Rating the cards - OHR

**fragle123** · **05-02-10**, **12:30 AM**

I wonder if one of you guys could help me by explaining how you actually go about the process rating a days racing as an example tomorrows card for Lingfield step by step as i am still trying to figure it all out, i have my set of standard times but dont know where i go from there to actually rating a race or if somebody could point me to an online document that i could dl that explains it
thanks.

andthereoff · **05-02-10**, **10:02 PM**

I find this hard too! mate,

I have the data alright but when attacking a card there is so many angles to take it confuses me, some days I might stick to good times on a certian ground or good times over a certian trip, even just times at a certain track - there is rarely a horse who can back up a good speed rating because there isnt the opportunities or they are simply not good enough too;

If a horse has won over 8f at Ascot in a serious time carrying 8-4, their is only 12 or so meetings at Ascot a year, with his rating shooting through the roof thats the class of race angle out the window, he's entered at Haydock next and 5 runs after that is at Salisbury, Epsom, Sandown, Pontefract and Hamilton - where am I meant to go from here,

This leaves me with only 1 solution, the ground!?, it could be 2 years untill a horse gets them same Ascot conditions, is he the same horse? NO 9 out of 10 times he's not, is he still in training, NO 9 times out of 10 he's not so whats the use of the data I have?

Spook insights, cheers fella

**Spook** · **05-02-10**, **10:51 PM**

I'm not completely sure I understand what's being asked in honesty?

**fragle123** · **05-02-10**, **10:51 PM**

what i am thinking of doing is just sticking to the AW tracks in the uk, i have read the tut thread but i cant see the wood for the trees as it doesnt make any sense to me at all as to where you just start to rate a days card in the first place im just so confused as to what i am suppose to do ?.

**Spook** · **06-02-10**, **12:06 AM**

You can't rate a card unless you hold ratings in the first place. In essence you need to build up a library of performances and then use this to assess the various horses entered. This will take about 3-6 weeks to compile, although the A/W circuit is much easier and you can probably begin after about 3 weeks. You could back-calculate them however if you want to hit the ground running. You'd probably be as well to give some thought though regarding how you intend to access information by way of a data retrieval system

**fragle123** · **06-02-10**, **10:38 AM**

Hi spook
ok so what i have to do is begin rating the past results that i have collected upto now using the winning times and class pars ect to be able to rate horses in future races.
As for retrieving data i got some code off Konex to download all the past results from the sporting life archives and put them into excel, and i have been playing about with a spreadsheet to do the calculations for me the past few weeks with some success so then once i have given speed ratings to these horses i then use them to assess the races that they are going to run in, so i presume that i should then be looking at how they achieved those ratings in eg Class,Distance,Course and then make a judgment as to how well they will perform from those is that right.
At the moment i have Mordins Standard Times to use but would i be better off using the RP's or creating my own to use or would it not matter which, its just that i have read in a few threads here that Mordins may not be the best ones to use.
Sorry if the questions seems obvious or daft but not having done this before i just want to be sure i am doing it the correct way in the first place.
thanks

**Spook** · **06-02-10**, **11:59 AM**

I wouldn't touch Mordins with a barge pole. He uses some hypothetical 'wonder horse' and one that presumebly has a JATO unit strapped on its back. Mordins standards are underwritten by possibility and not necessarily the result of what has happened. Personally I'd prefer to base a rating on demonstrable evidence and not some contrarians opinion which one almost suspects has been formulated as much with keeping his name in the spotlight. If you're going to aim at a/w tracks only then I suppose you coudl calcualte your own and if you get them right, you'll gain an edge. If you get them wrong, you'll be hampered. A/W tracks are quite predicatble and subject to much less fluctuation though and for the most part I wouldn't have thought you'd go far wrong simply using the RP's. The only things that disrupts A/W times are strong wind, and particularly high humidity, which can occur on any track. The rails aren't subjected to the same level of movement so the only unique thing concerns the harrowing of the track which you'll learn to detect. What normally happens is that it starts to ride slightly quicker before suddenly going slow again. If you used your own standards there would be a chance you'd simply hit this cycle at some point and draw incorrect conclusions. The RP's standards have been set over a much longer time frame and can absorb them thus as they'll just show up under the normal laws of variance. I'd save yourself the time and use the RP's. The only places I'd give serious sconsideration re-calcualting course standards at the moment would be Ascot, York, Haydock and Doncaster all of whom have had singificant drainage put in during recent times. I also think there's a strong case for re-calibrating French standards as they frequently throw up time based ground descriptions that would indicate that they haven't been updated for years

**fragle123** · **06-02-10**, **01:50 PM**

Hi spook
thanks for that info so for now i will use the RP's standards until i get a bit better at understanding what i am looking at/for, so at least now i can stop banging my head on the wall wondering how the horses get rated for future races as your just using it's previous ones for the new race it's running in ie a horses previous ratings are used for it running today and i take it my thinking that class distance course they were earned atare what i should be considering when making my judgement about its performance if it was running today.
thanks for your advice and help.

**The Duke** · **08-02-10**, **10:13 PM**

Just a quick note, the RP Standard Times on the Racing Post website are completely ****ed up to what they should be.

Warwick is a particular example of this, you'd be better off doing your own.

**fragle123** · **08-02-10**, **11:18 PM**

Hi Duke
Ok thanks i will look into it and see how i get on

**Natagora** · **15-02-10**, **10:40 PM**

I'll post my Southwell Standard Times as a guideline, these are what a Beyer rated horse of 108 should be hitting:

5f - 57.70
6f - 74.30
7f - 87.30
1m - 100.40
1m3f - 143
1m4f - 155.40
1m6f - 182.30
2m - 217

The 2 Mile one would be the dodgiest as there are few races over that distance at Southwell. Be intrigued to see how others relate to these times.

To make your own, I'd grab a copy of Beyer on Speed and Mordin on Time asap. I used a hybrid of their methods to develop my times and ratings.

**fragle123** · **25-02-10**, **12:43 AM**

Hi Natagora
If you dont mind me asking how are beyers speed ratings worked out, do you think they are better than Mordins way of doing them.

**Retroneous** · **13-04-10**, **06:55 PM**

Originally Posted by fragle123

Hi Natagora
If you dont mind me asking how are beyers speed ratings worked out, do you think they are better than Mordins way of doing them.

That's absolutely a matter of preference. I like the way Beyer does things, but you really should read one of his books to get a feel for it. Some will say that his ratings suit American racing far more than UK racing and they're right - to an extent - but its better than Mordin's "Look, I've just picked this rating from the moon because the horse looked like it was going quite quick at the end" style of doing things. You need to find what suits you.

One tip is that when rating races, remember that "impossible" is a word that shouldn't exist to you. If you rate a race and give a horse a rating of 92, when it previously has only run 78, 67 and 80, then that 92 isn't "impossible" at all, although many will tell you that it is, and that the horse should be rated between 67 and 80, with the field also downgraded to that level. If you check the maths, check them again and still get 92, then you may well have succeeded in your task of finding a horse - or several horses, if the race is a close one - to follow on its next run. Rule nothing out - but make sure your maths are correct!

**Natagora** · **14-04-10**, **10:17 AM**

Beyer compiles charts for each track with times and ratings alongside one another per tenth of a second, so for example....

99.8 109
99.7 108
99.6 107

Etc etc.

Obviously the gaps in ratings get bigger per tenth of a second as the distance of race decreases and vice versa. I find it an excellent system, especially if using your own standards. I then apply Mordin's class pars to the times of the races in order to discover the going allowance.

**fragle123** · **14-04-10**, **11:10 AM**

Hi Ret
Yeh the maths? thats the problem well for me anyway as i keep doubting that im doing it the right way as it never loooks right to me but i am still trying to learn this and want it to be right otherwise its a waist of time as the ratings will be wrong and it just seems that Mordins way dont feel right to me?.
I have found a formula and code that had applied to be patented back in the 80s for a software program by the looks of it and when i use this i have noticed that it gives ratings that are fairly close to the RP's, when i say close i mean within 10 - 15pts either way and as i understand it the RP ratings are adjusted for weight.
One other thing i have noticed this week when looking over the tutorial again at the beginning when going over the donny figures for calculating the track variance they have changed between then and now, so they cant be trusted. But i think these figures maybe better for me to use and i may stick with them after i have done some testing with them.

**Retroneous** · **14-04-10**, **11:23 PM**

Your best bet, fragle123, is to post your method of rating the horses. Post it step-by-step with numbers and the workings out for maybe the first three in a particular race. That way, we can all take a look, and chip in if we see anything massively wrong.

Posting the code that you've found might be interesting too - always good to see new methods and share and share alike and whatnot

**fragle123** · **15-04-10**, **10:06 AM**

Hi Ret
Its all in 2 pdf docs here is a link for anybody to get them from Free File Hosting Made Simple - MediaFire
I dont know anything about programming but it looks a bit like VBA to me probably wrong but you will understand it better than me. There is still some of it i am trying to figure out but the basic formula for a raw rating is distance*60/time+0.077*dist -4.92/0.008 this is used upto 9 furlongs and their is another one for more than 9 as you will see in the docs. it would be interesting to see what you all think.
Also when using weight to adjust your ratings as i understand it if you use say 10st as a base and a horse is 9st then you deduct 14 from the rating and above the base you add it on, do i have that the right way round? i did see it mentioned here somewhere but not been able to find it lol.
But any advice would be helpful as i just seem to be going round in circles with this, keep reading the tut over and over but what i am intending to do is just concentrate on the 4 AW courses here as the turf would be to much for me to keep upto date with because of my job i would just be playing catch up all the time.

**fragle123** · **15-04-10**, **10:22 AM**

Last night i just quickly rated the horses in the last race at kempton that had run on the aw before and this is what i got
Thrust Control 67 6th
Hemera 81 7th
Lisahane Bog 66 2nd
Sheer Force 58 4th
Soul Heaven 56 3rd
Ana Moutahabi 52 8th
These were just ratings without any adjustments or other things added to that formula in the above post, so do these ratings agree with what anyone else has for these horses.

**Retroneous** · **15-04-10**, **06:19 PM**

With regards to weight, you have the basics down correctly. If you were awarding 1 point per pound and using 10st as a base would see a horse carrying 11st being given an extra 14 points, whilst a horse running off 9st would have 14pts deducted from its final rating. Weight is a very tricky thing though, to be honest, as you would generally need to take distance into account as well. A horse winning with 11st on its back would generally need a bigger rating boost over 2 miles than it would over 5 furlongs - but (as with lots of this stuff) that's down to preference.

Rather than go through the previous run ratings for that last race, your better bet is to post up how you rated last night's race mate, and how you came to those ratings. Saves folks going back through the cards then to compare ratings. And I'll have to do it manually as I use a totally seperate rating system, so one race is easier for me.

One note regarding selections - I'm using a totally different scale and rating system, so I can't directly compare, but I notice that you have Lisahane Bog rated quite low there, which is probably based on the slow race it finished 2nd in last time out. The performance was decent, but the pace didn't suit it - hence the low rating. There is a danger to just referring to the last rating as this race shows. Lisahane Bog is a much better animal than the last run suggests and on my numbers (which are calculated totally differently, and are on a different scale, but I'll include them for reference) the horse ran an 87 last time, but 111, 123, 127 and 135 the four times before that. So you can't take the last rating as fact there - your selection process has to be organic.

Run us step-by-step through what you do to rate the last race at Kempton (ie. the result of last night's race rather than the ratings for each horse's previous run, Gold Rules 1st, Lisahane Bog 2nd, Soul Heaven 3rd) and we/I can pick it all apart and see if you're getting it right, so to speak

With regards to those PDFs - they are interesting, but I threw them out as soon as I saw that the author was using the horse's price in calculating the track variant! Anything that even refers to the SP when calculating a speed rating is (in my opinion) inherently flawed. The betting markets get it wrong - often - and saying "That horse should be rated higher because it went off at 33/1 and finished second" is just random.

Have a look at this site:- MAFSPEED SPEED RATINGS It breaks down a relatively standard rating process in step-by-step format and makes a decent amount of sense. It should serve as a decent base if the tutorials you've read so far are gettign you tied in knots!

**fragle123** · **18-04-10**, **11:49 AM**

Hi Ret
ok heres what i have when rating it the way it is explained on the mafspeed site for kempton
Gold Rules 92.19
Lisahane Bog 91.94
Soul Heaven 91.78
White Devil 91.72
Sheer Force 91.53
Thrust Control 90.59
Hemera 88.84
Ana Moutabahi 88.59
First i got the times,Classes and Pars for each race on the card so these were
17:35 +0.74 C6 6.9
18:05 +1.25 C6 6.9
18:35 +1.17 C6 6.9
19:05 +3.1 C5 6.6
19:35 +0.96 C6 6.9
20:05 +1.53 C6 6.9
20:35 +0.63 C4 6.3
21:05 +0.13 C4 6.3
i then subtracted the times from the class pars
6.9 -0.74 =6.16
6.9 -1.25 =5.65
6.9 -1.17 = 5.73
6.6 - 3.1 = 3.5
6.9 - 0.96 =5.94
6.9 - 1.53 =5.37
6.3 -0.63 =5.67
6.3 -0.13 =6.17
i then removed the 2 slowest times and added the remainder and div by 4
6.9 -0.74 =
6.9 -1.25 =5.65
6.9 -1.17 = 5.73
6.6 - 3.1 = 3.5
6.9 - 0.96 =5.94
6.9 - 1.53 =5.37
6.3 -0.63 =5.67
6.3 -0.13 =
This gave me 1.43 so i took this and added the time for gold rules 0.13*5 which came out at 7.81 and the deducted this from 100 to give me the rating 92.19
and for the rest of the runners in the race i got the beaten lenghts and div by 8 so Lisahane Bog was 2L behind 2/8 = 0.25 and subtracted this from the winners rating.
I noticed that in the mafspeed doc it said to remove the 2 fastest and the 2 slowest but in the tutorial on here it says just remove the 2 slowest as i read that Spook had tried the 2 and 2 but was not happy with doing it this way so i just removed the 2 slowest which i am right in thinking that they are the 2 biggest times?.
I can see that those first ratings i had done for the 21:05 were wrong after reading the mafspeed doc so thanks for pointing me in that direction, i have also noticed that the RP has changed its standard times from when i first used them last sept, the times have quickened for some of the distances.

**Retroneous** · **18-04-10**, **03:59 PM**

Originally Posted by fragle123

I noticed that in the mafspeed doc it said to remove the 2 fastest and the 2 slowest but in the tutorial on here it says just remove the 2 slowest as i read that Spook had tried the 2 and 2 but was not happy with doing it this way so i just removed the 2 slowest which i am right in thinking that they are the 2 biggest times?.

Just running through your post now, but I would say that the going allowance is one of the most difficult things to judge. If you stick to a hard and fast rule of removing the two fastest and two slowest, or just the two slowest, or even just the two fastest, then you will find yourself running into problems. I generally remove anything that looks "wrong."

For example, if a card (with all low class races) shows me +4.02, +4.31, +9.00, +3.75, +5.00 and +4.87, I'll check that +9.00 and see if there's any obvious reason that the going allowance sticks out a country mile. It could be a falsely run race, and that will affect ratings for every horse on the card. Removing the two slowest and two fastest isn't a "wrong" way of doing it of course, its just that at some meetings, that'll still leave you with a very dodgy going allowance.

Take this card, for example - all races are the same class (6), with "standard" fields. Official going is described as Good-to-Soft to begin with. Imagine we aren't using class pars or anything like that, just calculating the basic times:-

Race 1: +2.00
It starts to rain here.
Race 2: +4.00
Some heavy showers break through.
Race 3: +8.00
The rain holds off.
Race 4: +8.00
A torrential downpour begins.
Race 5: +12.00
And carries on...
Race 6: +13.00

If I take out Races 1, 2, 5 and 6, that gives us a going allowance of +8.00. Now, race 1 is going to be judged to have been run 6 seconds BELOW standard, despite actually being 2 seconds over in what were relatively standard (albeit slightly soggy) conditions.

I know you're focusing on the all-weather, and that takes some of the guesswork out of it straight away, as AW courses generally don't ride 10 seconds per mile slow or 10 seconds per mile quick as turf courses can.

You have to use your nose, so to speak, to sniff out the clues. There's no direct formula that can be applied to every meeting. Its a skill that you'll develop over time.

**Retroneous** · **18-04-10**, **04:09 PM**

Originally Posted by fragle123

slowest times and added the remainder and div by 4
6.9 -0.74 =
6.9 -1.25 =5.65
6.9 -1.17 = 5.73
6.6 - 3.1 = 3.5
6.9 - 0.96 =5.94
6.9 - 1.53 =5.37
6.3 -0.63 =5.67
6.3 -0.13 =
This gave me 1.43 so i took this and added the time for gold rules 0.13*5 which came out at 7.81 and the deducted this from 100 to give me the rating 92.19

You're all good up until here, I think.

You need to add the remainder and divide by 6, instead of 4, which I make to be 5.31, not 1.43.

This should give you Gold Rules on 73.45, and using the "beaten lengths/8" for the runners-up puts Lisahane Bog on 73.20 in second.

**fragle123** · **18-04-10**, **04:16 PM**

Thanks for that and thinking about it, it makes sense to look at what you have in front of you and make your decision based on that data rather than just sticking to a set rule for removing numbers.
Reading that link you put up has also helped me alot and thanks again for that.

**Retroneous** · **18-04-10**, **05:21 PM**

No problem chap, anytime

**TopCat** · **08-05-10**, **03:41 AM**

Originally Posted by Natagora

I'll post my Southwell Standard Times as a guideline, these are what a Beyer rated horse of 108 should be hitting:

5f - 57.70
6f - 74.30
7f - 87.30
1m - 100.40
1m3f - 143
1m4f - 155.40
1m6f - 182.30
2m - 217

The 2 Mile one would be the dodgiest as there are few races over that distance at Southwell. Be intrigued to see how others relate to these times.

To make your own, I'd grab a copy of Beyer on Speed and Mordin on Time asap. I used a hybrid of their methods to develop my times and ratings.

Natagora,
Would you like to elaborate on how you use a hybrid of the two methods, because I am trying to come to grips with speed ratings down here in Australia, but I was under the impression that Beyer's methods didn't work on turf.

Also, in the spirit of sharing, here are a couple of links to articles which all speed raters would find intriguing:

The Relationship Between Time

This one shows the effects of fatigue in a racehorse, and

Speed in the Thoroughbred

goes into speed. Both are a bit scientific but easy enough to follow. And one clearly shows that Secretariat was robbed!

Terry

05-02-10, 12:30 AM	#1 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Rating the cards I wonder if one of you guys could help me by explaining how you actually go about the process rating a days racing as an example tomorrows card for Lingfield step by step as i am still trying to figure it all out, i have my set of standard times but dont know where i go from there to actually rating a race or if somebody could point me to an online document that i could dl that explains it thanks.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

05-02-10, 10:02 PM	#2 (permalink)
andthereoff Banned. Join Date: Oct 2008	I find this hard too! mate, I have the data alright but when attacking a card there is so many angles to take it confuses me, some days I might stick to good times on a certian ground or good times over a certian trip, even just times at a certain track - there is rarely a horse who can back up a good speed rating because there isnt the opportunities or they are simply not good enough too; If a horse has won over 8f at Ascot in a serious time carrying 8-4, their is only 12 or so meetings at Ascot a year, with his rating shooting through the roof thats the class of race angle out the window, he's entered at Haydock next and 5 runs after that is at Salisbury, Epsom, Sandown, Pontefract and Hamilton - where am I meant to go from here, This leaves me with only 1 solution, the ground!?, it could be 2 years untill a horse gets them same Ascot conditions, is he the same horse? NO 9 out of 10 times he's not, is he still in training, NO 9 times out of 10 he's not so whats the use of the data I have? Spook insights, cheers fella

05-02-10, 10:51 PM	#3 (permalink)
Spook World Class Join Date: Sep 2005	I'm not completely sure I understand what's being asked in honesty?

05-02-10, 10:51 PM	#4 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	what i am thinking of doing is just sticking to the AW tracks in the uk, i have read the tut thread but i cant see the wood for the trees as it doesnt make any sense to me at all as to where you just start to rate a days card in the first place im just so confused as to what i am suppose to do ?.

06-02-10, 12:06 AM	#5 (permalink)
Spook World Class Join Date: Sep 2005	You can't rate a card unless you hold ratings in the first place. In essence you need to build up a library of performances and then use this to assess the various horses entered. This will take about 3-6 weeks to compile, although the A/W circuit is much easier and you can probably begin after about 3 weeks. You could back-calculate them however if you want to hit the ground running. You'd probably be as well to give some thought though regarding how you intend to access information by way of a data retrieval system

06-02-10, 10:38 AM	#6 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Hi spook ok so what i have to do is begin rating the past results that i have collected upto now using the winning times and class pars ect to be able to rate horses in future races. As for retrieving data i got some code off Konex to download all the past results from the sporting life archives and put them into excel, and i have been playing about with a spreadsheet to do the calculations for me the past few weeks with some success so then once i have given speed ratings to these horses i then use them to assess the races that they are going to run in, so i presume that i should then be looking at how they achieved those ratings in eg Class,Distance,Course and then make a judgment as to how well they will perform from those is that right. At the moment i have Mordins Standard Times to use but would i be better off using the RP's or creating my own to use or would it not matter which, its just that i have read in a few threads here that Mordins may not be the best ones to use. Sorry if the questions seems obvious or daft but not having done this before i just want to be sure i am doing it the correct way in the first place. thanks

06-02-10, 11:59 AM	#7 (permalink)
Spook World Class Join Date: Sep 2005	I wouldn't touch Mordins with a barge pole. He uses some hypothetical 'wonder horse' and one that presumebly has a JATO unit strapped on its back. Mordins standards are underwritten by possibility and not necessarily the result of what has happened. Personally I'd prefer to base a rating on demonstrable evidence and not some contrarians opinion which one almost suspects has been formulated as much with keeping his name in the spotlight. If you're going to aim at a/w tracks only then I suppose you coudl calcualte your own and if you get them right, you'll gain an edge. If you get them wrong, you'll be hampered. A/W tracks are quite predicatble and subject to much less fluctuation though and for the most part I wouldn't have thought you'd go far wrong simply using the RP's. The only things that disrupts A/W times are strong wind, and particularly high humidity, which can occur on any track. The rails aren't subjected to the same level of movement so the only unique thing concerns the harrowing of the track which you'll learn to detect. What normally happens is that it starts to ride slightly quicker before suddenly going slow again. If you used your own standards there would be a chance you'd simply hit this cycle at some point and draw incorrect conclusions. The RP's standards have been set over a much longer time frame and can absorb them thus as they'll just show up under the normal laws of variance. I'd save yourself the time and use the RP's. The only places I'd give serious sconsideration re-calcualting course standards at the moment would be Ascot, York, Haydock and Doncaster all of whom have had singificant drainage put in during recent times. I also think there's a strong case for re-calibrating French standards as they frequently throw up time based ground descriptions that would indicate that they haven't been updated for years

06-02-10, 01:50 PM	#8 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Hi spook thanks for that info so for now i will use the RP's standards until i get a bit better at understanding what i am looking at/for, so at least now i can stop banging my head on the wall wondering how the horses get rated for future races as your just using it's previous ones for the new race it's running in ie a horses previous ratings are used for it running today and i take it my thinking that class distance course they were earned atare what i should be considering when making my judgement about its performance if it was running today. thanks for your advice and help.

08-02-10, 10:13 PM	#9 (permalink)
The Duke The Iron Horse Join Date: Mar 2005 Twitter: @OHRacing Location: West Midlands	Just a quick note, the RP Standard Times on the Racing Post website are completely ****ed up to what they should be. Warwick is a particular example of this, you'd be better off doing your own.

08-02-10, 11:18 PM	#10 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Hi Duke Ok thanks i will look into it and see how i get on

15-02-10, 10:40 PM	#11 (permalink)
Natagora Claimer Join Date: May 2008	I'll post my Southwell Standard Times as a guideline, these are what a Beyer rated horse of 108 should be hitting: 5f - 57.70 6f - 74.30 7f - 87.30 1m - 100.40 1m3f - 143 1m4f - 155.40 1m6f - 182.30 2m - 217 The 2 Mile one would be the dodgiest as there are few races over that distance at Southwell. Be intrigued to see how others relate to these times. To make your own, I'd grab a copy of Beyer on Speed and Mordin on Time asap. I used a hybrid of their methods to develop my times and ratings.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

25-02-10, 12:43 AM	#12 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Hi Natagora If you dont mind me asking how are beyers speed ratings worked out, do you think they are better than Mordins way of doing them.

14-04-10, 10:17 AM	#14 (permalink)
Natagora Claimer Join Date: May 2008	Beyer compiles charts for each track with times and ratings alongside one another per tenth of a second, so for example.... 99.8 109 99.7 108 99.6 107 Etc etc. Obviously the gaps in ratings get bigger per tenth of a second as the distance of race decreases and vice versa. I find it an excellent system, especially if using your own standards. I then apply Mordin's class pars to the times of the races in order to discover the going allowance.

14-04-10, 11:10 AM	#15 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Hi Ret Yeh the maths? thats the problem well for me anyway as i keep doubting that im doing it the right way as it never loooks right to me but i am still trying to learn this and want it to be right otherwise its a waist of time as the ratings will be wrong and it just seems that Mordins way dont feel right to me?. I have found a formula and code that had applied to be patented back in the 80s for a software program by the looks of it and when i use this i have noticed that it gives ratings that are fairly close to the RP's, when i say close i mean within 10 - 15pts either way and as i understand it the RP ratings are adjusted for weight. One other thing i have noticed this week when looking over the tutorial again at the beginning when going over the donny figures for calculating the track variance they have changed between then and now, so they cant be trusted. But i think these figures maybe better for me to use and i may stick with them after i have done some testing with them.

14-04-10, 11:23 PM	#16 (permalink)
Retroneous Group 1 Level Join Date: Mar 2009 Location: Exeter, Devon	Your best bet, fragle123, is to post your method of rating the horses. Post it step-by-step with numbers and the workings out for maybe the first three in a particular race. That way, we can all take a look, and chip in if we see anything massively wrong. Posting the code that you've found might be interesting too - always good to see new methods and share and share alike and whatnot

15-04-10, 10:06 AM	#17 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Hi Ret Its all in 2 pdf docs here is a link for anybody to get them from Free File Hosting Made Simple - MediaFire I dont know anything about programming but it looks a bit like VBA to me probably wrong but you will understand it better than me. There is still some of it i am trying to figure out but the basic formula for a raw rating is distance60/time+0.077dist -4.92/0.008 this is used upto 9 furlongs and their is another one for more than 9 as you will see in the docs. it would be interesting to see what you all think. Also when using weight to adjust your ratings as i understand it if you use say 10st as a base and a horse is 9st then you deduct 14 from the rating and above the base you add it on, do i have that the right way round? i did see it mentioned here somewhere but not been able to find it lol. But any advice would be helpful as i just seem to be going round in circles with this, keep reading the tut over and over but what i am intending to do is just concentrate on the 4 AW courses here as the turf would be to much for me to keep upto date with because of my job i would just be playing catch up all the time.

15-04-10, 10:22 AM	#18 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Last night i just quickly rated the horses in the last race at kempton that had run on the aw before and this is what i got Thrust Control 67 6th Hemera 81 7th Lisahane Bog 66 2nd Sheer Force 58 4th Soul Heaven 56 3rd Ana Moutahabi 52 8th These were just ratings without any adjustments or other things added to that formula in the above post, so do these ratings agree with what anyone else has for these horses.

15-04-10, 06:19 PM	#19 (permalink)
Retroneous Group 1 Level Join Date: Mar 2009 Location: Exeter, Devon	With regards to weight, you have the basics down correctly. If you were awarding 1 point per pound and using 10st as a base would see a horse carrying 11st being given an extra 14 points, whilst a horse running off 9st would have 14pts deducted from its final rating. Weight is a very tricky thing though, to be honest, as you would generally need to take distance into account as well. A horse winning with 11st on its back would generally need a bigger rating boost over 2 miles than it would over 5 furlongs - but (as with lots of this stuff) that's down to preference. Rather than go through the previous run ratings for that last race, your better bet is to post up how you rated last night's race mate, and how you came to those ratings. Saves folks going back through the cards then to compare ratings. And I'll have to do it manually as I use a totally seperate rating system, so one race is easier for me. One note regarding selections - I'm using a totally different scale and rating system, so I can't directly compare, but I notice that you have Lisahane Bog rated quite low there, which is probably based on the slow race it finished 2nd in last time out. The performance was decent, but the pace didn't suit it - hence the low rating. There is a danger to just referring to the last rating as this race shows. Lisahane Bog is a much better animal than the last run suggests and on my numbers (which are calculated totally differently, and are on a different scale, but I'll include them for reference) the horse ran an 87 last time, but 111, 123, 127 and 135 the four times before that. So you can't take the last rating as fact there - your selection process has to be organic. Run us step-by-step through what you do to rate the last race at Kempton (ie. the result of last night's race rather than the ratings for each horse's previous run, Gold Rules 1st, Lisahane Bog 2nd, Soul Heaven 3rd) and we/I can pick it all apart and see if you're getting it right, so to speak With regards to those PDFs - they are interesting, but I threw them out as soon as I saw that the author was using the horse's price in calculating the track variant! Anything that even refers to the SP when calculating a speed rating is (in my opinion) inherently flawed. The betting markets get it wrong - often - and saying "That horse should be rated higher because it went off at 33/1 and finished second" is just random. Have a look at this site:- MAFSPEED SPEED RATINGS It breaks down a relatively standard rating process in step-by-step format and makes a decent amount of sense. It should serve as a decent base if the tutorials you've read so far are gettign you tied in knots!

18-04-10, 11:49 AM	#20 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Hi Ret ok heres what i have when rating it the way it is explained on the mafspeed site for kempton Gold Rules 92.19 Lisahane Bog 91.94 Soul Heaven 91.78 White Devil 91.72 Sheer Force 91.53 Thrust Control 90.59 Hemera 88.84 Ana Moutabahi 88.59 First i got the times,Classes and Pars for each race on the card so these were 17:35 +0.74 C6 6.9 18:05 +1.25 C6 6.9 18:35 +1.17 C6 6.9 19:05 +3.1 C5 6.6 19:35 +0.96 C6 6.9 20:05 +1.53 C6 6.9 20:35 +0.63 C4 6.3 21:05 +0.13 C4 6.3 i then subtracted the times from the class pars 6.9 -0.74 =6.16 6.9 -1.25 =5.65 6.9 -1.17 = 5.73 6.6 - 3.1 = 3.5 6.9 - 0.96 =5.94 6.9 - 1.53 =5.37 6.3 -0.63 =5.67 6.3 -0.13 =6.17 i then removed the 2 slowest times and added the remainder and div by 4 6.9 -0.74 = 6.9 -1.25 =5.65 6.9 -1.17 = 5.73 6.6 - 3.1 = 3.5 6.9 - 0.96 =5.94 6.9 - 1.53 =5.37 6.3 -0.63 =5.67 6.3 -0.13 = This gave me 1.43 so i took this and added the time for gold rules 0.13*5 which came out at 7.81 and the deducted this from 100 to give me the rating 92.19 and for the rest of the runners in the race i got the beaten lenghts and div by 8 so Lisahane Bog was 2L behind 2/8 = 0.25 and subtracted this from the winners rating. I noticed that in the mafspeed doc it said to remove the 2 fastest and the 2 slowest but in the tutorial on here it says just remove the 2 slowest as i read that Spook had tried the 2 and 2 but was not happy with doing it this way so i just removed the 2 slowest which i am right in thinking that they are the 2 biggest times?. I can see that those first ratings i had done for the 21:05 were wrong after reading the mafspeed doc so thanks for pointing me in that direction, i have also noticed that the RP has changed its standard times from when i first used them last sept, the times have quickened for some of the distances.

18-04-10, 04:16 PM	#23 (permalink)
fragle123 Claimer Join Date: Oct 2009 Location: Grimsby	Thanks for that and thinking about it, it makes sense to look at what you have in front of you and make your decision based on that data rather than just sticking to a set rule for removing numbers. Reading that link you put up has also helped me alot and thanks again for that.

18-04-10, 05:21 PM	#24 (permalink)
Retroneous Group 1 Level Join Date: Mar 2009 Location: Exeter, Devon	No problem chap, anytime