Uk Racing Handicap ratings... Any Experts ?

You are listing both Flat and Nationl Hunt Races, there is a different handicapping system for each

eg highest rated Flat horses are rated 135+ 

135 is only an average horse in Jumps Racing, where the top ones are rated 190+

Tobermory, that explains the huge drop in ratings but why the weight increase on 20th July ?

so Casual Garcia is rated 60 odd on the flat and 100 ish over jumps , indicating he is a moderate performer in both codes

I never look at such ratings....would rather rate horses myself...no point understanding the complicated workings of others minds IMO re such ratings.

Common sense is the main factor ....

Trainers need to know how system works though ........but not punters IMO....

The weight increase i guess could be due to running in a slightly lower class of race on July 20

If the July 15 race was for horses rated 0-70, and the highest rated had a mark of 70, and carrying 9-9, then Casual Garcia would carry 9-2

If the next race was for 0-65 rated horses, then he would have been near top weight himself.

When it was 2nd it was a NH race....the next 3 races were TURF......so tobermory has already given reason for drop.

How accurate can such ratings be...........the going in the horses last 8 races has had 7 different goings....... ANSWER - they are not IMO.........

yes, but Escapee is also querying why the horse carries different weights for 2 flat races he is on the same rating for .

Another reason this might happen would be weight for age .If a 3YO horse is top rated in a 3YO handicap, he will carry top weight.

But if he then runs against older horses he may still be top rated, but will carry less weight than one or more of the 4YOs.

A 3YO rated 65 might carry 7lb or so less than a 4YO rated 64, . The amount varies according to time of year.

Weight is certainly 1 of at least 100 FACTORS to rate a horse for some races.......and the GOING is probably the key factor to how a horse will fare.

I just believe the OR etc ratings etc just are not accurately produced to be of much use in weighing up a H/C race as to which horse has strong chance to win etc.....

There are just too many variables in assessing / rating a horse for a specific race and I suspect that the ORs just ignore many KEY FACTORS....therefore are not much use really to a punter.

They will be to a trainer placing a horse in races against known ORs of other horses, especially when the trainer knows just how much his horse will "imprve" when it gets its' ideal trip in the right type race and on the right going at the right time of year etc etc etc

I really should keep off this thread....as the originator wants "experts" on such ratings ......and just ignore them as irrelevant!

Sorry Escapee.....I will vacate the thread .....as T knows far more than most on here re how such ratings are stuck together, so you are in capable hands!!!!

Gl with trying to fathom them out!!!!

Looking up the races on the RP i see that actually the July 15 one was a 0-65 and the July 20 a 0-70.

The top weights in the 0-65 were rated 65 , 2 lb more than CG , and both carried 9-7, while he carried 9-2. This is because he was ridden by a jockey who was a 3 lb claimer .

On July 20, in the 0-70 , the top weight is rated 68 , and carries 10-0, CG - again rated 63 - carries 9-9, which again makes perfect sense.

I had a number of winners today and a close call on nap in comp. coming 2nd after giving rest a 20 lengths start.....beaten 1 length and finishing faster than the winner ...AGHHHHHH.

Horses I selected today that won at decent odds and their OR position in race.......


1st       Sibenek (IRE)      15/2   was rated 6th top rated/8 re OR

1st       Salpierre (IRE)      7/2fav was rated 6th equal top rated/10 re OR

1st       Conry (IRE)      11/4fav was rated 2nd top rated/11 re OR

1st       Peter's Gift (IRE)      20/1  was rated 7th top rated/10 re OR


Now ....how can OR ratings IMPROVE my guessing?

I just believe the OR etc ratings etc just are not accurately produced to be of much use in weighing up a H/C race as to which horse has strong chance to win etc.....

There are just too many variables in assessing / rating a horse for a specific race and I suspect that the ORs just ignore many KEY FACTORS....therefore are not much use really to a punter.

DFC, Many of your observations are correct there certainly. The Official Ratings are quite inflexible as they tend to take winning margins rather literally, and they do not account for factors such as the going, horses running over different trips, course form, differnt jockeys, and they also assume every horse always runs on its merits, which is rather dubious. The proof that the ratings are not the best guide is in the market: If it was perfectly accurate then every runner would be co-favourite in every handicap.

However you can use the rating system to bet. Using the same scale (on the Flat), 135+ for  Sea The Stars, 110 for a Group 3 horse, 60-70 for an average horse like  Casual Garcia, down to 25-40 for the worst horses in training , you can make up your own ratings, and include all the factors you mention that the Official Ratings discount , and if you are good at it , you will be able to price up races accurately and identify value bets .This is how Compound Magig and Top2rated operate on here i believe.

Thankyou tobermory, you've really helped shed some more light on the subject.

Do you know anything about 'speed' rating ? ( of which there seems to be a few flavours )
Any stuff you could write about, assuming I know nothing ?

DFC, I see from your threads that you probably use some sort of subjective ratings to back and lay horses in the same race, any chance of a vague outline ?( without giving away your secrets ofcourse ) would be great.


Basically, I'm trying to get a handle on the various methods of rating horses, how the market is formed/priced and how to start
deciding if the price good or bad or about right.
All I've done in the past is guess.



All view points welcome.

If anyone has an opinion, please feel free to write about it here.


Thanks to all in advance.

Most of horses above shortened in odds AFTER I posted my guesses on here...

The odds I got were :-



1st       Sibenek (IRE)      15/2   ...BF SP 9.00............My odds BF.. 22s

1st       Salpierre (IRE)      7/2fav ...BF SP 5.30............My odds BF.. 6.20

1st       Conry (IRE)      11/4fav        ...BF SP 4.10............My odds BF.. 3.55

1st       Peter's Gift (IRE)      20/1   ...BF SP 22.01.............My odds BF..28s



Of course it was after PETER'S GIFT won.....that BF SP shortened odds on SIBENEK and SALPIERRE ( CONRY was a short odds horse that was fav when I selected it ....but of course I don't select horses just because they are favourites .)......


Can ORs be useful at all to get such "value in odds".......................I suspect NOT!

Escapee, i know nothing about Speed Ratings tbh . I've never really looked into them as IMO race times are determined as much by the tactics of the jockeys on the day as by the abilty of the horse . Quite a few people on here though, who sound like they know what they are doing , do use them, and may enlighten you.

In general, horses are rated by distances, which are then converted to weight,typically 2 lb per length (though there is a complex scale based on the distance of the race)

So Horse A beats Horse B by 5 lengths , he will be rated 10 lb superior.

On the Flat, (i don't follow jumps really) the ratings indicate the following:

137+ One of the best horses of all time
130-136 A 'Superstar'
125-129 A very good Group 1 Horse
120-124 Good enough to place at Group 1, maybe win a below average one.
113-119 Group 2
107-113 Group 3
100-106 Good enough to compete in top handicaps (The Ebor etc) Listed Class, just short of Group class

About 50-100 is a handicapper, and can run in any race it is not too highly rated for ( eg 0-100 if rated 100 , 0-65 for Casual Garcia)
Below 50 i think is for horses not good enough for handicaps , so they run in Sellers.


To convert the ratings to prices you need to set a % chance on horses of different ratings beating each other.

Obviously 2 horses rated the same would be both be 2.0 in a 2 horse race.

I made this table a while ago to convert the difference in ratings to prices

.	Ratings Difference		Horse A:Odds To 1	Horse B:Odds To 1
.	0		1.00	1.000
.	0.5		0.92	1.087
.	1		0.84	1.190
.	1.5		0.80	1.250
.	2		0.74	1.351
.	2.5		0.66	1.500
.	3		0.60	1.667
.	3.5		0.54	1.850
.	4		0.50	2.000
.	4.5		0.46	2.173
.	5		0.42	2.380
.	5.5		0.38	2.631
.	6		0.34	2.941
.	6.5		0.31	3.226
.	7		0.29	3.448
.	7.5		0.26	3.846
.	8		0.23	4.348
.	8.5		0.21	4.760
.	9		0.19	5.263
.	9.5		0.17	5.882

As an example this is how the table above would be used to price up this hypothetical Mile Race.

eg Canford Cliffs v Goldikova is 1/1 , vs Paco is 0.42/1 , vs RVW is 0.38/1 making a total of 1.76/1 . I add the '1' in the Canford v Canford bit to produce the decimal price of  2.76

Horse	Rating		Canford Cliffs	Goldikova		Paco Boy	Rip Van Winkle		Decimal Odds	% Chance
Canford Cliffs	132		1	1		0.42	0.34		2.76	36.23
Goldikova	132		1	1		0.42	0.34		2.76	36.23
Paco Boy	127		2.38	2.38		1	0.84		6.6	15.15
Rip Van Winkle	126		2.941	2.941		1.19	1		8.072	12.39

That Race would be level weights of course. If you are doing this for a handicap you would have to adjust for the weights carried; so if you rated one horse at 70 and another at 63 but the latter was carrying 7 lb less , you'd need to rate that one 70 also, so long as you agreed that the weight would make the difference.

My ratings/prices converter table may be flawed possibly, and the ratings themselves are only as good as the individuals judgement.

Thats some good stuff tobermory.


I thought I'd create a table of the Exeter 13:50 with the current prices and weights adjusted for penalty and allowances to see if anything pattern becomes apparent.

Horse	wt net of Pen/Allow	OR	Current BF Price	win % ??
Scotsbrook Cloud	10-11	100	3.3	30.3%
Five Star Wilsham (IRE)	11-7	112	4.6	21.7%
Victors Serenade (IRE) (ex7)	11-11	104	4.7	21.2%
Saintly Lady (IRE)	10-9	98	9.2	10.9%
Kathleens Pride	11-2	102	9.4	10.6%
Captain Becket	10-13	99	12	8.3%
Casual Garcia	11-0	103	30	3.3%

Escapee, traditionally you have either to look for horses that are 'ahead of the hcapper' or that 'have dropped to a winning mark'.

The second is harder b/c of the trainer chicanery involved.

As regards the first, the hcapper cannot not up a horse by more than an amount that reflects its margin / ease of victory and what it beat on the day. It's poss. to identify anomalies by comparing ORs and RPRs / speed figs. for progressive horses, but of course these tend to be fully reflected in the betting.

*the handicaper cannot put up a horse by more...

Now this is an interesting thread!

An oldie like me can still learn something new about handicapping.

Been reading up on the OR and found out they update it every week
Going to do a lot more delving.

This URL to download all OR ratings to excel

http://www.britishhorseracing.com/inside_horseracing/media/ratings/

Great! Now to start manipulating, filtering, check results etc....

Using the latest OR will rate and test results to see if I come up with the same new OR, if not why it differs.

Have to get my head around Lbs, not that easy when I have been doing Kgs for so many years.

This from tobermory post

"In general, horses are rated by distances, which are then converted to weight,typically 2 lb per length (though there is a complex scale based on the distance of the race)

So Horse A beats Horse B by 5 lengths , he will be rated 10 lb superior."

For me 5 lengths equates to 7.5KGS which is about 16Lbs (Flat Racing) quite a difference! but has served me well.

As I stated in my Tosh thread a lot of different interpretations on how much each length is worth in terms of weight per distance and per class.

I wouldn't argue with your evaluations CM.

Timeform is like this apparently.....

Timeform rating represents "the merit of the horse expressed in pounds and is arrived at by careful examination of its running against other horses using a scale of weight for distance beaten which ranges from around 3 lb a length at five furlongs and 2 lb a length at a mile and a quarter to 1 lb a length at two miles"

My ratings are not based precisely on margins between horses, and i don't bet on handicaps so don't concern myself with how much difference a change in the weights would make. Sea The Stars beat Youmzain 2 lengths in Paris, but i rated  STS 136 and Youmzain about 124 as the race was over long before Youmzain was sighted.

Compound Magic

Distance	Lbs per length	Max weight advantage (same mark)	Weight advantage for each 1lb rise in class (same mark)	Weight / Class adjustment factor all other cases
1000	3.3	-14	-0.4	0.6
1200	3.1	-13	-0.4	0.6
1400	2.9	-12	-0.3	0.7
1600	2.7	-11	-0.3	0.7
1800	2.4	-10	-0.3	0.7
2000	2.2	-9	-0.3	0.7
2200	2	-8	-0.2	0.8
2400	1.8	-7	-0.2	0.8
2600	1.6	-6	-0.2	0.8
2800	1.4	-4	-0.1	0.9
3000	1.2	-2	-0.1	0.9
3200	0.9	0	0	1
3400	0.7	3	0.1	1.1
3600	0.5	6	0.2	1.2
3800	0.3	9	0.3	1.3
4000	0.1	13	0.4	1.4

Thanks for that info guys

As the difference from mine is the most at 1600 and over, will try some tandem ratings mine against those figures and see if what I use at the moment can be improved upon by using this info. All for improvements. After testing over a reasonable set of events will get back with my conclusions.

cheers

I do miss this sort of discussion