Coral Marathon - Sandown Saturday - an odds compiling exercise

davidjohnson said:
It's the ground issue that is causing the disparity and why any automatic rating to odds converter will always have Opinion Poll much closer to Manighar than his current odds. Opinion Poll's best form has come on heavy, can he reproduce it under these faster circumstances, I don't know, but will take a punt at the 6/1.
Click to expand...

An automatic rating-to-odds converter could factor context in, obviously.

At present, I prefer to keep it basic and to use it as a starting point for a more subjective approach.
 
davidjohnson said:
Given his chance on adjusted figures, the jolly could win this and not affect his rating.

At the prices though, I have to side against him despite suggesting on timeform radio that he'd take all the beating. It could be immaterial if they take him out, but I just don't think he's 6 times at likely to win as Opinion Poll.
Click to expand...

The current odds suggest it is about 3.3 times more likely that Manighar will win than Opinion Poll. 48.5% against 14.7%.
 
J Alfred Prufrock said:
At present, I prefer to keep it basic and to use it as a starting point for a more subjective approach.
Click to expand...
If your primary goal is the ongoing assessment of horses then that makes very good sense, I'd imagine. I'd prefer to keep a record of both the initial assessment based purely on merit, plus an adjusted tissue which brings more subjective matters into play.
 
Funnily enough, that is precisely what I am doing this morning for US. The delights of comparing an automated odds line based on ratings with the existing and subjective Morning Line, and both with SP, is not quite enough to keep my attention from wandering...
 
J Alfred Prufrock said:
Funnily enough, that is precisely what I am doing this morning for US. The delights of comparing an automated odds line based on ratings with the existing and subjective Morning Line, and both with SP, is not quite enough to keep my attention from wandering...
Click to expand...
I feel for you! How far away is the finished product?
 
Manighar isn't an odds shot for me here..major problems with the ground..as also has Opinion Poll.

King Of Wands @ 9.8
Aajel @ 12.0

are both overpriced as i think the front two should be 5/2 & 5/1 on this ground..there isn't that much between them imo

i'll dutch both of those..great value imo :cool:

the fav is a deffo lay at 1.86..thats a cert :)
 
EC1 said:
Manighar isn't an odds shot for me here..major problems with the ground.
Click to expand...

Worth pointing out that Timeform reckoned the going for most of Manighar's wins last season was no softer than good. The French can't be trusted with the going. :p
 
EC1 said:
Manighar isn't an odds shot for me here..major problems with the ground..as also has Opinion Poll.

King Of Wands @ 9.8
Aajel @ 12.0

are both overpriced as i think the front two should be 5/2 & 5/1 on this ground..there isn't that much between them imo

i'll dutch both of those..great value imo :cool:

the fav is a deffo lay at 1.86..thats a cert :)
Click to expand...

Excellent call EC1.
 
Desert Orchid said:
As with Man Of Iron last week, the question for me is whether tomorrow is the target for Manighar. He's the best horse in the race - no question - and weighted to win comfortably but is he 'off'? Is the Ebor the target? The Melbourne Cup? (At the moment, his OR is just about on the money so any handicap strikes me as being out for the time being.)

I won't be touching the race with the proverbial.
Click to expand...
So what about Manighar then? :whistle:
 
How much of a role does the potential pace scenario of a race play in the formation of a book for you boys? I've only began to grasp relatively recently that pace is everything in races; it's also probably the most difficult aspect of evaluating a race to grasp I find.

Taking today's race, I was relatively interested in Aajel on the grounds that he might get his own way out in front; instead, he's set what was at the very least a decent gallop (suspicion the leaders went off a shade hard?) and it seems it probably paid to be held up.

Surely any ratings-based approach must take pace into account as well. The notion of a horse running to a high speed figure is surely contingent upon being given the circumstances to hit that figure? Any automated approach that does not take pace into account is likely to overrate the chances of a hold-up animal in a race that is unlikely to be run at a decent gallop.

Cracking thread btw ~ sort of insight that makes this forum what it is.
 
Last edited:
For me, the likely pace scenario is always my starting point. If I can't visualise how a race is likely to be run, I really don't think I've got much of a chance of identifying what is going to win.

Regarding ratings, what is less important imo is the actual rating itself, but the confidence behind it and being able to acknowledge whether it could be significantly higher/lower and always then being able to play this in relation to the odds available. It's why I don't think it is worth the time and effort of compiling your own personal ratings when given the time constraints, you can get them from timeform/racing post etc for a relatively modest fee. What is much more important is being able to interpret them, acknowledge and understand that 2 horses could be rated 124 in a race but being able to understand why one is 7/4 and one is 14/1.

As I've said before I like to use speed figures as an aid to form ratings to give an indication of the likely strength of form. On virtually my first day at Timeform I remember being told the closer to timefigure for the winner of the race to the form figure, the more reliable the form is likely to prove. As a rule of thumb, there's nothing I've seen since to make me alter that view. I'd be lost trying to work out what is worth what in a Lingfield maiden without them!
 
trackside528 said:
How much of a role does the potential pace scenario of a race play in the formation of a book for you boys? I've only began to grasp relatively recently that pace is everything in races; it's also probably the most difficult aspect of evaluating a race to grasp I find.

Taking today's race, I was relatively interested in Aajel on the grounds that he might get his own way out in front; instead, he's set what was at the very least a decent gallop (suspicion the leaders went off a shade hard?) and it seems it probably paid to be held up.

Surely any ratings-based approach must take pace into account as well. The notion of a horse running to a high speed figure is surely contingent upon being given the circumstances to hit that figure? Any automated approach that does not take pace into account is likely to overrate the chances of a hold-up animal in a race that is unlikely to be run at a decent gallop.

Cracking thread btw ~ sort of insight that makes this forum what it is.
Click to expand...

lol - you starting to think like me - you will have the stopwatch out next - you anorak :D

with Betfair in running - reading pace is a good way of laying off liabilities
 
davidjohnson said:
For me, the likely pace scenario is always my starting point. If I can't visualise how a race is likely to be run, I really don't think I've got much of a chance of identifying what is going to win.

Regarding ratings, what is less important imo is the actual rating itself, but the confidence behind it and being able to acknowledge whether it could be significantly higher/lower and always then being able to play this in relation to the odds available. It's why I don't think it is worth the time and effort of compiling your own personal ratings when given the time constraints, you can get them from timeform/racing post etc for a relatively modest fee. What is much more important is being able to interpret them, acknowledge and understand that 2 horses could be rated 124 in a race but being able to understand why one is 7/4 and one is 14/1.

As I've said before I like to use speed figures as an aid to form ratings to give an indication of the likely strength of form. On virtually my first day at Timeform I remember being told the closer to timefigure for the winner of the race to the form figure, the more reliable the form is likely to prove. As a rule of thumb, there's nothing I've seen since to make me alter that view. I'd be lost trying to work out what is worth what in a Lingfield maiden without them!
Click to expand...

that "same" rating scenario is very prevalent in handicap races - 6 horses may have same RPR..but one is 2/1 and another is 20/1

so ratings alone can only play a small part in forming a tissue
 
davidjohnson said:
Regarding ratings, what is less important imo is the actual rating itself, but the confidence behind it and being able to acknowledge whether it could be significantly higher/lower and always then being able to play this in relation to the odds available. It's why I don't think it is worth the time and effort of compiling your own personal ratings when given the time constraints, you can get them from timeform/racing post etc for a relatively modest fee.
Click to expand...
Excellent point, DJ.

I compile my own because I don't have confidence in commercial ratings. If you get hold of commercial ones then end up poring over the form for hours to see how accurate they might be, you're as well doing it yourself.

RPRs are a fair bit closer to my own ratings than Timeform's are, therefore it wouldn't be surprising to learn that I'd have more confidence in RPRs than in the black book.

However, my ratings are only part of my race analyses. I'm really looking for improvers. These are the profitable ones, the ones tht end up making a mockery of the ratings! :)
 
Stating that pace is highly important in determining the outcome of many races - something with which most clued-up punters would agree - is not the same as saying you can predict it with accuracy, though it is well worth trying, imo.

Anything that can be reduced to numbers - as many things in racing can be, including predicted pace scenarios and degree of confidence in ratings - can be fed into a ratings-to-odds model. Plenty of people are doing this (very) successfully already.

But a ratings-to-odds model is still likely to benefit from intelligent human intervention, from where I am at, at least.
 
another point about the figures

I dont only use it for betting, it makes you understand better the quality of what you have seen.
 
If anyone fancies having a go at pricing any of Windsor's races up tomorrow, we could have "Figure 2" for this discussion. I can't say I hugely enjoyed doing them yesterday!
 
EC1 said:
from ...say a set of ratings for the last 3 runs - how would someone compile a tissue?
Click to expand...
My process is to go through the last six runs as listed on the Timeform racecard with the relevant ratings and make quick notes about those runs - reasons for below par (or standout) runs, patterns in form related to trip, ground, field size, track configuration, fitness (particularly noting any absences) and running style. I'll try to make an assessment about whether a horse is likely to run to its best recent form given likely conditions and will score each runner roughly. I try to categorize each runner in terms of their overall chance (outstanding, strong, fair, moderate, poor) and then put them in what I think is the correct order within that framework. I'll then price them up roughly and finally fine tune to 100%.

Clearly there is a method of doing so automatically based on the ratings and values which denote the strength thereof given conditions. I'm a tad old fashioned, but reckon we're doing the same job.
 
rorydelargy said:
My process is to go through the last six runs as listed on the Timeform racecard with the relevant ratings and make quick notes about those runs - reasons for below par (or standout) runs, patterns in form related to trip, ground, field size, track configuration, fitness (particularly noting any absences) and running style. I'll try to make an assessment about whether a horse is likely to run to its best recent form given likely conditions and will score each runner roughly. I try to categorize each runner in terms of their overall chance (outstanding, strong, fair, moderate, poor) and then put them in what I think is the correct order within that framework. I'll then price them up roughly and finally fine tune to 100%.

Clearly there is a method of doing so automatically based on the ratings and values which denote the strength thereof given conditions. I'm a tad old fashioned, but reckon we're doing the same job.
Click to expand...

i like to automise everything - its how my mind works i'm afraid:)

i've had a go through the years but never really been satisfied with what i came up with re tissues. Using a bit of free thinking judgement is easier i think.

Find contenders - non contenders

allot 80% to contenders

then try and put the contenders in rank order as you suggest...then try to allot a price

One thing that always throws me - i look purposely at a race with a very short price fav..just to test on....and try and see how a numerical solution could ever give an odds on price..and always find it difficult to rate a horse as having that strong a chance.
 
rorydelargy said:
If anyone fancies having a go at pricing any of Windsor's races up tomorrow, we could have "Figure 2" for this discussion. I can't say I hugely enjoyed doing them yesterday!
Click to expand...



19:40

TRIPLE DREAM 4.9
OCEAN BLAZE 5.4
STEELCUT 7.0
OUR PICCADILLY 7.4
DESERT POPPY (IRE) 8.4
MAKE MY DREAM 8.6
SOLEMN 8.8
EQUULEUS PICTOR 9.2
RIFLESSIONE 9.9
GREEN LAGONDA (AUS) 11.8
Overound 115%


20:10

RESURGE (IRE) 2.4
SEQUILLO 3.5
EFFIGY 4.1
PENCHESCO (IRE) 5.6
GENEVA GEYSER (GER) 7.0
ABERGAVENNY 7.4


Overound 111%
 
Back
Top