starting to construct a trifecta system using ranked average prizemoney as the selection method first to sixth. would like to know of what number combinations would be suitable. was thinking of first leg 1&2
2nd leg 2,3,4,5 3rd leg 3,4,5,6
all feedback is welcome

I think the combinations depend on the prices of the horses. Eg. If the top horse was $2.00 I wouldn't put it in unless the others were well over $10.00.
Tri. combinations I take are based upon horses I give decent chances to that tote punters leave out. If I take combinations popular with tote punters I will get big unders.

hi
As a guide one could use the place dividend to find the value for 2nd and 3rd.

$4.00 the place approx. or above. With trifectas you must find value as long runs of outs can cost you plenty up front.

I rather miss the small ones and get the bigger dividends there is a formula you can use that will give you the approximate return for a dollar for your trifecta combination.
x(xy-1)(z+1)
______________
x+1

Lets say we have three horses say 7/1 15/1 60/1 lets assume the 7/1 wins the dividend will be different if the 15/1 comes second to the 60/1 coming second.

This is because alot of punters dont put 60/1 horses for second. So the pool is the same say $30000 if 50 punters get it with the 15/1 coming second then the dividend is $600 but if the 60/1 comes second and only 10 punters get it then the dividend is $3000.

If you can find the value for second thats the key.

regards
ubetido

deleted

[ This Message was edited by: puntz on 2003-02-07 23:35 ]

thank you for your opinions
puntz, as for the book by malcolm knowles i think that came up recently on his second hand list, it went very quickly and unfortunately it was not me. i will keep my eyes out and let you know if i located one or two cheers

Try Garry Robinson at Horse Racing Australia, he keeps a lot of Malcolm Knowles books and systems. http://www.horseracingaustralia.hl....rget=BOOKS.html

ubetido

The trifecta formula suggested above is theoretically correct but in practice is seriously flawed.

If the odds of a horse are long odds on then the chances of that horse finishing third are substantially over estimated by the formula. Odds on chances don't run third very often, they win more than 50% races, run second a lot and so on. It is sometimes the case where, depending on the odds structure of the race, a 20/1 chance has the same chance of running third as the long odds on runner. But the given trifecta formula does not reflect such a reality.

Moral of the storey - dont put short priced favourites in for 2nd or 3rd in trifecta combinations as they represent poor value.

Disclaimer - I'm no expert.

thankyou for your response Dr. pangloss i do agree with you.

What do you think is a correct trifecta structure.

The formula i showed is the one Don Scott showed in his book winning way i think or one of the other ones.

regards
ubetido

I do not bet on Trifecta or First4 (so feel free to ignore this!) but I noticed something yesterday that may be worth considering. Have a look at the results for the Metropolitan (SR6) on NSW. The first 2 positions were filled by "value" runners (Victory Smile at $33 and Piachay at $4.20 the place), the 2nd favourite was 3rd (Dress Circle at about $5 win quote) and favourite was 4th (Red Trinket at about $3.80 win quote).

The Exotic Divs for the race were:
Exacta $358.80
Trifecta $3204.90
First4 $20336.80

Now if you assume that the win price is a close approximation of each runners chances in the race you can work out what the true dividends should be (adjusting the win price to remove the TAB "cut" and then expressing it as a percentage). Doing this you get:
Exacta (7-19) = $565.90
Trifecta (7-19-2) = $3064
First4 (7-19-2-18) = $10083

As you can see the Exacta paid was a big under given the long odds of the runners. The Trifecta was over the true odds by about 5% but the First4 paid double what it should have done as the chance of the favourite running 4th given the first 3 finishers was about 30%.

Maybe there is something to be said for leaving the favourites out of the first 2 placings in your Trifecta's/First4's and including them as the 3rd/4th selection? You certainly won't get many winners but when you pick one up it will probably be a good payout.

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

Ubetido

Allow me the liberty of re-phrasing your question to read, "What is the INCORRECT trifecta structure."

I'm not going to masquerade as some sort of latter day trifecta expert and so rely on statistics contained in 'Gambling into the Ninties - A better guide to odds', Hans EISLER 1990 Kangaroo Press. I know not of Mr Eisler's fate except to say as a mathametician and keen student of the punt he pioneered the application of computers in unravelling the mysteries of the turf.

Our friend crunched the results of 5,346 metropolitan races and compiled what he termed, "The Form Guide to ODDS".

The ODDS form guide revealed the success rate of Starting Prices calibrated from 1/3 through 100/1. For example, horses with SP 5/4 won 39.9% of races - thus if you could secure 6/4 a punter would break even in the long run. At SP 5/2 - needed 13/4 to break even. At 6/1 - 8/1 and so on and so on. The longer the SP odds, the harder it became to secure the break even odds at some stage during the betting.

He examined 306 races where the favourite SP was 5/4 or less. Using the favourite (F) as a roving banker with the field(X) he recorded the following trifecta results:

F,X,X $32,564 bet $25,856 return 20.6% LoT
X,F,X $32,564 bet $23,804 return 26.9% LoT
X,X,F $32,564 bet $23,902 return 26.6% LoT

Betting F on a straight out win only basis showed a 9.8% LoT.

Conclusion: What one should NOT DO in structuring a trifecta bet is transparent.

A successful alternative remains less obvious - at least to my eye.