Thanks for that.
I hadn't considered that the bettors at one track would be better than those at any other track.
Another gem picked up at the forum (subject to several days programming and data searching to test it out of course).
Thanks again.

KV

I'f be curious to know what you came up with KV, I have not the programming expertise to perform that kind of task. Any results would be greatly appreciated.

Duritz.

Sure thing, I'll let you know soon as I have anything interesting.

KV

Now way the punters are better at one track than another....if you consider they are all the same people then it would be in fact the track it self that produces more favourite winners.

If that is the case, and it sounds like a pretty sound argument to me, is it because some tracks have shorter straights (more favorites get boxed in on the turn or whatever)? Is it because more favorites get nobbled at some tracks? What is the reason? And if it is pure unpredictability rather than less astute punters can we really take any advantage from the different returns at different tracks or are we as likely to suffer the consequences of unpredictability as everyone else.

Every trck has it's own features and some horses run better at some tracks.....it is a very hard thing unless you study the tarck and the way it races....but here is one pointer...when looking in the form guide the horse with a "C" by it's name could be one to check out if it is in form

Flemington is rarely biased, Caulfield is rarely not. That makes Flemington a better place to do the form, and it means horses who don't deserve to win can win at Caulfield if on the right part of the track. That makes an analysts job harder at Caulfield than at Flemington.

We were talking about putting a dollar on each horse in each race at meetings some time ago and what the returns meant, well I've come up with a few results to mull over.

These are returns over the last 10 years on the NSW TAB if you plonked a dollar on pretty much every horse in pretty much every race (I say pretty much because data on the net just ain't perfect).

Over the 10 years 79.0% Returned overall (i.e.$79 back for every $100 bet).

1995 - 2001 is fairly consistent around the 79.6% area then progresses to 77.8% by 2004.
I guess the punters are getting more accurate.

There is only 2% between the highest and lowest yearly return so it's pretty consistent.

What puts the meaning of this ROI in perspective for me is the returns under different track conditions.

Avg Var from
Year to year
Fast Tracks 79.0% + or - 5%
Good Tracks 77.8% + or - 2%
Dead Tracks 79.6% + or - 5%
Slow Tracks 82.0% + or - 4%
Heavy Tracks 84.8% + or - 5%

You can see the punter does a more efficient job on Good tracks than any other sort. Heavy tracks overall are only 7% different and as many people say leave the punt alone on wet tracks it shows what a difference that 7% makes.

Metro tracks: 78.7%
Non Metro: 79.2%
Not much difference here.

1 to 8 Runners: 80.0%
9 to 12 Runners: 79.3%
13 to 25 Runners: 78.2%
This is a strange one. Does it mean we bet more efficiently in bigger fields? I'll be interested to hear anyones explanation of this one.

Canterbury: 77.5%
Rosehill: 78.1%
Randwick (Kens): 75.0%
Royal Randwick: 78.9%
Warwick Farm: 76.9%
Caulfield: 79.95%
Flemington: 82.7%
Moonee Valley: 82.4%
Ascot: 80.1%
Belmont: 77.0%
Eagle Farm: 76.95%
Doomben: 77.6%
Morphetville: 81.4%
Sandown: 80.05%
Hillside(03-04): 86.0%
Lakeside(03-04): 75.0%

6% ish difference from easiest track to hardest. Almost as much as Good track to Heavy track.

Well, Duritz said Flemington was a better betting prospect than Caulfield and the figures bear him out. We are 3% better at our job as punters at Flemington than Caulfield. About the same as we are better on Good tracks rather than Slow tracks.

Returns by the three TABs 1/8/02 to 31/12/04.

NSW Tabco UniTab
All courses: 78.0 77.6 77.3
NSW Races: 78.0 77.7 78.2
QLD Races: 77.9 77.9 76.4
WA Races: 77.0 75.8 75.9

Hope you find this interesting. Any ideas how we can use it to improve our betting?

KV

Use your imaginations with the headings above the figures. They shifted in the move from white page to green page. Sorry.

It is well known racing history that some tracks are harder to win on than others.

It has more to do with track layout than track bias [a different thing].

Cauifield is not much more than a trotting track [329m straight] that they tried [and failed] to fix and Sandown had it's hill. Canterbury is tiny [straight also only just over 300m] and hard to punt on. Ranwick is almost square and is notorious for taking your money too.

Basically size really does matter here. The larger tracks create a more even playing field [Brisbane tracks for instance but the huge race fields have to be avoided].