I read with interest an article written in this months edition of PPM by columnist E.J Mimmis about weights.

He was replying to a question which asked how do horses that have been beaten by a certain horse get a significant weight pull in there next meeting but that certain horse still beats it.

His reply was that he thought the bigger the rise in weight from last start the bigger the advantage.

He went on to say "... weight was one of a number of factors that i researched , finding a weight increase a far more positive factor than a weight decrease , as the follwing statistics , which are based on hundreds of thousands of individual horses runs , show:

Horses dropping weight by 3.5kg or more: 6.2% strike rate.

Horses dropping between o.5kg and 3kg:
7.4% strike rate.

Horses carrying the same weight:
7.8% strike rate.

Horses rising in weight between 0.5kg and 3kg:
11% strike rate.

Horses rising in weight by 3.5kg or more:
13.9% strike rate."

I thought those statistics were so unbelievable at first i had to re-read it to make sure i was reading it right.

In my ratings approach i gave horses rising in weight by 2.5kg or more a penalty while giving horses dropping by more then 2.5kg a bonus , i think i will be re-thinking that approach very closely now.

What does everybody else think about those statistics?

testarossa,
interesting stats....there are varied opinions as to the effects of weight...stats show topweights win the most number of starts (as a ratio to the number of starters) and in fact an english writer called russell clark believes the disadvantages of weight are much overrated....he cites his own stats on english races which show much the same trend as you did....on the other hand i have followed with interest neil's weekly stats on horses carrying 59kg or more...they distinctly show a marked loss if you back these horses ....i wonder what the stats would be with topweights carrying less than 59kg??

Below is a UK article backed by comprehensive statistics confirming that, from one race to the next, horses going up in weight win more races.

http://www.flatstats.co.uk/art****************************_racing.html

I am can confirm that this is correct on strike rate.

I am currently doing an analysis myself on weights and the impact of them over three years of saturday metro racing, when I finish, I'll post the results here.
It won't be for at least a week as I'm only a quarter way through it and it's very time consuming, but already this pattern seems correct.

E.J. has some great statistics he uses but some of his conclusions leave a lot to be desired given his statements on the Ausrace message system.

I have for a long time now largely discounted the effect of weight and since that time my profit has improved, partially because I believe weight doesn't have the effect people believe, but perhaps more importantly that everyone else still considers it important, which creates windows of value opportunity for me sometimes.

I can support EJ's stats in PPM with my own work from some time ago...which if I present them in the same categories as him show:

(Based on 3 years of NSW/QLD/VIC/SA Metro racing).

* < -3.5kg's = 6.5% SR and -25.5%POT

* -0.5 to -3.5kg's = 8% SR and -22.9%POT

* Carrying the same weight: 8.5% SR and -17.9%POT

* +0.5 to 3.5kg's =11.5% SR and -19.4%POT

* >+3.5kg's = 13.6% SR and -19.7%POT

Statistically speaking, EJ is of course right, a weight increase is a positive statistical factor and a decrease is a negative factor. EJ was answering a specific reader question and did so in an excellent manner but excluding that, these figures about weight change in isolation of other factors add no value to the handicapping process. They do nothing to help us make profit.

In racing, weight change takes place in the context of class changes i.e. easier or more difficult opposition from race to race. The reason then that horses rising in weight win more isn't because extra weight speeds them up, its because more often than not it means they are contesting an easier (or at worst same) level of race as previous. Taking this further, the reason horses higher in the weights win more is purely because they are better than the opposition, not because having a higher weight makes you run better.

So weight cannot be considered in isolation in a handicapping process and I believe anyone who applies penalties or bonuses based on weight changes or amnount carried alone is making a fundamental error in race and horse assessment.

The real issue in racing is whether the horse has the ability to win at the given level of race. If you are assessing weight it must be more in the context of the change in class and the horses ability to compete at that level.

If a horse is dropping in weight, it should not receive a bonus nor an autmoatic penalty because the stats say its a negative factor.

The view we must take is to assess whether the horse has the ability to win at that level of class, if it doesn't, then no amount of weight reduction will help it so why apply a bonus? If it has proven capable of winning at that class with the same or more weight in the past, then why would we apply a penalty for its weight reduction from last run?

If a horse is rising in weight and racing in a much easier class then we might conclude that it certainly has the ability to win and the weight will not be a hinderance. However if the horse is racing in similar class and only just managed to win last time, maybe the increase in weight is enough to bring it back to the field. Why would we apply a bonus in such a situation? Because the stats say a rise in weight is a positive factor and irrespective of other things, an increase in weight somehow magically improves a horses competitive ability in the upcoming race? I don't think so!

As you can see there is no easy answer, it depends on the context of the race and individual horse. That's why those who put faith in the search for statistical answers will in the long run always come up short. Not that I mind :smile:

They're my views anyway, interested in others comments. If time permits I will make another posting that examines class and weight changes when viewed together.

EI, Unless you wish to, there is no need to go to painstaking amounts of work and time to research such statistics.

If you would like to see in anything in particular then I am happy to help and save you some of that work (without becoming your person data bunny :smile: I can analyse these type of things over thousands of past races with win, place and profits in a matter of minutes.

Just post what you would like and I will do my best.

Thanks osulldj,

In fact your posting could not be truer and your assessment of statistics and how you apply them more accurate!

You cannot apply statistics solely on strike rate with out considering at least a few other variables.

Each race, in my opinion has to be treated as an individual event, but apply statistics based on the "weight" (pardon the pun) you give to each one.

The impact of weight over various distances has to be considered as well. Weight over 1000m or 1200m is vastly different to the impact of weight over 1800 or 2400m. This is where statistics can guide you in the wrong direction.

The reason that many of the topweights fail is not because of weight.
It is because they are weighted based on their best form not on their recent form. So you get out of form horses being given top weights and performing poorly.

They are then assigned topweight because they drop down a little in grade and are not at their peak formwise.

The ones that do win are usually in good form but have had luckless runs or beaten by a smaller margin.

So the stats are not a good guide if you only consider the raw data.
If you were to consider topweights which were beaten by less than two lengths at their previous start, over various distances I think you'd be get a better guide to the impact of weight.

As always it's the bigger picture to consider.

Thanks for your offer ossulldj, but I'm doing a rather involved study. It would be unfair to ask for such detail. In other words I am considering weight with other variables.

For example:
1. Strike rate of each weight per race.
2. Strike rate of each weight per runner.
3. Strike rate of both of the above over every individual distance.
4. Strike rate of all the above over sex.
5. Strike rate of all the above over age.
6. Rises and falls in weight for each of the above.
It just goes on and on.

It's more for my own curiousity than anything else as I'm kind of a stats nut and look at how best stats can be refined and applied to be useful.

Here are a few points to consider:

1. An increase in weight will NEVER improve a horse's performance.

2. An increase in weight will ALWAYS reduce a horse's performance.

3. The opposite applies for a decrease in weight.

4. The issue is whether the weight increase fully penalises the horse for the drop in class.

5. Similarly the issue is whether the weight decrease fully compensates the horse for the rise in class.

6. Some of the best long term bets may well be improving horses rising in class and dropping in weight.

[ This Message was edited by: Neil on 2002-11-29 12:37 ]

Stats on wt increase.

0-3.5kg=28%SR -19% Loss on turnover
4-5kg=40%SR +26.6% Profit on yurn over
5.5-16kg 36%SR -6.8% Loss on turnover

__________________
Cheers.