Hi guys, I have been using this website for a long time and to be honest i am still a bit confused as to how the actual raw figures are formulated.

I dont expect Martin to reply to this, he is very busy im sure running an excellent web site, my hat off to you Sir.



Now....i recognise that each point equates to fifth of a length of a racehorse. so 5 points over a mile - 1 length.
Now is this speed figure for all classes or for each individual class?

For example; A horse may register a raw speed figure of 55 on a class 4 race. The next race he may produce a speed figure of 64 over a class 5.

Does this mean he has actually gone faster in the class 5 race or is it relative to that class?

Also...what is the actual time difference between classes on average? I find it hard to actually determine the significance with a drop or step up in class.



Thanks guys, i read here regularly but have never posted. Great Website Martin. Thanks all.

Hi James,
Interesting points however:-

I thought SFs were calculated to pounds. The length difference coming from how many pounds are equivalent to one length. This is dependent on race distance. I use 3 lb for one length in sprints and 2 lb for one length in races from 1m to 1m 4f in flat races.

I don't think SFs are class dependent in the calculation but based on actual race time vs standard race time for distance run at that particular course, with the actual race time being adjusted +/- with an allowance for going.

The horses SF is then adjusted for the weight carried to a standard of 10 st for Flat and 12 st for NH. (Hence don't mix the two up if a horse runs both as its NH figure will be higher by default).

If a horse did not win a race in which its SF is calculated then the SF is further adjusted based on the distance it lost by, on the lines in para 1.

These are only my own thoughts based on how I calculate my own SFs which are not so far away from PF numbers whether this is coincidence or not, I'm not sure.

Maybe someone can clarify for us? rgds Dagil

I would also like echo your comments and thank Martin for the site and all the members for their interesting and informative posts

In addition,

I have just looked at a racehorse (Ten Down).

I looked at his form, he ran 2 races at Wolverhampton. Same track, distance and going (standard ofcourse) but one was class 4 and the other class 5. He won both races and was officially around a second from the average of each particular class.

The class 4 rating was 67.2 (1m 1.98s) slow by 0.78 seconds
The class 5 rating was 65.1 (1m 2.73s) slow by 1.53 seconds

So the post compare the times to the average at the track and class NOT calculating going etc.

This would make the average times at Wolverhampton over a 5 furlong trip:
Class 4: 2m-0.78s
Class 5: 2m-4.26s

Almost a 4 second gap between classes which thinking about it im pretty sure is correct. Even though the track average is calculated to class times it is clear now that Patternform does NOT.

Therefore.... 5 points on Patternform actually equates to 1 second on Patternform over 1 mile.


That is my theory now, whether its credible is another story but it seems pretty accurate...
Its actually amazing how accurate Martins figures actually are.

Hi James,
Firstly, I thought your initial post would have created far more interest and comments than it seems to have done so far. It is a pity no-one else wishes to contribute and/or help us on the posts above.

All who use this site make use of the Speed, Multi and Ranto figures given by it but apparently do not have any opinions on what they are using or what the numbers are or mean?

I can only give my opinion on what I do and assume in my own calculations and cannot really comment on how and on what basis the PF Speed Ratings are calculated or derived.

I only assume they are calculated back into pounds or back to a Standard Base number of 100 where a horse running to Standard Time, on perfect going carrying 10st would be rated 100 as opposed to 140 lb without the standardisation, as I have no privvy information on how PF ratings are arrived at I honestly do not know.

However, your comments on their relative accuracy as far as I can see are completely true. If I rate a race myself, I may not get excactly the same number as on PF but I do get the same order for all the horses.

Hence if I am completely honest, it is far less time consuming to use the figures on PF which are avaiable at the click of a mouse, than to plod through a race on one's own methods. That is why in my opinion PF is such a great site if used correctly

With regards to the 3lb and 2lb allowances for a distance of one length. These I belive are fairly standard and were used I believe by handicappers to asses a race. It is also a long time since I read Mordin but from memory he probably used the same figures as a base.

On Standard Times, I also do not fully understand how they are arrived at. You could spend hours with a stop watch and make your own or used the Standards Times published by the various organisations. I much prefer the latter.
However when I compile my own figures, the standard time itself in reality is not of any concern but the horses' performance +/- on standards time.

Therein lies the first fault with SFs is that in the calculation you must assume that a horse which can run at standard time at one racecourse can run at the standard time at another. All racourses are not the same and neither are the horses that run on them, for sure. You can limit sthis by only using SF figures in straight sprint races but even then all courses are different in gradient.

The second fault lies in the fact that you must then assume that a horse with an SF at one distance can run with the same SF at another. (Obviously, you are not going to take a horses SF at 5f and assume it will have the same at 1m 4f but the principal is there).

The third and most problamatic is the Going Allowance used. The going allwances as published are based on the race times and distances relative to the day in question. I have no idea where PF obtains the going allowance in the calculations but to a great advantage on this site a Going Filter is available and of far more use., than knowing the possible discrepency in the going allowance used to calculate the SF.

You can find on some days a large difference in the going allowances from one given race to another. The going allowance may in my opinion not only be subject to the actual going but also the pace at which the race was run at and race tactics in play.
PF lets you check the and possibly eliminate an SF based on the current day's going against the going on which any individual SF was calculated.


I don't really understand and can't answer the question about a horse running at an SF of 68 over a mile as it needs qualifying both in terms of the weight carried and the base number of the SF calculation standard. Maybe once again someone else can help in the discission?

On Official Ratings, given by The Handicapper, these are not SF figures. A horse rated 80 in any Hcap does not necisarily carry 80 lb. This is dependent on the race in question and the top weight of the highest rated horse.
A horse having a OR of 80 would carry 5lbs less than a horse having a rating of 85 but their actual weights carried are governed by the race in question not the actual weight of the OR. Thus the ratings define the weight that must be carried by the highest rated horse in the race based on the race rules and all the other horses in the race carry a less weight in pounds by how much they are rated less than the top weighted/rated horse.

On the question of Ten Down, the SF raw ratings I believe have nothing to do with the class of the race, only the speed at which Ten Down ran on each occassion. Obviously in the Class 4 race he would have (in theory) have had to put up a better performance than he did in the class 4 race. He could have well (in thery again) have won the Class 5 race with a higher SF than the Class 4 race. Class is very elusive.

Thanks for your comments earlier, no need to take your hat off, for sure .

I'm glad to post my opinion but maybe for sure there are some faults in what I write. rgds Dagil

I would agree with the first post made by James. This is generally speaking the standard method used to calculate speed figures, although you will come across variations of lb's per length and different ways to calculate going allowances. Speed figures have never really caught on for UK racing and i think it's widely accepted that one of the main reasons has to be the variety of racetracks used in this country, coupled with past and present draw bias among other factors. There are big players who swear by using raw unadjusted figures for UK racing and this is an endless topic for discussion with no right or wrong answers. If your serious about speed figures there is plenty of reading material available from across the pond in the US. Speed handicapping has been perfected in the US and became a victim of it's own success. Writers such as William L. Quirin led the way in expanding speed handicapping incorporating pace, trip, the horse's fitness and liking for the track among other factors. For what it's worth i think pace analysis is a MUST when considering any investment and it's often ignored due to the lack of scientific prediction. If taken by the subject it's ok to think outside of the box........better still invent the next box. Thoroughbred Handicapping - State Of The Art is a good read, first published in 1984 i think, it can be difficult to get hold of but worth the effort imo. I hope you have lots of time on your hands..........

Hi Nathan,
Many thanks for setting the record straight on SF figures. I will have to give the calculations I do another name it seems. I always use weight dependent ones in which all figures including race time +/- sec are converted back to lengths and then back into pounds dependent on the race distance using 2lb or 3lb per length dependent on race distance. My figures for a horse's past race are then simply corrected by addition or subraction of the horses weight in the race being analysed.

I fully agree with your comments on their use at UK turf tracks. I really only find them accurate in Sprint races when the going is good or firmer. Your comments on pace are very interesting, as in another way I said earlier the going allowance used is not always accurate because of pace and race tactics. this is probably why SF figures work better in Sprint races, there is less room for tactical manoevering as the horses are pretty much flat out most of the time I would think. At least more so than in a 1m 4f tactical race.

On the subject of SF figures and UK track being so different. Have you any thoughts on how they ought to work on Polytrack? These in theory are the most consistent surfaces we have.

One again, many thanks and it should be my hat of to you James this time

rgds,

Dagil

In my experience speed figures are more consistent at the AW tracks.

Just like to say thanks to the people who have participated in this thread. Really informative and im sure im not the only one who has learned a significant amount about the ratings.



Does anyone know or can justify how many points on patternform equals to over:
1 mile races
1mile 4f
2m 2m+

Although Dagil, i have agreed with you almost on everything you have said. I would have thought that lbs per length would be more (rather than less) than 3 over 1 mile 4 furlong distances.

I mean longer distances (2m+), the speed ratings vary enormously actually making them quite innefective to take as face value (interesting reads on pace and their effect to racing times).

Over the jumps and longer distances, i dont even look up patternform ratings because i cant work out how many lbs per length most horses equate to. Some opinions and feedback would be really appreciated.

Hi James,
Thanks for your comments and I think the same ought to be returned to yourself for sparking off the debate. Hopefully we will all learn something from it.

On the point of weight vs distance, maybe we are not understanding each other, so I will explain in a bit more detail where I am coming from.

I assume that in Sprint Races that 3lb will slow a horse down by 1 length and in races 1m to 1m 4f that 2lb will have the same effect.

The theory goes as follows, if 3lb slows a horse down by one length at 5f 6f then it should take half that weight to slow the horse down if the distance is doubled ie 10f or 12f
The calories expened in theory in the work done by carrying 3lb over 5f are the same as carrying 1.5 lb over 10f

3lb*5f = 1.5lb *10f = 15 lb f work .... (not 15 lb/f, there is a difference).

The actual numbers I use, I believe come from studies a long time ago and were (if still not are) used in Handicapping and penalties in WFA races in British Racing. They used to be published in official tables.

It appears that probably due to race distance and pace over 10f it is better to use 2lb rather than 1.5 lb as the horses run slower per furlong and thus do not expend the same energy per furlong in the actual running of the race. (ie you walk one mile or run one mile).
1.5 lb per length can be used for distances over 1m 4f but at these distances SF figures in my opinion are not of much use as below.

SF effectively treat horses as machines, in that they can and will always run to the SF. This is obviously not the case in reality but it does give an indication of the relative ability between horses.

With regards to distance I fully agree with you. from my own experience SF calculations are more accurate in sprint races. Hence I believe Nathans comments about pace being taken into consideration. The longer the distance the more room there is for race tactics and pace variations, all of which affect the SF calculations.
Also going has a big effect and maybe the attributed going allowance for a paricular race is not true. If a race is run on soft ground at a slow pace, it is possible too much weight is placed on the going and not the pace. The false going allowance rsulting increases the SF figure.
Hence my question to Nathan about AW tracks in particular Polytrack.

In theory once again in my opinion a straight Sprint race on Polytrack should provide the best comparable SF calculations between horses. The surface is the same and there is not so much room for race tactics.
My only dilema is when I use AW calculations in Hcaps the numbers are always very close but in reality that is how they should be if the handicapper is doing his job properly and the SF figures are calculated being properly.

The lengths per second changed this year from 5 lengths per second in 5f races to 6 lengths per second. The BHA were also going to alter this with going changes. I have just picked the following up on a forum where we were involved in discussing the implications for those of us who compile our own ratings. The link is the official changes when the going alters etc

Only registered users can see links on this forum! Register or Login on forum!

For those that use lbs per length consider the following. With the old method 5 lengths = 1second and lbs per length.

5f = 3lbs per length or 3ft per pound which would make the length of a race horse in stride 9ft long.

From a Time perspective under the old sysytem, the standard horse covered 55f in one second. (5l =1sec or .20 seconds per length)

Now divide 55 by 5 (Lengths ) and your race horse in stride now = 11ft. Which is a major contradiction as your race horse is 9ft long in stride according to lbs per length and 11ft long in stride when using feet per second.

Good Luck

Monster

Excellent addition to this thread Monster, for discussion,and well done for the information.

I always used in races that a horses length was judged as 8 feet.

Thus ...when a horse finishing in a sprint race... it would be doing so at around 40 mph. (Obviously this only applies to the first few horses going past the post finishing strong ..thus also assumes they are flat out in doing so)

Using my old method you would get around 0.18 second per length based on a horse being 8 ft long and finishing at a speed of 40 mph. (this gives 5.56 lengths per second and the BHA have just increased their figures from 5 to 6).

My figures also come from way back when the official distances were done visibly on an actual length basis as judged by eye and later a camera etc not the time difference between horses as is now ... ie the distance a horse was judged to be in front of another horse by visible lengths, according to the judge or camera ... . not the seconds on the watch.

Now lengths distance in a race are not based s on the actual distance between horses in terms of true lengths (feet etc) ... but the time between horses, and this time ... is then converted back to lengths for official distances ... as per the web page address you give above. There I think lies the anomally. They have to adjust the length/sec because they cannot adjust the measured time in sec.

Before we converted the "true as seen" distance (in lengths) into seconds, .... now the absolutely true seconds are converted into "artificial lengths", not real distances as such , anymore ?.

Are horses getting longer? ... from the original 8 or 9 ft now to 11 ft ??

No not IMO but maybe they are getting faster on average as athletes are also doing these days.
This IMO really ought to be put as such ... into a horses finishing speed gradually increasing over the years not the horses getting longer, I think?

Thus the time per length run is getting shorter, not the horse's length getting longer. Hence as observed by BHA the time per length is decreasing and not it is not the horse's length that is increasing, causing the adjustment I think?

The horse is still the same length ... it just runs it faster, that's all.

Maybe now we should increase our lb per length in the calculation to accomodate it?

Or maybe in reality ..... we should all realise that horses no longer lose by real lengths anymore but by timed seconds or fractions of such. These seconds are then converted back into "BHA" official lengths as per the tables on the web page kindly posted above ..

Our horses haven't grown from around 8 ft or 9 ft to 11 ft or more in length overnight ...... the BHA has changed the baseline on which we all worked on by changing their definition of a length in seconds run not the size or length of the horses involved. Now a horse that was 8 or 9 ft is 11 ft because the BHA have changed the clock.

It is still a horse and a has a horses length, unless it is racing, then its length is measured in seconds, not feet. ..... (and forgive the joke but my length is still measured in inches ) Thank God it's not seconds there !! I would have complaints!!

We'll just have to get used to it, that's all

Once again many thanks Monster for a very interseting and informative post.

Hi Dagil

You could measure in MPH

Ie using RP Standard Times & Lush Lashes race as an example

RP Standard Time 12f Newmarket 147.5 Seconds or 36.61MPH
Lush Lashes 37.22 MPH

In this example,take off the RP Going allowance.

Lush Lashes Average speed 36.43MPH

2nd horse was one & quarter lengths away Average speed 36.38mph

Compare both against the figure for standard 36.61mph etc for your speed rating.

Good Luck

Monster

Hi Monster,
Thanks, using MPH maybe a solution to look into.

I think in a nutshell the point I was making above is that a few years ago Speed Calculations were based on Actual Race Times against Standard Race Times, with the going allawance taken into consideration. This can still be used for the winner.

However for placed horses the times were calculated from the horses losing distances. ie in my case I used to use 0.18 sec/l or the reciprocal of that being 5.56 l/sec

The length was actual distance as best measured by eye or camera scale. (I personally used to ignore a sht hd, and then take a hd as 1/8L, a nck as 1/4 L the rest as quoted in the race result).

Now the length is not an actual physical distance between horses as such but one that is calculated from the time distance between horses..

This is further complicated as per the PDF you kindly directed us to that this calculated 'length' distance is also dependent on the going not only time. So the BHA are building in a different and variable (as opposed to truel) length distance dependent on the going based on time, ... obviously to try to be a little more accurate in the "standard length" quoted in official race result distances.

However, this being the case the BHA are also building in an "approximate going allowance" in themselves in the distances quoted, I believe?

The only way I can see to get back to the original method of Speed Calculations, is to take the quoted official length differences by the BHA dependent on the prevalent official going, .... as the times per length will differ depending on going and in AW the racecourse also..... then use these times for the placed horses (rather than quoted distance), then recalculate the race from these using Standard Race Times and Going Allowances as in RP etc.

Now instead of what used to be a fairly straightforward, if somewhat laborious calculation, we have the compounded problem that the BHA official length quoted is dependent on both time and going ,, and on AW surfaces also the racecourse involved becuase it is no longer true distance but one recalculated from time as per the PDF file.

ie a length at Sthl, does not mean the same physical distance as a length at Wolves or Lingfield, and neither of these is the same as a length at Great Leighs, even if all courses are running at STD and .... neither of these are the same on turf depending on the prevalent going??

Hence your horses growing fro 8 or 9 ft to 11 ft etc, ... it is not the horses that are varying but the BHA's definition of a length in the race results depending under what course and conditions the horse runs.

So instead of using one calculation which covered all going dependent on the giong allowance to calculate lengths into time ...., the "distances" are now time, going and racecourse dependent before they can be calculated back to actual times for calculation purposes.

The MPH can easily be used for the winner as can the old method to whatever standard one personally uses, however the speed of the second horse has now to be calculated from the BHA distance back into time dependent on the course on which the race was run and the BHa definition of the official going (as it varies with these now),... not the measured length distance with the going allowance as before Otherwise the going allowance is used twice over, once by the BHA in their assumed not measured distance of the placed horses based on the PDF ... and then once again by ourselves in the Speed Calcs.

The going allowance now, ..(I think?) .. need only be used once the whole race has been brought back to actual time, based on the official race time of the winner and the times of the placed horses back calculated from the distances as per the PDF file. Then this can be compared against Standard Time with the RP or whatever going allowance used etc.

I wish you the same Good Luck and once again thanks for your input, much appreciated by me for sure. Dagil

Maybe, one day, rather than the above, in which the BHA calculates artificial length distances from race times..... They will just kindly provide us with the actual times, for all the horses.... rather than their own calculations of distances from such times which they already must have?

Thus we would then not have to read a race based on their assumptions into how many seconds a length is dependent on a course and it's going but we would have the actual time in seconds (and parts of) for all the horses involved.

They already do this with athletes and F1 GP drivers etc I'm sure it could be provided for horses... if the BHA wishes to do so ... that is ?? The information must already be avaialable.

They... The BHA... for sure should already have this information for every race run, otherwise they could not calculate their artificial length distances as per the PDF in the first place

Maybe the "Old Boy's Club" doesn't want this information to be published as such ?? For sure the reason cannot be a lack of technology, it already exists as we well know, from other sports .... so why the Red Herring, converting the seconds into artificial lengths.

Why not just quote the true race time for all the competitors in seconds?.... as in atletics or F1 races etc. ..... They must have them but they will not publish them in British Horse Racing for some obcure reason.

I don't know.... any ideas on this? Anyone ??