Impact Values

[Lescoach]

I am currently using Impact Vales to rate all runners in a race and also to produce an odds line.
Has anyone got any thoughts and does anyone no where i can get further information on this subject.
Thanks

[Lescoach]

Impact Values are the statistical relative weights assigned to various handicapping factors. They are calculated by dividing the percentage of winners with a certain characteristic by the percentage of actual starters with that characteristic. I am using 8 factors at the moment but i would like some more e.g I V"s for Track conditions, Barriers if there are any.
HELP

[enjay]

Equestrian Pub PPM Magazine some years ago had a listing ofseveral values. I do have the ;mags stored away somewhere and can find if ;you need them......

[enjay]

Equestrian Pub PPM Magazine some years ago had a listing ofseveral values. I do have the ;mags stored away somewhere and can find if ;you need them......

[Lescoach]

I got mine from Equestrian in 2000 but i think they did some before that but unsure

[TheDuck]

If it's statistics your using and you have the actual raw data, factor analysis might be more appropriate. It tells you the significance of the data points but in relation to each other. It will also tell you if those 8 you are looking at now are important to begin with, what their relevance is to each other and so on.

You then create a discriminant function or do some clustering to find out which group of factors, in which relative combination, most often represent winning horses.

You can then create a probability model that will help you determine when to wager on this group.

Or you could do like my wife does and pick the cute jockey and/or funny-named horse.

-Duck

Hi, I've got a home made program that uses impact values to select the most likely horse. I take into acount barriers, weight, win %, place %, prize money, track & distance record and track condition etc.

For all this I get around a 30% strike rate and if I blindly bet on the selections I get a long term gradual lose over the past 3 years.

For the preceding 4 years it gave a good profit at the same strike rate. so why the change? My only conclusion is that with the dramatic increase of computor systems to make selections, these statistically most likely selections are more heaverly backed than before, thus reducing the dividend. A check of starting prices bears this out with the average now around $2.90 down from $4.60

I have made a couple of additional conditions to my selection which if passed I place the bet. This has resulted in a return to profit, though still not to the previous levels.

My advice is to keep written records as well as computor files. Computors crash and discs lose data over time. Then at least once a year review your selections, results for patterns, good & bad, make adjustments and check back over your records to see if that change has long term merit.

I have observed various short term trends or patterns which may be the basis for some of the commercially (scam) software products. I had one pattern a couple of years ago I spotted just as it stopped. One of the bottom 2 rated horses were getting up 1 in 12 races with average div of $52. This meant for a $24 outlay I could of got back $52 = $28 profit. I watched and noted last season if it continued before I pumped money into TAB/Bookie pockets. Thankfully my caution was well founded as I would of lost big time.

Good punting,
Horse Cents

[hermes]

I was reading an article somewhere a while back about the impact of the computer revolution on punting. Had an impact definitely, but not a huge one, it was saying. In fact the comforting thing in the article was that all the computer power in the world has only managed to budge the punt a millimetre or two, mainly by draining profits out of favourites. The punt was a very tight thing already, so its been made more difficult, but there's still hope for the computer semi-literates.

The article - sorry don't recall the author etc. I only glanced at it - said the impact of computers reached its limit by about 1996-7. Further advances in computers haven't enhanced their tipping performance greatly. Extra gigs don't help. Then there was stuff about chaos factors etc. explaining why there are real limits to the predictability of a race anyway.

Can the mathematically minded tell me just how unpredictable is a horse race?

I gather that the best computers in the world, with the best software and the best brains behind it, selecting on all standard races, hits a wall somewhere just over a third winners. Why?

The zip rating system in the Sportsman is "objective" computer-driven handicapping. Gets under a third winners (but worth following because its a better paying third than market favourites).

Hermes

[Equine Investor]

On 2002-08-08 14:54, hermes wrote:
Can the mathematically minded tell me just how unpredictable is a horse race?
Hermes


:lol: Ever tried to pick the winner?

Seriously though, your point is extremely pertinent to what every punter is trying to achieve. The inherent problem with any form analysis, computer program etc is this number one hurdle...

The best horse in the race does not win with a high enough strike rate because of various factors which cannot be measured.

How fit is fit?
How does the horse / jockey feel today?
Interference?
Bad Luck (As a separate entity to interference).
The list just goes on and on.

Until we invent something that will take into account every variable and assign a value to it, software programs and the like can only be a tool to help you assess available data and make an educated guess.

[becareful]

Hermes,

I think the answer to that is that horse racing is too unpredictable to ever expect more than 30% winners if you take a "pick the winner" approach. One of the big problems is that any system must be based on predicting the future events based on what has happened in the past - we all know that that doesn't always work! You are always going to get winners who have vastly improved on their previous efforts that systems will usually miss (usually the really good paying ones!). Also you often get horses that should win on paper but run like they are carrying 100kg. And of course there are horses like those that Equine Investor follows where their true ability is not reflected in the recent results - again any system based purely on these statistics will generally miss these horses.

This is why I believe that to be successful you must try to look for Overs rather than trying to pick the absolute winner - find those horses that have a better chance of winning than the odds that are on offer. Picking the winners 30% of the time will send you broke when the average div is only $2.50. If you can find the horses paying $8 when they should be $5, or $15 when they should be $10 or $20 instead of $15 then you can make money - of course you wont win every race, or even every day but you will win in the long run.

[TheDuck]

Good points. I would like to echo two.

First, the 'soft' factors. I don't have enough past performance data so I can't provide any precision but the point is well made. Because the numbers are easy to deal with we do our analysis on those. And we've seen the numbers aren't the whole story. It's unreliable to look for high/low numbers. Patterns are often more important. And they require more work to derive. We also forget to use all of the numbers. For example, we report 30%, 70%, etc. But we don't report variance, error rate, and so on. If I told you I guarantee 50% strike with 15% POT but 700% variance the swings would destroy your bank.

Also, I have found an odd trend in the objective of computer programs. They seem to be trying to predict the win. Even the papers submitted by (apprently) super-intelligent scientists on citeseer and from universities are doing this. We aren't trying to pick the winner. We don't really care which horse wins, it's US that wants to win. If you review strategies by punters there seems to be a lot of spreading the bets around, focusing on place rather than win, tactics like dutching and so on. We really don't care who comes out ahead on the track as long as we come out ahead at the wicket (or with the bookies, as the case may be).

Well, I think that should be enough rambling for now. :smile:

-Duck

P.S. If anybody does have any historical data I would love to crunch it. Especially the soft stuff like 'rallied 3 wide' and other things that people are afraid to statistically analyze with numerical computers. If you're curious about how to analyze this stuff an easy to read book (well, easier than regular statistics books) is "Marketing Research: Methodological Foundations".