Any Mathematicians here??????

[redfloyd]

Trying to come up with a 1 to 10 rating figure on how each horse finished its race regardless of placing. This is just for a 1 to 10 rating on how the horse finished, so for example:
<O:p</O:p

Example 1: 18 horses in field 800m 400m Place Margin
17 11 2 0.5
Example 2: 10 horses in field 800m 400m Place Margin
9 8 1 0
<O:p</O:p
Example 3: 22 horses in field 800m 400m Place Margin
2 7 12 6.2
Example 4: 24 horses in field 800m 400m Place Margin
2 2 1 0
I am trying to come up with a formula which will rate each horses finishing so I can apply it to the rest of my program. To date, this is the only part of my program which is open to subjective input - something I am trying to eliminate.

Are there any mathematicians here who may come up witht he formula needed so i can apply it to my program?

[lomaca]

Trying to come up with a 1 to 10 rating figure on how each horse finished its race regardless of placing. This is just for a 1 to 10 rating on how the horse finished, so for example:
<O:p</O:p

Example 1: 18 horses in field 800m 400m Place Margin
17 11 2 0.5
Example 2: 10 horses in field 800m 400m Place Margin
9 8 1 0
<O:p</O:p
Example 3: 22 horses in field 800m 400m Place Margin
2 7 12 6.2
Example 4: 24 horses in field 800m 400m Place Margin
2 2 1 0
I am trying to come up with a formula which will rate each horses finishing so I can apply it to the rest of my program. To date, this is the only part of my program which is open to subjective input - something I am trying to eliminate.

Are there any mathematicians here who may come up witht he formula needed so i can apply it to my program?

Hi!
I am sure I could help but I maybe missing something, could you please make it a bit clearer?
Like what the 800 and 400m mean? sectional positions? if so, should'nt be two of them?
Cheers

[lomaca]

Hi!
I am sure I could help but I maybe missing something, could you please make it a bit clearer?
Like what the 800 and 400m mean? sectional positions? if so, should'nt be two of them?
Cheers
Sorry!
Upon reading your post again, I see what you mean, it's just the formatting of the post that threw me.
Still not clear though, without knowing the position of the other horses it is hard to put a number on one horse without comparing it to the rest!?

Will consider the problem.
Good luck
Ps.
If you are serious about it, give us an Email address.

[Chrome Prince]

Hmmm, it's a little open to subjectiveness.

An 800m,1 400m,1 finished 2 margin 2.50L

Is this worse or better than your example?

Or do you want to rate them on lengths made up?

[crash]

I'm sure such a rating number in a horse's last race could be dredged up to suit the distance parameters required, but I'm trying to work out [considering the amount of things during each particular race that would effect individual runners at various distances], what possible use such a rating number could be. Especially taking into account a horse's particular running style, jockey, barrier position and at what track etc. etc. that will all influence the outcome of a race for each horse in it and what position it was in at various distances in a race.

For instance, at the 800m and the 400m a winning backmarker [eventual winner of race] will be about last and a losing leader [eventual last] will be about 1st.
In the last 400m their positions will go through a rapid transformation covering almost every position in the field from first to last. The leader will be covering them as it looses ground and the backmarker will be covering them as it gains ground. In fact this could all happen within the last 300m, 200m or even the last 100m. First to last and last to first.

There might be some use for such a rating figure for on-pace grinders [only], but even that would be of iffy use.

Of far better possible use which could be give a rating number, would be sectional times, but even these figures [suitable to USA dirt tracks and Aust. trots] have debatable worth on Australian race tracks, considering very few of them have the technology involved to gather that info. for a start and also the widely diverse layout of our tracks and movable barriers.

[system]

shouldnt he just use horses last 800,or 400 sectional? he,s only looking for melb cup lead up races.and only betting in the cup itself.
i think red is only looking for horses finishing off there races well over shorter distance,s like 2000m 2400m

[woof43]

Hi,
The following method is relatively crude, compared to what is possible. (... think of how you can use this technique not only to correct Calls, BUT ALSO BY Class) But I would also continue to think about the ramifications of changing the data domain.

First set yourself up a table of OBSERVED historical probabilities for getting any given call value. Express them as PERFORMANCE PERCENTILES. So in my example, you end up with eight percentile floor cutoffs, like (a made-up example):

Call 1 = 82.2%
Call 2 = 71.1%
Call 3 = 50.7%
Call 4 = 41.3%
Call 5 = 30.8%
Call 6 = 21.7%
Call 7 = 09.5%
Call 8 = 00.0%

Now, for any given formline where you have a horse and he scored a Call of 3, you know he performed AT THE 50.7 PERCENTILE LEVEL OR ABOVE. So throw away his original Call, and substitute his PERCENTILE score. Do that for all of a horses lines. Add up the percentile scores, and divide by the number of lines you used, and you get the horses AVERAGE PERCENTILE SCORE, which you use as a CORRECTED AVERAGE Call

To project the 800 Call on TODAY'S race, simply get the percentile rating for all horses, put them in order, and give the highest horse a projected break of 1. Next highest a projected break of 2, and so on, just using a simple RANKING of the PERCENTILE AVERAGES.

Important Note: To get a more accurate rating for each horse, don't take the FLOOR percentile for that call, but take the mid-point between the floor and the ceiling. For example, if a Horse has a Formline with a call of 3, don't use 50.7 for his score, instead use ((50.7+71.1)/2) which is the midpoint of the percentile step.

The "trick" here is that you are NOT getting a correction for each formline. Instead, you are getting a corrected OVERALL average for the Horse by changing the domain of your data from CALLS to PERCENTILE PERFORMANCE RATINGS.

And finally, and OBVIOUSLY, this general technique is pretty valuable when applied to other problems of this type commonly found in the handicapping world.

[KennyVictor]

I just KNOW there's gold in what you say Woof, if only I could understand it.

KV

[Chrome Prince]

okay, I'll say it.....

What is a call?

racecall?

[lomaca]

Hi,
The following method is relatively crude, compared to what is possible. (... think of how you can use this technique not only to correct Calls, BUT ALSO BY Class) But I would also continue to think about the ramifications of changing the data domain.

First set yourself up a table of OBSERVED historical probabilities for getting any given call value. Express them as PERFORMANCE PERCENTILES. So in my example, you end up with eight percentile floor cutoffs, like (a made-up example):

Call 1 = 82.2%
Call 2 = 71.1%
Call 3 = 50.7%
Call 4 = 41.3%
Call 5 = 30.8%
Call 6 = 21.7%
Call 7 = 09.5%
Call 8 = 00.0%

Now, for any given formline where you have a horse and he scored a Call of 3, you know he performed AT THE 50.7 PERCENTILE LEVEL OR ABOVE. So throw away his original Call, and substitute his PERCENTILE score. Do that for all of a horses lines. Add up the percentile scores, and divide by the number of lines you used, and you get the horses AVERAGE PERCENTILE SCORE, which you use as a CORRECTED AVERAGE Call

To project the 800 Call on TODAY'S race, simply get the percentile rating for all horses, put them in order, and give the highest horse a projected break of 1. Next highest a projected break of 2, and so on, just using a simple RANKING of the PERCENTILE AVERAGES.

Important Note: To get a more accurate rating for each horse, don't take the FLOOR percentile for that call, but take the mid-point between the floor and the ceiling. For example, if a Horse has a Formline with a call of 3, don't use 50.7 for his score, instead use ((50.7+71.1)/2) which is the midpoint of the percentile step.

The "trick" here is that you are NOT getting a correction for each formline. Instead, you are getting a corrected OVERALL average for the Horse by changing the domain of your data from CALLS to PERCENTILE PERFORMANCE RATINGS.

And finally, and OBVIOUSLY, this general technique is pretty valuable when applied to other problems of this type commonly found in the handicapping world.

Hi!
I was thinking along the same lines, but I also included "Class" since I thought it to be one of the main factors. Have some results but so inconclusive as to be utterly useless.
Have to agree with Crash, that running style, ( I don't have this data and never use it) is probably more important in this instance than anything else.
Next of importance would be the barrier pos. - track condition, combination (hence riding tactics) and that's just to start with, in other words we are looking at an endless array of probable outcomes for little or no return.
I think, that sometimes intuition and experience in handicapping is the best course of action.
One example: yesterday at Sandown race 4, there was a horse called "LION DREAM" form line 0094, nothing to write home about, but if you look at it's last performance on a heavy track and the fact it was obviously improving (0-0-9-4), this was a horse to put some money on, now this decision will never come from systems or even ratings. I am happy to say I had 5*50 on it. And only the ability of a superior jockey beat it to second place.
Good luck

[lomaca]

I just KNOW there's gold in what you say Woof, if only I could understand it.

KV
Hi KV!
I may be way off here, but I think I did the same as Woof, Looked at all the races where the sectionals were available, and created a table where the relationship of finishing positions to sectional positions were written as a percentage of occurance, like how many pos. 5 at 800m and at pos. 3 at 400 m finished first, sec. etc. etc.

Then I did the same including the class of race as well.
I examined the results, but sadly they led nowhere.
Then I gave it away.
As they say, if you don't succed for a couple times give it a rest! No use flogging a dead horse.
Cheers

[woof43]

Hi
When modelling how the Public wager, one WOULD use the Classification of each race, this is something the Public sees, but in Handicapping you need to use a Statistical description, in the example provided I would classify the race by the average Standard deviation of all runners.

The other important part as you mentioned is the accumulated physical Track bias which is present as we get closer to the finish...in my opinion i would be looking at little closer to the start if i was using going to use performance percentiles

[Wunfluova]

What is a call?

American terminology for sectional markers. Their calls would equate to our positions in running at the 800m, 400m etc.

I am a little confused here as I remember Woof warning against averaging 'in running' data for greyhound racing because it was 'ordinal data' yet as far as I can see this is a similar exercise for horse racing. No doubt he will explain the differences in due course.

[lomaca]

Hi
When modelling how the Public wager, one WOULD use the Classification of each race, this is something the Public sees, but in Handicapping you need to use a Statistical description, in the example provided I would classify the race by the average Standard deviation of all runners.

The other important part as you mentioned is the accumulated physical Track bias which is present as we get closer to the finish...in my opinion i would be looking at little closer to the start if i was using going to use performance percentiles

I agree.
Good luck

[lomaca]

American terminology for sectional markers. Their calls would equate to our positions in running at the 800m, 400m etc.

I am a little confused here as I remember Woof warning against averaging 'in running' data for greyhound racing because it was 'ordinal data' yet as far as I can see this is a similar exercise for horse racing. No doubt he will explain the differences in due course.
HI!
There a no monkeys on the dogs backs anymore! (no jocks!)

[Wunfluova]

There a no monkeys on the dogs backs anymore! (no jocks!) No, most of them are in the betting ring ripping up tickets. http://forums.ozmium.com.au/images/icons/icon7.gif

[Horse Whisperer]

I dont know if any of he other guys really covered this but another flaw in getting a rating based on the finishing is the pace at which a race is run, a slow pace obvioulsy isnt going to favour the back markers and a fast pace is not going to suit a front runner. Also another thing that cant be factored in is whether a horse is blocked for a run coming down the straight or if the track has any bias in it or for that matter the length of the straight and how tight the last bend is, for example the difference between Doomben & Eagle Farm. I think there is to many factors to take into account before you could before you could confidently come up with a mathematical rating to represent a horses finish.

[woof43]

Technically yes it is, the major reason not to use the Calls in greyhound racing was an is because of the bias in the person recording the "official" calls, especially the "jump" call, calls in greyhound racing if used need to be corrected track by track

But back to your original question in horse racing they might just be discrete? Or maybe even poorly sampled (or poorly measured!) continuous data? If you throw calls in the discrete category, then it's perfectly fine to do averaging. And the same would go for poorly measured continuous data.

[lomaca]

Also another thing that cant be factored in is whether a horse is blocked for a run coming down the straight
Hi!
Tell me about it,
Today Gosford race 3,
My selections were: Buzzy Henry, Anextra, Master Magic, Son of Kintyre.
Put 2*20.00 on Anextra and had a quinella with it for the other three.
What happens? major interference in the straight, maybe it would not have won,
but it definitely would have ran second.
OK I got my place divi of $4.10, after protest, but I missed out on the quinella???
Is there justice in racing??
Yes we get it in our favour as often as not!
Good luck.

[Chrome Prince]

American terminology for sectional markers. Their calls would equate to our positions in running at the 800m, 400m etc.

I am a little confused here as I remember Woof warning against averaging 'in running' data for greyhound racing because it was 'ordinal data' yet as far as I can see this is a similar exercise for horse racing. No doubt he will explain the differences in due course.

Thanks for that.
Cheers.

[crash]

'Calls' in horse race? As long as they are racing on a totting track or greyhound track [set-up for it], no problem. At a lot of tracks you are flat out getting an '800m' call. Where they were at the 'turn' is as good at it gets on most tracks except the major metro's and where they were at the turn is everything to do with class of race, jockey's ability and/or luck in a race, track layout, conditions, distance, running style, and barrier draw to suit it [or not].

More sobering is to have a look at the form lines of many winners for their last 4 runs. There is often a lot of 5,6,7,8,9,0 in them due to many, many variables as previously mentioned in some posts here, that often have nothing to do with ability. I'd be more wary of 2 straight wins [a big rating figure for most stat. methods] in the last 2 starts than 2 unplaced runs. Like glue-on shoes, very few horses win with either attached to their names at their next start :-))

I think if a lot of punters put a 1/10th of their energy into learning how to handicap as they put into gathering 'stats' and trying to come up with winning rating systems [often over a LOT of years], they would already be ahead of this game. Sure, messing with numbers will bag you some winners [so will my aunt Molly throwing darts], but common ****** sense and a bit of reasoning, will come up with a lot more.

[Chrome Prince]

The stats mean diddly unless you KNOW the running style of the horse.

As always, it's what you do with the stats, how you apply them and more importantly how you understand them which leads to success or failure.

The problem is that the majority of punters base any findings on limited stats and misinterpret the findings.

Just have a look at some of the systems bandied round based on one or two longpriced winners or not enough data.

I would hazard to guess that less than 5% of the punters that do gather stats, know how misleading stats can be, UNLESS based on logic, large data samples and interrogated correctly.

200 races is not even close to providing accurate reflections, unless you have identified a trend which makes up more than 10% of the profit. (I say trend, not profit - meaning 10% of the runners are overlays).

There are ways to ensure that the system or stats is likely to be a hiccup or an ebbing/flowing short term result, or a longterm money spinner.

[Chrome Prince]

If you are looking for in running positions the Queensland racing website has a great pdf format racebook which gives 4 positions in running, also a handy head to head account of previous matchups.

[redfloyd]

Thanx to all the replies.


At the end of the day, I am just trying to get a rating of 1 to 10 on each horse's finish in a particular race. I'm not using this as a system to bet on, but more as a standard to grade each horses finish.

So for example:

800m 400m Margin Place
12 3 0.5 2

You could look at that horse and say: well thats a finish of an 8/10

then another horse will be:

800m 400m Margin Place
2 9 7.5 15

and you'd say: that horse had a finish of 3/10

Not sure if that makes sense or not, but it's that rating out of ten that i want to be able to find a fomula for so that it's a standard across the board??

If you'd like more information, my email address is: redfloyd1@hotmail.com

Cheers
Red

[redfloyd]

The formating is out, can't help that, so if you want email me, please do?