SPEED RATINGS

[Zilzalian]

I will put this one out there before the flat gets going and it might interest @MCLARKE and anyone else that is thinking of Speed Rating races.

Unless you are prepared to put in over 30 hours per week you will never cover all the races so you need to pick your pool of horses carefully. So i will suggest a pool of horses that will serve you well all year round and be more efficient after the first year which will give you time to experiment and that is 2yo races on the all weather. If you do the figures for these races you will need to bear in mind a few things..

1. The type of all weather surface, Tapeta (Newcastle, Southwell, Wolves) the others inc Dundalk are all Polytrack.

2. Don't be in a hurry, collect the data and use it to small stakes or save it until you have a substantial database. Minimum of 9 months to a year.

3. Don't get too bogged down with draw bias the better AW Jockeys/trainers can overcome it.

4. Identify the better AW jockeys/trainers (that consistently overcome the draw, it is good information.

5. keep the data updated daily (don't put it of until weekend etc)

6. understand the difference between wet and dry sand (think of a beach)

7. check the results and the race comments of every race against your data.

8. bear in mind you only need to rate a race down to 6 and often 3 places at the most, this is because the figures below 5th or 6th are of very little use.

[MCLARKE]

Right, I'm diving straight into the deep end !

Be gentle with me, I'm an absolute beginner at this.

I was going to concentrate on the all weather but as @Trotter says, we are now probably past the peak season for the AW.

After rereading some of Nick Mordin's work he does state that the racing post standard times are valuable but compiling your own average times would produce more accurate figures. So, in for a penny, in for a pound.

I will start with Ascot (purely based on the alphabet).

I have taken data for 2022 (I am going to use 2023 as test data). For each race I have listed the date, class, distance and winning time. Most of this data I have been able to download automatically.

 1. Convert the data into seconds per mile. This shows the the fastest time was in a 5 furlongs sprint and the slowest time in the Queen Alexandra Stakes over nearly 2 miles 6 furlongs. Logical so far.

2. Calculate the average time for each distance. 

2. Calculate the average winning time by distance and by class. So as an example for 6 furlongs there were 8 races at class 1 with an average time of 73.73 seconds, class 2 was 74.03 seconds and class 3 was 75.13 seconds. Convert these into seconds per mile.

3. For each race subtract the winning time from the average for that distance and class.

4. Calculate the average for that meeting. This is the going allowance for that meeting. The highest adjustment (1.53 secs / mile) was in June when the going was good to firm, the lowest adjustment (-2.85 secs / mile) was in October when the going was good to soft.

I disagree to a certain extent with Mr Mordin at this stage, he takes a subjective view of each meeting, in an example of a 6 race meeting he excludes one race because it was "probably won by an exceptional horse" and for 2 others he excludes them because he would "guess that they were slow-run affairs". That leaves him with just three races to calculate the going allowance.

5. Subtract the going allowance from the winning time.

6. Subtract the adjusted winning time from the average winning time for that distance. The highest figure was for Pyledriver in the King George VI and Queen Elizabeth stakes. Intuitively this feels right. The winning time was 149.49 secs (100.00 secs / mile), the average winning time for that distance is 102.62 secs / mile. The going adjustment for that meeting was -0.10 secs / mile, giving Pyledriver a rating of +2.72. The lowest rating was Oh This Is Us in a class 3 classified stakes with a rating of -3.87.

7. Convert the ratings to a scale. Not strictly necessary but conceptually better. The top rated horse could be given a rating of 140 and the lowest 0, possibly as a comparative to the BHB official ratings.

8. Calculate the ratings for the remaining horses in each race. I will use Nick Mordin's assumption that every fifth of a second is counted as one length.

 

So there it is, my starter for ten. It seems logical but I may be missing something. The next step is to calculate these numbers going back a number of years for Ascot and also for all other courses. Time consuming but doable using Excel. Hopefully the effort will be worth it.

 

 

 

[Trotter]

Good luck with that @MCLARKE 

To be honest I've never considered compiling my own race times. Being very much a 'pen and paper' guy it would just take me too much time. 

I think it would be fairly easy for me to become 'someone who compiles speed ratings' rather than 'someone who uses speed ratings to help find winners' ! ......... so I'm quite happy sticking to the winter AW and having the summer off !

I'm still toying with the idea of rating the top race days ......... maybe days with Group Races on the card but I'll see how that goes

 

[MCLARKE]

1 hour ago, Trotter said:

I'm still toying with the idea of rating the top race days ......... maybe days with Group Races on the card but I'll see how th

It would be good if you could do that, we could then compare our results

[Zilzalian]

2 hours ago, Trotter said:

Good luck with that @MCLARKE 

To be honest I've never considered compiling my own race times. Being very much a 'pen and paper' guy it would just take me too much time. 

I think it would be fairly easy for me to become 'someone who compiles speed ratings' rather than 'someone who uses speed ratings to help find winners' ! ......... so I'm quite happy sticking to the winter AW and having the summer off !

I'm still toying with the idea of rating the top race days ......... maybe days with Group Races on the card but I'll see how that goes

 

When i first did speed figures i had to use index cards, this was very time consuming, with the advent of computerisation it was 1000 x faster especially when sorting the data for each new race. I must admit i loved having all those index cards but they were a bugger to keep sorting and flipping through even though they were alphabetical to the name of the horse.

[Zilzalian]

2 hours ago, MCLARKE said:

Right, I'm diving straight into the deep end !

Be gentle with me, I'm an absolute beginner at this.

I was going to concentrate on the all weather but as @Trotter says, we are now probably past the peak season for the AW.

After rereading some of Nick Mordin's work he does state that the racing post standard times are valuable but compiling your own average times would produce more accurate figures. So, in for a penny, in for a pound.

I will start with Ascot (purely based on the alphabet).

I have taken data for 2022 (I am going to use 2023 as test data). For each race I have listed the date, class, distance and winning time. Most of this data I have been able to download automatically.

 1. Convert the data into seconds per mile. This shows the the fastest time was in a 5 furlongs sprint and the slowest time in the Queen Alexandra Stakes over nearly 2 miles 6 furlongs. Logical so far.

2. Calculate the average time for each distance. 

2. Calculate the average winning time by distance and by class. So as an example for 6 furlongs there were 8 races at class 1 with an average time of 73.73 seconds, class 2 was 74.03 seconds and class 3 was 75.13 seconds. Convert these into seconds per mile.

3. For each race subtract the winning time from the average for that distance and class.

4. Calculate the average for that meeting. This is the going allowance for that meeting. The highest adjustment (1.53 secs / mile) was in June when the going was good to firm, the lowest adjustment (-2.85 secs / mile) was in October when the going was good to soft.

I disagree to a certain extent with Mr Mordin at this stage, he takes a subjective view of each meeting, in an example of a 6 race meeting he excludes one race because it was "probably won by an exceptional horse" and for 2 others he excludes them because he would "guess that they were slow-run affairs". That leaves him with just three races to calculate the going allowance.

5. Subtract the going allowance from the winning time.

6. Subtract the adjusted winning time from the average winning time for that distance. The highest figure was for Pyledriver in the King George VI and Queen Elizabeth stakes. Intuitively this feels right. The winning time was 149.49 secs (100.00 secs / mile), the average winning time for that distance is 102.62 secs / mile. The going adjustment for that meeting was -0.10 secs / mile, giving Pyledriver a rating of +2.72. The lowest rating was Oh This Is Us in a class 3 classified stakes with a rating of -3.87.

7. Convert the ratings to a scale. Not strictly necessary but conceptually better. The top rated horse could be given a rating of 140 and the lowest 0, possibly as a comparative to the BHB official ratings.

8. Calculate the ratings for the remaining horses in each race. I will use Nick Mordin's assumption that every fifth of a second is counted as one length.

 

So there it is, my starter for ten. It seems logical but I may be missing something. The next step is to calculate these numbers going back a number of years for Ascot and also for all other courses. Time consuming but doable using Excel. Hopefully the effort will be worth it.

 

 

 

This reads a bit like how i first started (we have read the same book) but it didn't take me long to find so many flaws ( you have already found 1 🙂) so i went fishing 3 days and nights and gave it a shit load of thought came back and totally simplified it and it worked pretty well now here is something that might interest you and @Trotter when i applied the methodology to the aw racing something quirky happened namely that it became less good at predicting the winner but very good at predicting the places my best guess on this, it is because back then AW racing was very low grade piss poor horses running on just 3 tracks all different. Southwell (fibresand), lingfield and then Wolves all three tracks have changed since then and it has now become pretty close to turf (equivalent to good to firm) believe it or not so you can at a pinch switch between the two for the majority of Runners. I would be quite happy to have a face to face chat with you on messenger or some such to discuss things  but i will not put it on here it would be pointless and take up to much time.

[Trotter]

19 minutes ago, Zilzalian said:

When i first did speed figures i had to use index cards, this was very time consuming, with the advent of computerisation it was 1000 x faster especially when sorting the data for each new race. I must admit i loved having all those index cards but they were a bugger to keep sorting and flipping through even though they were alphabetical to the name of the horse.

I used to have a card index file for 3 mile chasers but even with such a specific group of horses it pretty quickly got out of hand. The problem in the olden days was that the Weekender or Sporting Life, later The Post, would only give you up to 3 previous runs in the form. If you wanted to know more than that you had to collect the weekly results pull out from the Raceform Weekly ?

I became self employed in 1996 so gave up betting and racing for the next 12 years ........ when I came back to it I couldn't believe the amount of stuff available on the internet as regards form. Everything was so much easier and complete

Unfortunately though I was never computer literate enough to get into all the data management ..... I'd never used a computer at all until I was about 4 years into self-employment when I thought I'd better take the plunge or I'd soon be Un-employed rather than self-employed. So I only learned how to do what I needed to do for work

I'm still a pen and paper chap for keeping records of speed figures etc 

 

[Zilzalian]

1 minute ago, Trotter said:

I used to have a card index file for 3 mile chasers but even with such a specific group of horses it pretty quickly got out of hand. The problem in the olden days was that the Weekender or Sporting Life, later The Post, would only give you up to 3 previous runs in the form. If you wanted to know more than that you had to collect the weekly results pull out from the Raceform Weekly ?

I became self employed in 1996 so gave up betting and racing for the next 12 years ........ when I came back to it I couldn't believe the amount of stuff available on the internet as regards form. Everything was so much easier and complete

Unfortunately though I was never computer literate enough to get into all the data management ..... I'd never used a computer at all until I was about 4 years into self-employment when I thought I'd better take the plunge or I'd soon be Un-employed rather than self-employed. So I only learned how to do what I needed to do for work

I'm still a pen and paper chap for keeping records of speed figures etc 

 

The funny thing is i had to teach myself how to use data bases and how to write formulae to produce/automate the inputs so the incentive was there and it was quite enjoyable which always helps even though the cat got kicked (figuratively speaking) a good few times before i mastered it to the level i needed.

[MCLARKE]

I am fortunate in that my financial background meant I had to use computers for much of my work. I remember back in 1981 when I was a trainee accountant at British Steel there was a shiny new PC in the corner of the room that everybody was scared of but I took the plunge and started to use it. In those days it was a twin floppy disk set up with the system disk in one side and the work disk in the other (512kb capacity, I used to hate the disk full message). Lotus123 was the spreadsheet of choice, I used to love that (especially writing macros) and it took me till the mid 2000s to ween myself off it and move to Excel.

Before I retired a few years ago I just used to bet on Saturdays, typing data from the Racing Post into my spreadsheets (I was too tight to subscribe to a data service).

Since being retired I have had more time (too much according to my wife) to analyse and test data and I am reasonably happy with the returns from my AW systems.

I would hope to have all my inital work done in the next couple of weeks before the flat season starts to kick in properly. Where it leads me we will see.

[harry_rag]

9 hours ago, Trotter said:

Unfortunately though I was never computer literate enough to get into all the data management ..... I'd never used a computer at all until I was about 4 years into self-employment when I thought I'd better take the plunge or I'd soon be Un-employed rather than self-employed. So I only learned how to do what I needed to do for work

I'm still a pen and paper chap for keeping records of speed figures etc 

Wow, it’s hard for me to imagine doing anything meaningful in terms of systems and data analysis just with pen and paper. I guess you could create the ratings, it would be just a lot more hard work. But it would be impossible to do any meaningful analysis or manipulation. I’d say I’m obviously ahead of you in that I make good use of spreadsheets day in day out but I’m behind the other two in terms of databases and importing data.

I think it would be worthwhile you getting to grips with entry level spreadsheets. Anyone can type in words and numbers to start with, it’s as easy as writing them down. Doing the basic sums with your numbers is easy enough too and then being able to sort the data by various criteria. Put it this way, anything you can feasibly do with a pen and paper isn’t going to take much learning. It’s also fairly easy to go beyond what you can do with pen and paper. Even if you don’t go beyond that stage you’ll have sped up the process and extended the scope of what you can do.

Think of it this way, you can learn how to build and repair a car or you can just learn to drive one. If you simply learnt how to be a passenger spreadsheet wise you’d reap a huge benefit in terms of time and effort.

[Zilzalian]

49 minutes ago, harry_rag said:

Wow, it’s hard for me to imagine doing anything meaningful in terms of systems and data analysis just with pen and paper. I guess you could create the ratings, it would be just a lot more hard work. But it would be impossible to do any meaningful analysis or manipulation. I’d say I’m obviously ahead of you in that I make good use of spreadsheets day in day out but I’m behind the other two in terms of databases and importing data.

I think it would be worthwhile you getting to grips with entry level spreadsheets. Anyone can type in words and numbers to start with, it’s as easy as writing them down. Doing the basic sums with your numbers is easy enough too and then being able to sort the data by various criteria. Put it this way, anything you can feasibly do with a pen and paper isn’t going to take much learning. It’s also fairly easy to go beyond what you can do with pen and paper. Even if you don’t go beyond that stage you’ll have sped up the process and extended the scope of what you can do.

Think of it this way, you can learn how to build and repair a car or you can just learn to drive one. If you simply learnt how to be a passenger spreadsheet wise you’d reap a huge benefit in terms of time and effort.

Add to this that Apache open office is free and in my experience it is perfectly adequate for all we need to do and so much more. Their excel spreadsheet equivalent is simpler to use than MS office excel (i have used both in the past)

[harry_rag]

3 minutes ago, Zilzalian said:

Add to this that Apache open office is free and in my experience it is perfectly adequate for all we need to do and so much more. Their excel spreadsheet equivalent is simpler to use than MS office excel (i have used both in the past)

I can endorse Libre Office in a similar vein. The fact that perfectly adequate free software is available means you might at least give it a go!

[Trotter]

46 minutes ago, harry_rag said:

Wow, it’s hard for me to imagine doing anything meaningful in terms of systems and data analysis just with pen and paper. I guess you could create the ratings,

To be fair, I am into my 70s ........  the old brain is a bit too addled to start doing complicated things with computers, particularly as I never learned to use them apart from what I taught myself

I'm happy enough to study the form in the old fashioned way and just form the ratings and keep them on a Word document ....... I don't keep a record for each horse just jot down the winners rating then I can quickly look up a horses last 2 or 3 figures by reference to the race winner of the races he ran in

I've never really been attracted to systems or data analysis ........ I'm interested in the horses that are running today and I'm happy enough studying today's race without reference to what other horses have done in the past. I've got the form of today's runners and that's all I need !

[Trotter]

I see that Chazzesmee is declared for the Lincoln on Saturday

He won the Irish Lincoln a few days ago and I did that day's ratings and had him with a figure of 90 which I think is pretty low for a top handicap winner. I intend to rate the Donny card on Saturday .... be interesting to see what he does.

[Systemight]

As a young man i did speed figures for two year olds on the flat(there was no all weather racing back then) and had some success,and certainly a lot better knowledge of the form book.

However, while i started with enthusiasm,i soon found the workload a bit of a chore,particularly when the losing runs hit.What finally put me off then(as still happens now),was the constant moving of the rails making distances longer or shorter than the officially stated.Thus (to my eyes at least), made the ratings at best unreliable.

Interesting to see him mentioned here, and as an author i think Nick Mordin had few peers,indeed his book "Betting For a Living" was in my opinion the best book on horse racing betting that i have ever read.His book on speed ratings however lagged behind Andy Beyer's work "Beyer On Speed"(again only in my opinion).

One must remember that Nick Mordin's phoneline (as there was back then) was regularly listed as one of the most unprofitable Horse Racing service around, costing his devoted followers dear-clearly the man could talk the talk,but couldn't really walk the walk-well not when the money was down anyway.

Nick was last heard of having retired from Journalism to go and live in South Africa,a big loss to us all, as he had many thought provoking and innovative ideas...

[Trotter]

One problem if you really want to get obsessive about compiling your own figures is course configuration in the sense that courses might have variable going on different parts of the track. If you think about a large tract of land out in the country with dips and hollows and undulations it's evident that water will gather in the lower lying parts of the track ........ and the going will be softer there. So for example there might be a marked going difference between the round course which goes out into the country and a straight track........ and many race courses have both round and straight tracks

Very difficult to assess this with a small sample size of maybe 3 races on the straight and 3 on the round at a particular meeting

That's one advantage of AW racing ....... the tracks are flat (Ling has a bit of undulation) so the 'going' tends to be uniform across all distances

 

[MCLARKE]

On 3/21/2024 at 10:23 PM, Systemight said:

As a young man i did speed figures for two year olds on the flat(there was no all weather racing back then) and had some success,and certainly a lot better knowledge of the form book.

However, while i started with enthusiasm,i soon found the workload a bit of a chore,particularly when the losing runs hit.What finally put me off then(as still happens now),was the constant moving of the rails making distances longer or shorter than the officially stated.Thus (to my eyes at least), made the ratings at best unreliable.

Interesting to see him mentioned here, and as an author i think Nick Mordin had few peers,indeed his book "Betting For a Living" was in my opinion the best book on horse racing betting that i have ever read.His book on speed ratings however lagged behind Andy Beyer's work "Beyer On Speed"(again only in my opinion).

One must remember that Nick Mordin's phoneline (as there was back then) was regularly listed as one of the most unprofitable Horse Racing service around, costing his devoted followers dear-clearly the man could talk the talk,but couldn't really walk the walk-well not when the money was down anyway.

Nick was last heard of having retired from Journalism to go and live in South Africa,a big loss to us all, as he had many thought provoking and innovative ideas...

I wonder if Nick's lack of success with his phoneline was the same reason with good tipsters today, the bets quickly become overbet.

I will persevere for the time being. I was intending to produce speed figures for both GB and Irish racing, however the Irish don't have the same classification ratings as the GB so I'll stick with GB for the time being. 

[MCLARKE]

53 minutes ago, Trotter said:

One problem if you really want to get obsessive about compiling your own figures is course configuration in the sense that courses might have variable going on different parts of the track. If you think about a large tract of land out in the country with dips and hollows and undulations it's evident that water will gather in the lower lying parts of the track ........ and the going will be softer there. So for example there might be a marked going difference between the round course which goes out into the country and a straight track........ and many race courses have both round and straight tracks

Very difficult to assess this with a small sample size of maybe 3 races on the straight and 3 on the round at a particular meeting

That's one advantage of AW racing ....... the tracks are flat (Ling has a bit of undulation) so the 'going' tends to be uniform across all distances

There are many problems with speed ratings, I suppose the evidence for me will be if I can make any money out of it !

The small sample sizes used to calculate the all important going allowance is a concern, especially if you take out the outliers. The going often changes during a meeting and is probably different between the first race and the last race.

 

[Trotter]

3 minutes ago, MCLARKE said:

There are many problems with speed ratings, I suppose the evidence for me will be if I can make any money out of it !

The small sample sizes used to calculate the all important going allowance is a concern, especially if you take out the outliers. The going often changes during a meeting and is probably different between the first race and the last race.

 

I try to mitigate the problem by only taking my going allowance from handicap races where I know the rough level of ability of the horses in the race. I work out in advance what rating horses of this class should achieve, compare this to what they actually achieve, strike out the fastest and slowest and use the average of the others as my going allowance

Although I'm not going to bother with UK Turf Racing (too much!) I did rate Donny on Saturday 

I had the expected figure for the 4 handicaps in time order, the actual figure and the difference

91- 42 - slow 49
99 - 54 - slow 45
85 - 28 - slow 57
85 - 31 - slow 54

these are all within the same ballpark which suggests they might be reasonable !

I strike out the 45 and the 57 and average the other 2 to give my final figure of Slow 52

So the Lincoln winner gets 54 plus 52 = 106

 

 

 

[Zilzalian]

2 hours ago, MCLARKE said:

There are many problems with speed ratings, I suppose the evidence for me will be if I can make any money out of it !

The small sample sizes used to calculate the all important going allowance is a concern, especially if you take out the outliers. The going often changes during a meeting and is probably different between the first race and the last race.

 

One thing i learnt early on was that it is far too easy to overcomplicate things and far to difficult to stop yourself overcomplicating things. At the end of the day it is just a tool and we can all use a chisel to remove a screw. It is no coincidence that the most popular and successful ratings system was Phil Bull's Time and Form = Timeform.

[MCLARKE]

22 hours ago, Zilzalian said:

One thing i learnt early on was that it is far too easy to overcomplicate things and far to difficult to stop yourself overcomplicating things. At the end of the day it is just a tool and we can all use a chisel to remove a screw. It is no coincidence that the most popular and successful ratings system was Phil Bull's Time and Form = Timeform.

I'm trying to keep it simple but not being very successful !

With regard to calculating the going adjustment I have five options (at least)

1 Use all the races at a meeting

2 Ignore the slowest and the fastest

3 Take the 3 fastest

4 Only include the handicaps

5 Take the 3 with biggest fields

I'll start with option 1 but will be interested in others opinions

 

 

 

[Trotter]

25 minutes ago, MCLARKE said:

I'm trying to keep it simple but not being very successful !

With regard to calculating the going adjustment I have five options (at least)

1 Use all the races at a meeting

2 Ignore the slowest and the fastest

3 Take the 3 fastest

4 Only include the handicaps

5 Take the 3 with biggest fields

I'll start with option 1 but will be interested in others opinions

 

 

 

I've tried all those in the past ..... with the exception of number 5 (but I can see the point of it)

In some ways I don't think it matters as long as you use the same method consistently. Where I have changed tends to be at the start of a new AW winter season where I decide to do something different this time round

At the moment I'm happy with my method of only using handicaps. I'm using RP Standard Times which measure something like 'a rating that a 100 rated horse carrying 9 stone should achieve over a mile on good ground'

As you look at a handicap race you can quickly see what a horse carrying 9 stone would be rated. Let's say it's 88....... you'd expect it run 12 lbs or 4 lengths slower than RP Standard. If it runs 18lbs or 6 lengths slower then the going allowance for that race is 'Slow 6lbs or Slow 2 lengths' ........ I use lengths

then you can average out the going allowances for the qualifying races or strike out the fastest and slowest etc to get your average for the meeting ....... which you can then apply to all the races

so I think my method takes account of the class of horses running at the meeting

 

[Zilzalian]

2 hours ago, MCLARKE said:

I'm trying to keep it simple but not being very successful !

With regard to calculating the going adjustment I have five options (at least)

1 Use all the races at a meeting

2 Ignore the slowest and the fastest

3 Take the 3 fastest

4 Only include the handicaps

5 Take the 3 with biggest fields

I'll start with option 1 but will be interested in others opinions

 

 

 

I once used option 2, then i thought hmm sometimes there are 7 races 8 and even 9 so what i did was use just 5 races i knocked off the fastest and slowest (if only 6 races i took out the extreme)

[Trotter]

I can see the point more in knocking out the slow races. I guess the faster races are probably run at a truer pace whereas the slower races might just be dawdling or excessive holding up or no-one wants to front run and therefore more likely to produce false times or times that reflect something other than the conditions

[MCLARKE]

I've parked setting my own standard times for the time being after going round in circles many times.

I've decided initially to use the Raceform Standard Times.

So my revised method is :-

 

1. For each race compare the actual time to the standard time.

2. Convert these into seconds per mile.

3. Calculate the average variance for each class of race. This gives the seconds per mile slower than standard time by class.

 

4. For each race amend the variance by the class adjustment.

5. For each meeting calculate the average of the adjusted variances. This is the going allowance for that meeting.

6. For each race amend the variance by the going allowance. This gives the final variance.

 

This has created some interesting and unexpected numbers. In my top ten I would have expected to see the winners of the big group 1 races at Royal Ascot etc. However this is not the case. The top rated are :-

1. + 1.68. MINZALL. Haydock Group 1. No surprise here as it was a new course record. 

2. + 1.62. LIVE IN THE DREAM. Sandown Class 2 Handicap. Won 4 times from the next 14 runs at prices of 28/1, 12/1, 5/2 and 11/8. This include the Group 1 Nunthorpe at York so obviously a class horse. 

3. + 0.82. ANMAAT. Haydock Group 3. Won 2 from the next 3 at prices 23/10 and 13/10.

4. + 0.67. TANMAWWY. Ffos Las Class 3 Handicap. A poor record afterwards with 1 win at 13/8 from 7 runs before being shipped to the US. 

5. + 0.63. DUBAI MIRAGE. Salisbury Class 2 Handicap. Won 1 from the next 4 at 11/2.

6. + 0.63. GOOD LUCK FOX. Catterick Class 5 Handicap. Ran 11 times since without success (although none of them were on heavy going). Not the sort of horse you would expect to see with a high speed figure. The going was heavy and all the races were run upto 27 seconds slower. Perhaps speed figures achieved on heavy going should be ignored or at least treated with caution.

7. + 0.59. REGAL REALITY. Windsor Group 3. The time was 1.99 seconds faster than standard despite the going being good to soft. Won 1 from next 7 at 5/1.

8. + 0.43. AZANO. Redcar Class 3 Handicap. Faster than standard and also carrying top weight of 10 st 2 lbs would indicate that this was a top class performance. However only won 1 of the following 15 races at 5/1.

9. + 0.39. SHEIKH MAZ MAHOOD. From the same meeting as GOOD LUCK FOX. Won 1 from the next 4 at 15/2.

10. + 0.36. SEANTRABH. Catterick Class 5 Novice. Going was Soft (Heavy In Places). Won 1 from the next 7 at 15/2.

11 + 0.36. PYLEDRIVER. Ascot Group 1. Won 1 from next 2 at 7/2.

12 + 0.29. STAY ALERT. Newbury Group 3. Faster than standard but no success in the next 5 races (although they were all higher class races).

13 + 0.28. CHIPSTEAD. Catterick (again !) Class 2 Handicap. Won 3 from the next 14 at 16/1, 12/1 and 12/1.

 

A lot more work to do but so far looks encouraging.

 

 

 

[Trotter]

Looking good McLarke ! ........... 👍

Good to see that the class variances do increase as you go down the ranks ..... this is what you'd expect logically and it would have been a bit of a problem if it was any different 

your new system is now very like mine as regards the steps involved in coming up with a final figure 

you obviously have the advantage of running a huge database whereas I work out the ratings day by day at the end of the meeting and store that info for future reference. So for example if I was looking at the 3.00 at Newcastle tomorrow and I'd been keeping up AW ratings all winter I'd have a stored rating for every run of all the runners over the last 5 months

How would you go about looking at this race ?

By the way tomorrow's card at Newcastle might be a good card for anyone wanting to have a crack at producing their own speed figures after the meeting has finished ...... good quality racing on an AW surface

[Trotter]

just noticed from your list that Chipstead is running tomorrow ....... Newc 3.35

[Zilzalian]

19 minutes ago, Trotter said:

just noticed from your list that Chipstead is running tomorrow ....... Newc 3.35

My figures suggest Coachello, there will be no hiding in this so you need a 7f horse in my opinion, I don't write of Chipstead  (its in my f/c) but i just think Coachello will be stronger at the finish.

[MCLARKE]

3 hours ago, Trotter said:

Looking good McLarke ! ........... 👍

Good to see that the class variances do increase as you go down the ranks ..... this is what you'd expect logically and it would have been a bit of a problem if it was any different 

your new system is now very like mine as regards the steps involved in coming up with a final figure 

you obviously have the advantage of running a huge database whereas I work out the ratings day by day at the end of the meeting and store that info for future reference. So for example if I was looking at the 3.00 at Newcastle tomorrow and I'd been keeping up AW ratings all winter I'd have a stored rating for every run of all the runners over the last 5 months

How would you go about looking at this race ?

By the way tomorrow's card at Newcastle might be a good card for anyone wanting to have a crack at producing their own speed figures after the meeting has finished ...... good quality racing on an AW surface

Yes it is encouraging that the results appear logical, notwithstanding that Catterick appears to produce fast horses.

I am still in the analysis stage at the moment, I haven't yet worked out how to use them !

Strangely the service I use for my data only provides the finishing time in the horse's previous race. I can calculate the finishing time in the current race for most of the historical data (probably 95%). For 2023 it's probably 80% and for current races I will beed to input the data manually. I have all the rest of the data so it'll probably only take a couple of minutes per day.

My next stage is to take each horse for 2022 and analyse it's speed figure with it's future races and see what conclusions I come too.

Once I have all my data completed for 2023 I would be able to take any race and look at the speed figures achieved historically and calculate who has the best speed figure. It may then be a case of deciding if the current race has similar conditions to the race in which the best speed figure was achieved (distance, going, OR, time of year etc).

 

[MCLARKE]

21 hours ago, Trotter said:

just noticed from your list that Chipstead is running tomorrow ....... Newc 3.35

Surprised at his price 33/1 considering he'd won over this distance on the AW 4 weeks ago and he'd also won off this OR before

Came 3rd