- This topic has 48 replies, 14 voices, and was last updated 16 years, 11 months ago by
.
-
Posts
-
been reading anew book CLASS FIGURES but it deals with raceform speed figs i prefer racing posts what would be considerd a good speed figure as compaired to its racing post handicap figure in each race 10pts less than its race postform mark or what trying to get a comparison i know one has weight inclueded and one does.nt
class tells,
Briefly, Racing Post Ratings(RPR) and Topspeed are on the same scale and both are calculated and adjusted for weight carried.
Topspeed generally lags behind RPR because not all races are run at a true pace, so that while every race is assessed for RPR, only a fraction of races produce worthwhile speed ratings.
The Topspeed rating never exceeds the RPR because it is a vital peice of evidence in deciding what the RPR for any horse’s performance should be. If a horse achieves a Topspeed rating of 100, the RPR must be at least that figure.
A Topspeed figure within 10lbs of the RPR for any horse’s performance tends to frank the form. If you are considering ability based on the two ratings, you have to be aware of the fact that the Racing Post gives the rating for a horse’s best performance, which might have been achieved over a different course, or distance and possibly on different going than today’s conditions.
That certainly used to be the case but the RP time figures have been altered so that for the flat it is based on a standard of 100, carrying 9 stone, and for the jumps 135, carrying 12 stone.
Hi robert99
do you know how they calculate their topspeed figure?<br>I am sure it’s no big secret, or is it?
byefrom<br>carlisle<br>
No secret at all. The Racing Post and the Weekender have both carried articles detailing exactly how the Topspeed figures are compiled.
The RPR and the Topspeed rating are complementary and measure the same thing i.e. a horse’s ability relative to others.
The Topspeed rating follows directly from the race times and calculated going allowance using Racing Post standard times. These standard times are based on RPRs(standard= best achievable,RPR = 100, 9st, good ground).
If the Topspeed figure comes out at 120, it is certain that the RPR cannot be less than this figure because the basis of the RP handicapping method (and Timeform) is that a horse rated below this figure(120) cannot achieve a higher speed rating. In short, if a horse can run that fast, it must be worth the rating.
There are a whole number of horses running today with a Topspeed higher than their RPR.
In addition the idea that RPR & Topspeed are complementary may be true in terms of how they are expressed. However, as far as I understand it, Topspeed is a measure of what a horse has actually achieved with referrence to a standard time and the RPR is a measure of what a horse ought to achieve with referrence to it’s current ability.
I would be very careful in mixing the two, there’s no logic in it and it would seem quite a lot of backfitting. For an independent speed rating try Pattern Form.
<br>
(Edited by Pompete at 10:05 am on May 18, 2007)
Pompete,
I think we may be at cross purposes, which may be my fault for trying to answer a complex matter in brief.
What you see in the Racing Post as part of the daily racecard is a horses BEST topspeed figure from the last 14 months or so, adjusted for the weight carried today.<br>The RPR is what the handicapper believes the horse is capable of running to today, adjusted for weight carried.
When I’m refering to Topspeed ratings and RPRs being complementary, I am refering to handicapping(or rating) a race ON A PARTICULAR DAY. If you are working out ratings today for York, say, you would give each horse a Topspeed rating and A RPR for the race it ran in. In these York races, the Topspeed would never be greater than the RPR. Well, not today, but if the form of the race doesn’t work out, the handicapper might go back and reduce the RPR, which could take it below the (fixed) Topspeed figure. If the form works out very well, The RPR might be raised retrospectively,but the Topspeed figure is fixed.
The reason you see Topspeed ratings greater than RPRs is because the BEST Topspeed figure is fixed whereas the RPR is variable. If the RPR falls as a horse loses form and is downgraded, the Topspeed figure will then be above it.
I don’t think this happens too often – I could only find one example at York today – All You Need in the 3.50, where the form from the race at York on 29th July,2006 was presumably downgraded in retrospect.
If you look at the results on the RP website after about 4 days, you will see the RPR and Topspeed side by side. I can assure you that if the race has been run at a fair pace, the two will not be far apart and the RPR will almost certainly be above Topspeed for each race. You might find the odd exception, but it would be very unusual. I would be surprised if you didn’t see the logic in how these ratings are compiled once you understand how it’s done.
carlisle,
I’ve rooted out my dog-eared cuttings from the Racing Post.<br>There are three sheets:<br>1. How Topspeed ratings are calculated.<br>2. The scale for converting weight into time for given distances.<br>3. The weight for age(WFA) Scale – not that important now because the ratings are adjusted for WFA on the racecards and on the website, but not in the results.
If you pm me with your email address, I will scan them and send them to you as an attachment.
Hi Artemis
thanks, here’s my email. I calculate my own version of the Solidus ratings, so I do understand how speed ratings work. I would like to get inside Topseed, more out of curiosity than anything else.
Having said that a good idea might occur to me in the process. Plus it will stop bugging me. I still have my old booklet explaining all about StopWatch ratings, ah those where the days……. Sheight!
cheers from<br>carlisle
(Edited by carlisle at 5:58 pm on May 25, 2007)
Before attempting to reply in detail to this posting, please can I say that my comments here refer to the calculation of speed ratings for races that have already been run and not the ratings that appear alongside any given card in the Racing Post on a daily basis.
I compile my own speed ratings, using the same 0-140 scale as Topspeed (&others), which allows direct comparison to be made with a set of conventional form ratings. In assessing the results, there is no reason why, for a given performance, a winning horse cannot post a speed rating that is higher than the equivalent form rating. Such occurences do, however, call for close examination to ensure that a horse is not being overrated from a time perspective. The calculation of ‘raw’ speed figures, before any adjustments have been made for, say, the going is an objective exercise. However, calculation of the final adjusted figures does require an element of subjectivity and it is possible to assume that the compiler of the Topspeed ratings is filtering his results in some way that limits the speed / form rating differential – perhaps artificially so. If you look at the Topspeed figures for a given meeting that has occurred in the past, you will also note that on the majority of occassions there is at least one horse on the card that achieves a speed rating within 1 to 3lbs of its form rating for the same race, which suggests some deficiency in the methodology being used by Topspeed. There are also numerous occasions when the going allowances published by Topspeed for a given (past) meeting are identical for races taking places over straight and round tracks, when it is clear there are significant differences between the two due to ground conditions and / or wind speed / direction. Bearing, in particular, this latter point in mind it is clear that the Topspeed figures are unreliable and should be treated with substantial caution. The ‘standard times’ used by Topspeed also seem untrustworthy and can underestimate performances – for example a rating of 107 for Manduro in April’s Earl of Sefton at Newmarket. By comparison, my rating was 119 and the comment I made at the time was, ‘A smart performance by the clear-cut winner, who has Group 1 form and should continue to hold his own at the highest level’. Similarly, Mahler in the Queen’s Vase at Ascot, Topspeed 67, my rating 107 and comments at the time, ‘Fair staying performance from the winner; could go on to make his mark in this division’. The assessment of race times is fascinating and, in this country, difficult to deal with; it is a pity that so much rubbish is written on the subject.Hi scallywag,
Welcome to the Forum.
I’m a great fan of the RPR ratings/Topspeed combination – as you may gather from my posts on the forum. However, I recognise that form study is a very inexact ‘science’ and your points about Topspeed’s methods are fair.
I have corresponded with Dave Edwards(Topspeed) from time to time and I think he is also aware of the limitations of awarding speed ratings to thousands of races a year.
The RP seem to have a decent theoretical model of the relationships between time, weight and distance, but it is only based on averages(weight and distances) and best performances under given conditions(time). The margins of error are so great that there are bound to be differences of opinion about individual performances, so I’m not at all surprised by other speed figure compilers disagreeing with Topspeed. The two examples you quote look outside of the range of disagreement I would expect and are probably worth looking at to see why such a discrepancy exsists.
It may be that your own model has a different set of standard times to the RP and therefore your assessment of the going was different, or there could be other variances. It would be interesting if any Timeform subscribers could supply their figures for the races in dispute.
The VDW fans are still discussing races from 1978(see VDW threads), so we might not get to the bottom of this for another 25 years.
I looked at topspeed sometime ago and I don’t think his going correction is very good and I disagree with the way he calculates his standard times.
dave,
A bit of a woolly, (hurried?) reply, not your usual reasoned statement.
Anything in particular about the going correction and standard time calculations that you disagree with? They are very much connected.
I know Artemis, I was busy laughing at your ‘we’ll get around to discussing this in 25 years’ statement ..
These are just my opinions obviously. I wanted to have a look at the topspeed figures primarily to find out why they are so bad. They seem to be the worst speed figures going, when a publication like the RP has the resources to produce practically anything they want to. My inquisitive mind wants to know why.
First of all, the standard times have a standard for 2YO’s and a standard for non-2YO’s .. this implies that the actual calculation doesn’t include a factor for horse age/sex, weight/length, not to be confused with age for weight. Omitting this would make certain values almost meaningless, as we know males mammals are bigger than female ones. In the case of horses males are heavier and longer.
The going correction is also a problem for me. I prefer the method where you use the 3 shortest distance races to calculate the correction against a standard time for class (class pars). That way you aren’t (or you are less likely) including false bits of ground and logically speaking most of the races will finish over the same ground as the shortest 3 distances anyway, if that makes any sense.
Dave,
There is only one standard time for each distance regardless of the age of horses. The course record times are shown seperately in the RP for 2yos and older horses.
The going correction is often different for races run on the straight part of the course from those run on the round part of the course, for obvious reasons. This is a very difficult area and can give rise to significant discrepancies among ratings compilers. Experience of rating races over the same courses over many years is invaluable.
As you say, it’s all about opinions, but form and speed ratings are my special interest, and I’m confident that the RP does a very good job given the limitations of this type of analysis. There are several organisations and individuals who do something similar, yet I don’t think any of them are any better than the RP. From the cost angle, the RP is at present only the price of an internet connection, or a fraction of this if you surf a lot.
It doesn’t explain why they are so bad though, if you do the tipster analysis on the site and check topspeed it makes pretty woeful reading and that’s him not tipping a horse in every race.
You would expect a pin sticker to get a 20% SR if he stuck to the top 3 in the betting with no analysis at all.
- You must be logged in to reply to this topic.