Hi All,

There is a theory that says by the time we come to the last race of the day, a large majority of punters (let's call them mugs) are looking to "get out" and recover previous losses. To do that they tend to steer away from favourites and bet more other non-favourite runners. The theory states that this then increases the average price of favourites in those races (while their strike rate doesn't decrease) and provides a better result relative to other favourites.

Overall market efficiency issues are of great interest to me and I believe there may also been some more formal research on this phenonemon outside of Australia. I may end up turning this topic into a broader article on my website at which time I will hunt down that info.

Anyway, onto the topic at hand.

The initial stats I posted on another thread showed the following:

Strike rate and profit of favourites at NSW TAB dividends for NSW, QLD, VIC, SA Saturday Metro racing from Mar 99 to End Jan 03.

Race SR% POT%
1 - 33.9% : -14.8%
2 - 33.0% : -17.5%
3 - 31.6% : -11.1%
4 - 29.7% : -12.4%
5 - 30.6% : -9.8%
6 - 27.5% : -15.7%
7 - 24.9% : -18.6%
8 - 28.8% : -2.9%
9 - 34.3% : +31.6%
10 - 9.1% : -54.5%

Race 10 only had 11 ocurrences for one winner so the POT% figures is a little misleading.

Based on these upfront stats, there certainly seems a bias to facvourites being more profitable in the later races.

If we combine races 8,9,10 the result is 244 winners from 843 bets (28.9% SR) and a loss on turnover of just 0.8% or $67 if you were using $10 flat bets.

Obviously each of these races aren't the last of the day as in everycase there is a race 9, then race 8 isn't the last of the day. However since i'm unable to split them in that way and since there isn't a huge occurence of meetings with 9 races, I think it's appropriate that we can consider the group of races 8-10 as being the last race of the day.

If we compare this to races 1 to 7 we see the following:

Races 1-7: 30.1% SR and -14.3% POT
Races 8-10: 28.9% SR and -0.8% POT

Quite interesting?

It would seem that the theory might have some legs....or as with many statistical patterns it could just be coincidence with no validity in the assumption that it will continue.

Either way, Focusing back on races 8-10 I thought I would explore further.

Statistics show that no matter which way you look at it, 2&3YO's are by far the worse value of any favourites...I feel it's because their form and ability is so obvious that everyone jumps on, (shortening the price) as well the hype factor so many young horses have placed upon them (further deteriorating value).

So I thought I would see what happens if we eliminate these historically poor value favourites.

With no 2&3YO's the stats for favourites in races 8-10 show:

Selections: 673
Winners: 200 (29.7%)
Win POT%: +4.1%
Place POT% +1.9%

It's worth pointing out that the longest losing sequence was still 16..plenty enough to send all you staking plan types broke :smile:

The comparable results for races 1-7 are a 28.8%SR and -11.9% POT. Still a big difference.

So with two simple rules - Favourite in the last race of the day but eliminate the historically bad value 2&3YO's you have a system that BASED ON PAST RESULTS has shown a small profit at TAB prices.

I do have data from Feb03 to April 03 on another PC and when opportunity permits I will run the same over that information and see what is thrown up.

There are other rules from these two that can increase the profit further, but I might save them for another post or a more detailed article for my site.

Interested in any comments people have.







[ This Message was edited by: osulldj on 2003-05-21 17:11 ]

[ This Message was edited by: osulldj on 2003-05-21 17:16 ]

Hi Osull
Where is your web site? :smile:

djosull
Well, Silly me. I've just worked it out.
What a coincidence.we have corresponded before.

do those poor figures for 3y/os refer to races for 3y/os or 3yos in general ie races for 3yo and up & wfa

I REMEMBER RUNNING THROUGH SOME FIGURES WITH PAUL SEGAR SOME YEARS AGO.
THE GENERAL CONSENSUS WAS THAT FAV BACKERS SHOULD GO HOME AFTER RACE R4.
I DIDNT AGREE WITH IT THEN AND DONT AGREE WITH IT NOW.
THE LINE OF FAVORITISM CAN CHANGE WITHIN THE LAST FEW SECONDS OF A RACE WITH A MAJOR PLONK.

http://www.adrianmassey.com/flfav/index.htm

The forensic analysis on UK racing found above tells a different story from the Australian experience.

40,000 races from 12 years of flat racing yields:

Last Race 31% win 60% place -12% PoT

First Race 35% win 65% place -7% PoT

Furthermore, the Last Race was most definitely the very worse race of any race on the UK race programme on which to support the favourite. It seems the UK 'mug punter' psyche is very different to ours.