. Agree about the multiplication of numbers though. It gets very abstract and remote from what its supposed to be about, namely horses.

Forget form pattern rules - bollocks.
A 8th placing may have been a better run than a win. Might & Power ran 8th in the Epsom (certainty beaten) then came out and won the Caulfield Cup by 6 lengths.

I use relative simple rules, but they are not "simplistic" like some of these.

The way to use simple rules is to have several rules for one criteria. ie you could have 3 simple rules just for a horse's last start performance. The combination of the 3 rules provides a powerful filter, yet is based on logic and not abstract like these multiplying rules.

There are exceptions to every rule - you can also catch some of these winners by looking at what these exceptions could possible be - so if a horse doesn't fit the rule but falls into one of the exception areas don't eliminate it if meets your other criteria.

Quote
"5. TURFIES TOP RATED.
Simply multiply the most recent finishing position by 5 - the next most recent finishing position by 3 - the next by 2 and the next by 1 and add them up the horse with the lowest or equal lowest total is the qualifier."

FAULTS - DOESN'T TAKE INTO ACCOUNT - CLASS, BAD LUCK, BEATEN MARGIN. TOO ABSTRACT.



Quote
"7. NINES QUALIFIED.
Simply total the best three finishing positions of last four starts any with a total of 9 or less qualify."

FAULT - DOESN'T TAKE INTO ACCOUNT CLASS, BAD LUCK, WEIGHTS ETC.... WHAT IF THE HORSE RAN 3 1STS CARRYING 50-52KGS, BUT NOW HAS TO CARRYING 59KGS??



Quote:
"8. LAST START WINNERS.
Self explanatory."

FAULT - ELIMINATES ABOUT 50% OF WINNERS. I DON'T HAVE EXACT FIGURES HERE, BUT PLENTY OF WINNERS DIDN'T WIN LAST START.



Quote:
"9. FORM PATTERN.
Last four starts must show constant improvement or at least level ending in either Win or Second, each start other than fourth last must be in first 9.
e.g. 0882 qualifies 0021 does not. 6642 qualifies 6543 does not 0711 qualifies 1121 does not."

WHAT'S WRONG WITH 1121??? LOOKS CONSISTENT TO ME???

FAULTS - MISLEADING, DUE TO CLASS FACTORS, WEIGHT, BEATEN MARGIN, BAD LUCK ETC...

The criteria from Sportsman and Wizard may have some merit. Readers of those papers may be able to tell you.





[ This Message was edited by: chief on 2002-06-23 13:54 ]

Chief wrote:

"A 8th placing may have been a better run than a win."

Really? But 8th is 8th and 1st is 1st. It must mean something if a horse wins. 4th isn't 3rd, either. It must surely mean something if a horse can get in the money. Isn't a miss as good as a mile? And aren't there horses that don't know how to win, so to speak? They run well, look nice but more often than not run a 2nd. Conversely, people tell me some horses have a good understanding of what a race is about and know what winning is. If I'm given the choice between a winner and an unlucky 8th, I'd prefer the winner. And what does "unlucky" mean here?

Are you suggesting Chief that I stop giving too much validity to finishing places, and start looking at how the horses ran, not just where they finished? But the name of the game is winners (or placegetters at least). If I look at how a horse ran last few starts, rather than its capacity to put together enough to win (or place), mightn't I select horses that run well but don't necessarily win? What factors should I consider? Speed in last 600? Lengths from the winner? Again, when is a horse "unlucky"?

Do what you like. I don't care if you lose money. Only trying to help.

Hermes...

You are quite right when you are holding a losing ticket, 4th is 4th and 2nd is 2nd.

But I think what chief was trying to get you to see...is that numbers on paper are not the be all and end all. This is why most computer software to predict winners doesn't work. There are variables in horse racing.

For instance, if a horse runs second 8 lengths from the winner....isn't a horse that runs 10th placing only 4 lengths from the winner a better horse / run?
On paper it can be deceiving, and so can some formguides / stats.

Looking at statistics ALONE doesn't work.

Also a horse may have been blocked for a run, got bumped in the running by another horse, cast a plate and slipped on the home turn and flashed home for second placing. That is a lot better run than a soft run second from another horse.

When assessing form and stats for a horse always take into account the overall picture such as win / place percentages, prizemoney, form over the distance and in the conditions. Always forgive a horse a few bad runs as long as it doesn't become a pattern in it's form. Look at trials formguides, and stewards reports, also trackwork times if your horse runs a well below average run.

[ This Message was edited by: Equine Investor on 2002-06-23 21:08 ]

The point is any one stat doesn't tell the full picture. On Saturday 15 out of 22 flat winners in Bris, Syd and Melb did not win their last start. Therefore, if you eliminated horses just because they didn't win last start - you miss out on a heap of winners. What one must do using stats is combine them and then they become powerful.

Horse A - Ran 8th last start, but won two starts back, drops 3.5kg from last run where it was only beaten 1.8 lengths after being trapped 3 wide without cover. Is also racing over its pet distance today - 1200m at Rosehill for instance after running over 1100m last start which is short of its best. Has barrier 5. Odds of 6/1.

Horse B - Won 3 starts in a row, but is rising 4.5kgs, has barrier 15 in a 1200m race at Rosehill, dropping back from 1400m to 1200m. Odds of 7/2.

I know which horse I would rather bet on!!!

All advice is greatly appreciated and heeded. I take the point. Stats are one dimensional and a horse race is a 3D thing. The attraction with the numbers, I suppose, is time. If you are pushed for time on Friday nights/Sat mornings you need some reliable shortcuts. But I realise any shortcut can be no substitution for proper analysis. And Mr Turfaccountant, whose methods I was looking at, makes claims for them based, he says, on large-sample database trials (albeit several years ago now). I think the idea would be to combine various methods as so many filters. The triok would be finding a combination of factors that didn't filter out too many races.

Thanks again

Hermes