My kennel

[retriever]

Beware of what you wish for because the poor draw stats topic has just bitten me in the bum. I have a possible bet for Wen and this mornings 48 hr decs show a negative draw. I will post below some of the notes i have made concerning this.

Wen 4th Kemp 8.0 : Got No Dollars @

A 7 fur 0-70 with 12 run. Dollars shows top off my ratings @ +1 and has the looks of one placed to win. His recent form 0-7 may have included some gaming because all of those runs where in the OR seventies which looks an ask too much, but the BHA have finally relented and dropped him to OR 69 which is still 2lbs above his highest winning mark but based on some of his placed form runs is def doable.

These double digit size fields appear to suit with all four previous wins in same and the profile suggests that his 49 day return should not prevent. Two of his wins have been at Kemp with the most recent coming over this CD Jan last year. 5lb claimer Luke Catton was shown as booked at the 4 day stage and his allowance takes the sting out of the gross 9-8. He was also on top for Dollars most recent win.

The animal has won an 8.0 pm winter race at Kemp so no concerns there and the stable appear in good nick with eight winners during Dec. As ever there are niggles involved with the main one being his 6 box which emulating via 12 run - Hcaps shows a worst of the lot 4% S/r but his previous CD win was when drawn 5-12 and his 16Mar22 CD run when drawn 6-12 resulted in him being done only half a len in a 0-80 so subject to price i may chance my arm.

I'll have a look at that race later when HRB brings up the race.
Usually it would fall out of my scope as I generally confine myself to 9 to 11 runners only.
.

 

[mick]

No worries retriever the only reason i shared my early doors thinking was because i feel the best way to example something is via a relevant pre race selection.

 

[ARAZI91]

Hi mick Thanks for your input. The help I seek? That was a while ago, and I think I have progressed a bit since that post, although we are always learning, aren't we!!



Hi ARAZI91 I am probably going to disappoint you here as what I do regarding the 'micro-draw' angle is very much 'hands-on' and race specific.
I wish I could develop data banks and statistics with computers and programming but I wouldn't know where to start. I have a fair understanding of spreadsheets and love trying to work formulas. I use OpenOffice but need to get Excel.

Regarding the draw I do it via HRB race by race. The bias ( if any ) comes down to stall positioning in relation to the bends over the distance. All weather tracks work better than the Flat as there's obviously more data.

As an example of what I look for lets take a race tomorrow :- 2.00 Chelmsford ( 9 runners )

I would first look at this System Builder category having put in the race parameters :- 'Position In Stalls'

View attachment 126776

This gives :-

View attachment 126777

Now you can see the bias here to the low numbers which is what you would expect at Chelmsford......LH anti-clockwise track...with the horses grabbing the rail. This just gives me a 'picture' to use along with other factors, obviously, Age, Weight, OR, H/g, Sex, etc. The odds market stays out of it until the end where I look for the best value - usually from my shortlist of three.

Now I know HRB doesn't adjust for non-runners and from the above we can see an anomaly with Stall 7. Looking further by highlighting it we see :-
View attachment 126778

This shows in part the shuffling due to non-runners. However it is still quite high in that area, so with a further look, I see that if a fancied runner is running from that area I know not to discount it too early - especially if it is able to break fast and get a front-running position.

Usually it's the case that the higher draw numbers will see more shuffling than the low numbers. Stall 1 will see none I would say. Where can it go?
If stall 2 is a non-runner they are not going to move them all up. It will just become a vacant hole that the race loader can dive into quickly.

All tracks are different and I know we are dealing with small sample sizes but over a certain threshold ( say, double the number of runners.....9 runners......minimum 18 races ) a small bias might be seen. Some tracks, for example, may have the stall positioned at a tangent to the first bend and those in the centre of the stalls have the shortest distance to the bend.

This is what I mean by 'micro-draw' and it's very much 'hands-on' and not easily converted to numbers and stats. They do sometimes move the stall positioning ( different tractor drivers....lol ) on some of the wider tracks which means an 'advantage' is negated. Just have to watch for it.
Goes without saying - replay watching is a must.

And just for interest the above 2.00 Chelmsford race I got down to these three. Probably do a combination Exacta and pick one for a win bet.

View attachment 126781


Incidentally, Bobby On The Beat, raced on Saturday, I put up in the HOD Saturday Super Heinz thread working to similar lines.
Got 6/1......SP 11/4 ARAZI91 can verify.
.

Good post and not disappointing me at all retriever - i like your approach and is something i'm well aware of when doing the routine of replay work, i would venture to note that the opposite of your scenario - where a horse run's well against a perceived "bias" is also worth bearing in mind. Bend's on round track racing should always be a source of interest. Two areas in particular - those bends that are close to the start, hence making early position an even more crucial factor and those that are positioned before they turn into the straight, generally this is where the pace starts to quicken - any moves made here have a high energy cost, especially if a horse is wide.

 

[ARAZI91]

Hi mick Thanks for your input. The help I seek? That was a while ago, and I think I have progressed a bit since that post, although we are always learning, aren't we!!



Hi ARAZI91 I am probably going to disappoint you here as what I do regarding the 'micro-draw' angle is very much 'hands-on' and race specific.
I wish I could develop data banks and statistics with computers and programming but I wouldn't know where to start. I have a fair understanding of spreadsheets and love trying to work formulas. I use OpenOffice but need to get Excel.

Regarding the draw I do it via HRB race by race. The bias ( if any ) comes down to stall positioning in relation to the bends over the distance. All weather tracks work better than the Flat as there's obviously more data.

As an example of what I look for lets take a race tomorrow :- 2.00 Chelmsford ( 9 runners )

I would first look at this System Builder category having put in the race parameters :- 'Position In Stalls'

View attachment 126776

This gives :-

View attachment 126777

Now you can see the bias here to the low numbers which is what you would expect at Chelmsford......LH anti-clockwise track...with the horses grabbing the rail. This just gives me a 'picture' to use along with other factors, obviously, Age, Weight, OR, H/g, Sex, etc. The odds market stays out of it until the end where I look for the best value - usually from my shortlist of three.

Now I know HRB doesn't adjust for non-runners and from the above we can see an anomaly with Stall 7. Looking further by highlighting it we see :-
View attachment 126778

This shows in part the shuffling due to non-runners. However it is still quite high in that area, so with a further look, I see that if a fancied runner is running from that area I know not to discount it too early - especially if it is able to break fast and get a front-running position.

Usually it's the case that the higher draw numbers will see more shuffling than the low numbers. Stall 1 will see none I would say. Where can it go?
If stall 2 is a non-runner they are not going to move them all up. It will just become a vacant hole that the race loader can dive into quickly.



And just for interest the above 2.00 Chelmsford race I got down to these three. Probably do a combination Exacta and pick one for a win bet.

View attachment 126781


Incidentally, Bobby On The Beat, raced on Saturday, I put up in the HOD Saturday Super Heinz thread working to similar lines.
Got 6/1......SP 11/4 ARAZI91 can verify.
.

One point though ....If Stall one is a NR , stall 2 becomes stall 1, if stall 2 is a NR , stall 3 becomes stall 2 and yes they do move them all up. It's just a matter of loading them in and does not require any thought by the stall handlers really. Only time you'll see a vacant stall is a very late withdrawn horse usually down at the start.
I should imply that i mean the horses in the intended stalls, not the actual stalls themselves lol
There is one commercial data mob who claim they are the only ones who account for this and use it as a selling point - Flatstats - but that's a complete falsehood as mick has shown Raceform Interactive has a NR filter, Proform and also Timeform account for it as well - as for HRB, well there is so many great features on there, plus it would be very difficult for Chris to go back and re-code all the stalls - fortunately there is an easy fix if your willing to download the data from HRB into Excel -as if you think about it there is another number sequence in a race which 99.95% of the time will follow the same order as the "correct" stalls order - finishing position you just need to sort, copy and paste and then resort. From there you could use a pivot table to get "corrected" win % etc .

 

[retriever]

Thank you ARAZI91
I think that is what I've grabbed onto where I've thought they don't get moved up - if there has been a late withdrawal.
Just checked a couple of recent races where Stall 2 is a non-runner ( 13-12-22 6.45 Wolv. & 22-12-22 8.00 Southw. ) and yes, it goes 1,3,4,5 etc.

In a way, by using Position In Stalls I am half way there as 1st, 2nd, 3rd Inside & Widest should be pretty accurate because I'm fixing the number of runners ( which relates to the actual number that ran ). I just need to be aware that 4th, 5th, 6th and possibly 7th, 8th and 9th the figures could be a little fuzzed. I only do 9 to 11 runner races.

On the whole though I think it suffices as a generalised visual tool and there are many other factors that play a part.
.

 

[retriever]

if your willing to download the data from HRB into Excel -as if you think about it there is another number sequence in a race which 99.95% of the time will follow the same order as the "correct" stalls order - finishing position you just need to sort, copy and paste and then resort. From there you could use a pivot table to get "corrected" win % etc .

There's those words again - pivot table!! - - I need to learn more.
First though, I need to get a decent laptop ( my current one is driving me nuts! ) and a copy of Excel.
.

 

[mick]

There's those words again - pivot table!! - - I need to learn more.
.

Less is best as the table appears to pivot when you attempt to place your glass for yet another refill.

 

[markfinn]

One point though ....If Stall one is a NR , stall 2 becomes stall 1, if stall 2 is a NR , stall 3 becomes stall 2 and yes they do move them all up. It's just a matter of loading them in and does not require any thought by the stall handlers really. Only time you'll see a vacant stall is a very late withdrawn horse usually down at the start.
I should imply that i mean the horses in the intended stalls, not the actual stalls themselves lol
There is one commercial data mob who claim they are the only ones who account for this and use it as a selling point - Flatstats - but that's a complete falsehood as mick has shown Raceform Interactive has a NR filter, Proform and also Timeform account for it as well - as for HRB, well there is so many great features on there, plus it would be very difficult for Chris to go back and re-code all the stalls - fortunately there is an easy fix if your willing to download the data from HRB into Excel -as if you think about it there is another number sequence in a race which 99.95% of the time will follow the same order as the "correct" stalls order - finishing position you just need to sort, copy and paste and then resort. From there you could use a pivot table to get "corrected" win % etc .

Proform do adjust the stalls for records ,( I was not aware any other firm did so , BHA don’t or at least did not and when I asked they had no plans to do so) so for stats proform do have them but that’s no use on the day , you still need to adjust your on the day order - one thing I have sort of noticed it’s usually the lower weights that are withdrawn , whether that’s just in the races I look at or not I can not say . Good luck with the Blog retriever

 

[retriever]

Good luck with the Blog

It's only going to be bits here and there. It's a bit too much in the open for my liking. This forum has loads of members that don't sign in and loads that can read without registering. I like to help ( and get help ) where I can in my limited way but it is possible to lose an 'edge' by being too open.
I've already said too much - so this blog will self-destruct in 2 minutes !!
.

 

[mick]

retriever Re your above while i would not disagree on balance may i ask which of your edges would be lost by posting on this public part of the forum and how do you evidence this. I have read that Bookmakers encourage there employees to read racing forums but i suspect this is more to do with seeking patterns and connections Re people backing the same horse as opposed to gleaning winners, but if it is the latter i pride myself in providing plenty of disinformation.

Re lurkers in general they are so busy doing so that they do not have the time to get the best from any worth having finds. ? PS : I note Got No Dollars is 17/2 > 7/1 with WH and as this is your Blog i now expect you to find and expose the member responsible for rattling my market by asking Hills for a fiver at the bigger odds.

 

[retriever]

It's funny actually - your horse is the only one that has moved with WH according to Oddschecker.

And when the rest of the bookmakers price up Got No Dollars I bet they go 11/2 tops and WH fall into line. A couple of hundred lurkers to this forum and your good name will see to that!

I have no evidence but it stands to reason that if a system that has been winning for the last few years gets published and out in the open, it won't be long before those selections get over-bet, the value eroded and a loser overall. I have my eye on a promising trainer. This year I expect them to be on a level footing to BSP before any filtering. Next year ( 2024 ) they will probably get noticed and over-bet. But if a big song and dance was made about them this month or next that 'edge' would be eroded much sooner.
.

 

[ARAZI91]

There's those words again - pivot table!! - - I need to learn more.
First though, I need to get a decent laptop ( my current one is driving me nuts! ) and a copy of Excel.
.

Mate you can do everything that you need in open office/libre office - you just need to know some extra formula.
Google sheets has lots of functions and utility as well.
But i'll tell you.... get "RStudio for Dummies" or something, give yourself a week with the first few chapters and some good youtube tutorials and you'll only go back to Excel type software for displaying stuff. Once you realise how easy it is you will never look back.
Dive in and take the plunge.

 

[mick]

I have no evidence but it stands to reason that if a system that has been winning for the last few years gets published and out in the open, it won't be long before those selections get over-bet, the value eroded and a loser overall. I have my eye on a promising trainer. This year I expect them to be on a level footing to BSP before any filtering. Next year ( 2024 ) they will probably get noticed and over-bet. But if a big song and dance was made about them this month or next that 'edge' would be eroded much sooner.
.

With the bottom line for myself being its all about the price then your above makes for an interesting read regretfully these days i suspect the real edge is being able to obtain the price, although some contrarian or minority thinking as part of the selection process does increase the prospects of the price being offered in the first place. !

 

[markfinn]

With the bottom line for myself being its all about the price then your above makes for an interesting read regretfully these days i suspect the real edge is being able to obtain the price, although some contrarian or minority thinking as part of the selection process does increase the prospects of the price being offered in the first place. !

Just scanning the RP there are already 3 odds on antepost favs ( apparently there was only 1 this time last year) at the Cheltenham Bun festival.

Which can only be good news for short price Exchange layers ,

 

[retriever]

Mate you can do everything that you need in open office/libre office - you just need to know some extra formula.
Google sheets has lots of functions and utility as well.
But i'll tell you.... get "RStudio for Dummies" or something, give yourself a week with the first few chapters and some good youtube tutorials and you'll only go back to Excel type software for displaying stuff. Once you realise how easy it is you will never look back.
Dive in and take the plunge.

I've ordered the Dummies book ARAZI91 - should be with me by the weekend
I have always wanted to do more on coding etc. I know I have the aptitude for it.
.

 

[markfinn]

I've ordered the Dummies book ARAZI91 - should be with me by the weekend
I have always wanted to do more on coding etc. I know I have the aptitude for it.
.

What is RS standing for please

 

[retriever]

R stands for retriever! of course!!

I don't know markfinn - just the name of the programming language I suppose.



R is an open source statistical environment and programming language that has become very popular in varied fields for the management and analysis of data. R provides a wide array of statistical and graphical techniques, and has become the standard among statisticians for software development and data analysis. R For Dummies, 2nd Edition takes the intimidation out of working with R and arms you with the knowledge and know-how to master the programming language of choice among statisticians and data analysts worldwide.

EDIT : Just seen that a Ross and a Robert developed the language at the University of Auckland in NZ in the early 1990's.

Also, RS is for RStudio - a user interface editing tool.
.

 

[markfinn]

R stands for retriever! of course!!

I don't know markfinn - just the name of the programming language I suppose.

View attachment 126840

R is an open source statistical environment and programming language that has become very popular in varied fields for the management and analysis of data. R provides a wide array of statistical and graphical techniques, and has become the standard among statisticians for software development and data analysis. R For Dummies, 2nd Edition takes the intimidation out of working with R and arms you with the knowledge and know-how to master the programming language of choice among statisticians and data analysts worldwide.
.

Interesting good luck / let me know how you get on

 

[markfinn]

retriever - Great result for the Seagulls - Kop moaning as usual, the game is boring enough without his and others touchline antics

 

[gerry]

Had a look at your horse mick in 8 00 i had already bet one in the race last night at 3/1 now 9/4 sisters in the sky the reason i bet it was i thought last couple of runs where improvment yet dropping in mark for i consider improved runs. Also only won once with holly doyle on it and she hasnt been on it for while last time she rode it run good race round here so looks like trying day.
Your horse i would put in same catogory last couple races improved runs yet dropping mark so very interesting race.
sister in the skys finished behind yours three runs back which was got no dollars last race giving dollars 4lb now getting 4lb and got no run in that race but i feel sister in the skys has improved last couple runs and with holly back on looks fancied to me.
So will stick a little on forecast too.
Both have done better times here than that race they both run in when sister in the skys was fourth to dayman was slightly faster.
But got no dollars run behind biggles even faster and down 5lb but sister in skys down 13 lb from dayman race. and from distance beat gives sea the skys ed ge so lets hope both run big races.