Cheltenham Stats Question

M

Maruco

Super Moderator
Staff member
Joined
Jan 7, 2004
Messages
4,787
Location
The Shire
For those of you that hold stats, does anyone hold stats for returning winners and returning placers by number and percentage. What I'm trying to put together is hurdles to hurdles, hurdles to chases, chases to chases, handicaps to handicaps, graded to graded, graded to handicaps, handicaps to graded, and novices to open company. Plus anything else useful I haven't thought of. I don't have a baseline for this and it's a lot of work (that I'll do for next season if need be), and practically I don't have the time left to compile it.

I have a whole host of stuff I compile and update but I've realised I don't have anything on this and it strikes me as though it's something I should know and keep. If anyone can help I'd be most grateful.
 
I cannot help you Maruco with this but I will pass on many years of personal experience of horse racing. You can take the advice or ignore it as is your will. Past statistics of races totally unconnected to the form of todays horse form lines is as relevant as the colour of the horse, or what the first letter of the name is. You really are wasting a lot of your time and effort. That would seem a shame to me personally as i have found that time wasted in life is very valuable, i'd spend time on other aspects if asked.
 
Last edited:
I've never really been a stats follower, I say that in the conscious mind, because subconsciously we all do a degree of statistical elimination in our processes.
With the competition for places being so tight, most of the handicaps have become very constricted too.
This has caused some stats to morph over the last 5 years and getting ahead of these could be a slight edge.
Going to try doing a weighted average using some stats; I experimented with last years Pertemps and the results were startling.
If any one has any ideas as to what the breakdown of elements should consist of, moreover, what you think should weigh heavier, I'm all ears.
 
Last edited:
The Pertemps is historically my poorest race Max. In fact I've been eliminating stats this year rather than adding to them. The following are my current aide memoirs:

Don't usually win their qualifier. Probably 2nd, 3rd, or poss 4th, unless a cosy 5th or 6th. Must watch all qualifiers for quiet qualifying rides
The 2 from 19 that did win there qualifier ran in Feb qualifiers

If Irish came from Leopardstown qualifier. Usually a placed horse there will place here

Last six rated between 138 and 148 but overall quality is improving so cap out at 150

The market indicates to Ignore 7 yo and under at single figure prices

2 x 6yo recently, but 10 and above terrible

5yo terrible record unless Pipe which do place

If full range hcap no more than 11st 2lb. Where is the cut off likely to be this year, and where should the top weight consideration be?

Last time winners do well which suggests they may have needed to win to be sure of a run. Also maybe why the weight stat is important

JP runners do very well and seriously punch above there numbers

3 x selections. One on progressive form, one on stats, one that appears to have been hidden and running for a mark for at least 12 months
 
Last edited:
Tying yourself in knots over whether a 6yo or 7yo should be ruled out in these big handicaps seems like an awful waste of time.
 
My interpretation is that is that the reason is that most 7 year old hurdles are pretty exposed Slim. The check and balance on that is if they are lightly raced for some reason or they're an obvious plot job tumbling down the handicap from the right stable. The line above is simply a headline to make me check them so I don't tie myself in knots.
 
Last edited:
Remember Buena Vista and Willie Wumpkins lads ?

Some horses thrive on the place, the hustle and bustle , large fields and fast pace while others do not.

The previous Festival performances page in Racing Post is the one i pay most attention to.

Then again I am a low stake ew lucky 15 kind of guy who wants a good run at a price.
 
Maruco said:
My interpretation is that is that the reason is that most 7 year old hurdles are pretty exposed Slim. The check and balance on that is if they are lightly raced for some reason or they're an obvious plot job tumbling down the handicap from the right stable. The line above is simply a headline to make me check them so I don't tie myself in knots.
Click to expand...

Can you name me a stat that you've mentioned that isn't factored in by the market? When clowns like Tom Lee come on the TV quoting stats I'm physically ill.
 
Last edited:
Nothing like doing your own research.

I decided on doing a weighted average because you can mix and match the various parts of the elements within the data, and match them with whole elements. For example; I can apply different weights for the various ages, and on the other hand; I can apply and compute weights for such things as; course form, ratings, owners, Breeding, etc, all within the same sum.

I have looked at the last 15 years in order to give a decent sample size. I have plotted the first 5 finishers in each race, reason being; the bookies are offering nearly 21% of the field to place, the prices are good, and it throws up a lot of outsiders in the market.

The race distance dropped one and a half furlongs down to 3 miles in 2005 when it moved to the stiffer New course.

I plotted 3 groups; all 15, the first 10, and the last 5, the latter in particular, because of the increased quality and competitive nature of the compressed handicap, and as I mentioned earlier; I think the trends are morphing.

I’ve read all the stats previously available and can blow some of them away as either false or most definitely misinterpreted.

I’ll begin by stating that in percentage terms, 5 year olds are best by a country mile, but as there are only 2; one won’t get in, the other won’t stay with a furlong start; so I’m going to completely eliminate them.

6 year olds have the most wins; 6 with 3 of them in the last 5 years. But their overall 5 place record reads only 21%; joint 4th with 8 year olds. 7 year olds are the worst of the 3 greater numerical contesters, that of; 6, 7, and 8.

10 year olds performed best in the first 10 year sample and 2nd best only to the 5 year olds in the last 5 years, and 2nd best overall. Discounting the 5 year olds leaves them clearly top. Only one entered and it's a Mullins plot!

Drawing conclusions from age alone is not my intentions here. As I said earlier; I want to do a weighted average using all or most of the, most relevant trends.

I’ve listed some of them below and would be grateful to anyone who offers an opinion in aiding order of importance, or who can add anything else of value.

I will post my final entire field finding on here asap.

Thanks in advance Max.

Age
Ground.
Irish Trained.
Particular Trainer.
Particular owner.
Course Form.
Distance Form.
Rating.
Price.
Beaten Fav last time out.
1st Time head gear.
Breeding.
Weight.
Profile specific.
Seasonal Runs.
Last time out winner.
Qualifiers.
Font Runner.
Days since last run.
 
Last edited:
The only stats I give consideration to are the trainer's record in the race, and horse's with previous Festival form (as alluded to by Eddie).

For me, other stats such as age, won LTO etc are just as likely to lead me away from the winner, as towards it. As you would expect, these stats are never used in isolation from form study.
 
Bets just aren't hard enough to find to bury yourself in stats that the market already factors in. I'd rather spend by time focusing on getting on.
 
There's a difference in looking for value and looking for winners. There's plenty of opportunity to guarantee profit at the festival with the extra e/w places on offer if that floats your boat. Others enjoy studying the form by whatever method and putting their opinion on the line.
 
TheBear said:
There's a difference in looking for value and looking for winners. There's plenty of opportunity to guarantee profit at the festival with the extra e/w places on offer if that floats your boat. Others enjoy studying the form by whatever method and putting their opinion on the line.
Click to expand...

I love people that study form, they find the best bets but there has to be a trade off. If you're tying yourself in knots going through ststs that are already factored into the price then whats the point? At the end of the day you have to look at the hourly rate for the over indulgence in stats.
 
Last edited:
I agree if you're working them out yourself but if someone's already done them for you then why not use the ones you deem significant? If that enables you to eliminate a third of the field you'll be a damn site quicker coming to a decision than someone sitting through video after video of every single runner.

Grassy says it could turn you away from the winner which I agree with but I just accept that as collateral damage. I can't find every winner.

What I will say is that combining the right stats with form study has drastically improved my strike rates in the handicaps.
 
I suppose it's a matter of taste.

If I'm beat, I want it to be because I missed something in the form, the I've subsequently identified. That's a useful lesson I can take away, and apply next time. I don't want it to be because I've excluded 6yos because they have a poor record.

Slim's right. The market must factor in these stats to a degree anyway, and if that's the case, then you are potentially spending time looking for an edge that isn't there.
 
Last edited:
The Four days of the Festival has been exceptionally good to me over the years. The reason is because I put a considerable amount of work in and I enjoy doing it which is a big part of it. That's hardly tying myself in knots.

As for the stats having no edge and them being factored into the market, that may be partially true. But when they throw up multiple win and place horses every season it becomes less relevant. I'll also add that I keep my own stats and never read anything else available. I'm sure there are plenty of stats that I use that cross over to others that are freely available, but equally I'm sure I have several others that don't, and therefore i do have an edge.

I know many disagree with a statistical approach, but how can I agree with them when it works for me. Or maybe I just get lucky every year! The bottom line is there are multiple approaches you can take to finding winners, and whatever approach turns a good profit surely can't and shouldn't be knocked.
 
Grasshopper said:
If I'm beat, I want it to be because I missed something in the form, the I've subsequently identified. That's a useful lesson I can take away, and apply next time. I don't want it to be because I've excluded 6yos because they have a poor record.
Click to expand...

I can assure you that anyone who keeps stats and uses them properly would never do that Nick.

You use them in combination with form study and you take an interpretation in the context of the race and each individual horse. And I disagree with Bear as I would never use them to quickly eliminate a third of the field.

Max made the perfect observation above in the respect that you are looking to profile both win and place horses. Why would you want to ignore a 33/1 placed just to back an obvious a 3/1 winner. The truth is you want to back both.
 
Maruco said:
The Four days of the Festival has been exceptionally good to me over the years. The reason is because I put a considerable amount of work in and I enjoy doing it which is a big part of it. That's hardly tying myself in knots.

As for the stats having no edge and them being factored into the market, that may be partially true. But when they throw up multiple win and place horses every season it becomes less relevant. I'll also add that I keep my own stats and never read anything else available. I'm sure there are plenty of stats that I use that cross over to others that are freely available, but equally I'm sure I have several others that don't, and therefore i do have an edge.

I know many disagree with a statistical approach, but how can I agree with them when it works for me. Or maybe I just get lucky every year! The bottom line is there are multiple approaches you can take to finding winners, and whatever approach turns a good profit surely can't and shouldn't be knocked.
Click to expand...

I don't have a problem with much of what you're saying here, you're essentially building your own model. Your pocket will answer the question of how predictive it is against the market. I think great bets can be sourced for almost nothing these days. I'd teach someone starting out now how to read and beat markets rather than read form which would not have been the case twenty years ago.
 
Last edited:
Maybe I'm inefficient in going through the card but if I was to study the form of every runner then I'd be breaching that hourly rate rule that Slim alludes to. I can't bring myself to look at the handicaps until the day before because it would just take me so long to do it and I'm not really losing any value. The prices are normally best on the morning of the race with the extra places on offer.

I apply my method and come up with a shortlist without even looking at the market. If I then see a strong favourite which I'd ruled out through trends/stats early I may take a 2nd look but unless I find something compelling I feel that by applying that method my shortlist has a good chance of having found value.

What isn't value is hanging onto the coattails of these talking horses such as The Storyteller, Tombstone et al.
 
TheBear said:
What isn't value is hanging onto the coattails of these talking horses such as The Storyteller, Tombstone et al.
Click to expand...

I completely agree with that. All the best bets next week will be in handicaps because those markets are completely inefficient right now.
 
One other thing to add. There are plenty of non-race specific stats to overlay. Eddie mentioned returning winners so I'll use that as an example. There are 40 of them this year. They won't all win. I do think as many as 8 will, and possibly 13 will place. If I'm right that's more than half of them. It's even more significant when you consider that's 21 horse in 28 races. In other words a 75% race strike rate. In fact it's slightly better than that because you can take out the Bumper, Triumph, and Fred Winter, so make that 21 in 25 races and an 84% race strike rate.

The flip side is there will be approaching 500 horses running next week. What will the strike rate be for the remaining 460 by comparison! Surely that's a stat you want on your side when you're analysing races? I've just used one stat there and it's pretty compelling. There are many more.
 
Maruco said:
One other thing to add. There are plenty of non-race specific stats to overlay. Eddie mentioned returning winners so I'll use that as an example. There are 40 of them this year. They won't all win. I do think as many as 8 will, and possibly 13 will place. If I'm right that's more than half of them. It's even more significant when you consider that's 21 horse in 28 races. In other words a 75% race strike rate. In fact it's slightly better than that because you can take out the Bumper, Triumph, and Fred Winter, so make that 21 in 25 races and an 84% race strike rate.

The flip side is there will be approaching 500 horses running next week. What will the strike rate be for the remaining 460 by comparison! Surely that's a stat you want on your side when you're analysing races? I've just used one stat there and it's pretty compelling. There are many more.
Click to expand...

So previous years winners will be missed in the market? No selection process no matter the strike rate can work if you're not taking the right price.
 
Last edited:
Back
Top