SPEED RATINGS

[MCLARKE]

Fro @black rabbit

Good luck with the ratings Mike.  I think I posted a note ages ago about a guy I use to see regularly at the races back in the 70’s/80’s ‘Milo’ he’d a system where he would note from the Raceform Handicap Weekend edition (don’t think it’s published anymore) all horses which had won a race in an above average time (denoted by prefix ‘B’ in result column) the horse would have to have one of the following caveats from the race reader – ‘ran on well’ ‘quickened clear’ easily won’ ‘comfortably’ he claimed he got the system from ‘Phil Bull’ which was probably bullshite but he did appear to keep finding frequent winners.

The ’Rub’ being that any horse who could quicken at the end of a fast run race warranted serious respect in future races, I have loosely applied this over the years when a horse from these categories has been unplaced a few times and then slipped on to a nice handicap weight.

[MCLARKE]

Yes I believe the ability to quicken at the end of a fast race is a rare attribute

I may have some old data from a few years back that I can test it on, I'll add it to my speed to do list

[Zilzalian]

19 hours ago, MCLARKE said:

Yes I believe the ability to quicken at the end of a fast race is a rare attribute

I may have some old data from a few years back that I can test it on, I'll add it to my speed to do list

 

20 hours ago, MCLARKE said:

Fro @black rabbit

Good luck with the ratings Mike.  I think I posted a note ages ago about a guy I use to see regularly at the races back in the 70’s/80’s ‘Milo’ he’d a system where he would note from the Raceform Handicap Weekend edition (don’t think it’s published anymore) all horses which had won a race in an above average time (denoted by prefix ‘B’ in result column) the horse would have to have one of the following caveats from the race reader – ‘ran on well’ ‘quickened clear’ easily won’ ‘comfortably’ he claimed he got the system from ‘Phil Bull’ which was probably bullshite but he did appear to keep finding frequent winners.

The ’Rub’ being that any horse who could quicken at the end of a fast run race warranted serious respect in future races, I have loosely applied this over the years when a horse from these categories has been unplaced a few times and then slipped on to a nice handicap weight.

I had a similar/same system as your guy but noted all the placed horse that achieved a bigger speed figure than the winner due to weight differences, my biggest earner was a horse of Mark Johnsons called Yavanas Pace (switched from Ireland) that came forth in a class C handicap it ended up winning a Group 1 after winning the November handicap and many other races usually at decent odds. There is gold in them there speed figures lads.

[MCLARKE]

I have now analysed my data for 2023 and the results are not as encouraging as they were for 2022.

3 of the top 10 were recorded on the all weather at Newcastle.

Of the remaining 4 were recorded on heavy going, 2 on soft and 1 on good to soft.

 

This prompts 2 questions.

1. Should all weather be treated as a separate entity.

2. Should speed figures recorded on soft or heavy going be ignored ot at least be treated with suspicion.

 

[In Play Race Reader]

IMO Newcastle is a bit of an anomaly as regards Speed Ratings as most races are run at a crawl early then sprint finish especially on the round course ,if your using all races to get your going allowance there’s going to be some difficulty ,,I know raceforms SR’s for Newcastle are always out of sync,,Cover up recently won off 83 and was rated 123 ,which is par with Group 1 ,,I use race par finishes to find out which races have been close to truly run 

[Trotter]

14 minutes ago, In Play Race Reader said:

Cover up recently won off 83 and was rated 123 ,which is par with Group 1 ,,I use race par finishes to find out which races have been close to truly run 

The Racing Post speed rating for that recent win was 87 ..... which seems a lot more reasonable than 123 !

[In Play Race Reader]

12 minutes ago, Trotter said:

The Racing Post speed rating for that recent win was 87 ..... which seems a lot more reasonable than 123 !

Yes ,I think I rated it about 78 ,,which is decent mid-top handicapper on my scale ,,but Raceform do often go big  at Newcastle,,used to always be big on Southwell’s old surface 

[Zilzalian]

2 hours ago, MCLARKE said:

I have now analysed my data for 2023 and the results are not as encouraging as they were for 2022.

3 of the top 10 were recorded on the all weather at Newcastle.

Of the remaining 4 were recorded on heavy going, 2 on soft and 1 on good to soft.

 

This prompts 2 questions.

1. Should all weather be treated as a separate entity.

2. Should speed figures recorded on soft or heavy going be ignored ot at least be treated with suspicion.

 

Point 1. There is an argument for separation but in my opinion it is a small argument.

Point 2. If the ground is soft through heavy then they are useful on that going but for all other goings by my system they never come out on top anyway unless the horse in question has run to a figure on the other "goings". Your going adjustments based on the little i know about your "going corrections" are obviously flawed.

[MCLARKE]

If I restrict my rating to horses that achieved a rating on good or good to firm going then the top ones are -

+ 0.03 PADDINGTON. Sandown Group 1. No surprise as this was the Coral-Eclipse. Won its next race.

-0.03 ONCE MORE FOR LOVE. Newmarket Class 2 Handicap. Ran 2.38 secs faster than the group 3 contest on the same card. Unraced since.

-0.10 HUKUM. Ascot Grade 1. Lost its next race (Arc De Triomphe).

-0.12 REGIONAL. Haydock Group 1. Unraced since.

-0.19 FOX JOURNEY. Newbury Class 4 Handicap. Ran 2 times more on turf (placed at 11/1 and 5/1).

-0.25 SHALLOW HAL. Carlisle Class 5 Apprentice Hadicap. Unplaced in next 5 runs.

-0.26 REGIONAL. Haydock Listed. Lost its next race but then won the Group 1 Betfair Cup at Haydock with an even higher speed rating at 10/1.

-0.26 AUDIENCE. Newmarket Group 3. 4 runs since, lost 3 times on soft or good to soft going but came 2nd at 20/1 on good to firm.

-0.31 FIRST SIGHT. Yarmouth Class 3 Handicap. Transferred to Meydan where it lost 2 races.

-0.33 LOVE LIES. Ayr Class 5 Maiden. Won next race at 10/11 but then lost next 2.

-0,34 WARM HEART. York Group 1. Won 2 of its next 4 runs in France and the US at odds of 3/1 and 12/5.

 

Most of these appear logical and would have shown a profit if backed afterwards.

I will be interested how REGIONAL performs on its reappearance since it recorded 2 out of the top 7 speed figures. It has an entry in the Group 2 Duke Of York Stakes at York on the 15th of May.

My next step now I am reasonably happy with my calculations is a detailed analysis of each horse in 2023.

 

 

[MCLARKE]

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

8 hours ago, Zilzalian said:

Your going adjustments based on the little i know about your "going corrections" are obviously flawed.

A bit harsh my friend !

Remember what I said at the beginning "Be gentle with me, I'm an absolute beginner at this".

I may amend them at some stage but I'll run with them to start off with.

I think the logic may be flawed for longer distances on soft to heavy going because as the distance gets longer the impact will be greater in terms of seconds per mile compared to the shorter distances.

As an example, the Bath meeting on heavy going on the 18th October.

The longer distances are considerably slower compared to the standard than the shorter distances. This results in the sprints having much higher speed figures than longer races. 

I have a nagging doubt I'm missing something in my logic, I'm sure somebody will point it out.

[In Play Race Reader]

7 hours ago, MCLARKE said:

 

Edited April 4 by In Play Race Reader

[Zilzalian]

8 hours ago, MCLARKE said:

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

A bit harsh my friend !

Remember what I said at the beginning "Be gentle with me, I'm an absolute beginner at this".

I may amend them at some stage but I'll run with them to start off with.

I think the logic may be flawed for longer distances on soft to heavy going because as the distance gets longer the impact will be greater in terms of seconds per mile compared to the shorter distances.

As an example, the Bath meeting on heavy going on the 18th October.

The longer distances are considerably slower compared to the standard than the shorter distances. This results in the sprints having much higher speed figures than longer races. 

I have a nagging doubt I'm missing something in my logic, I'm sure somebody will point it out.

My apologies i didn't for one second intend my response to seem harsh i was merely answering the two points. For what its worth i have never solved the riddle of why on average the highest speed figures are distance biased. One thing i do know is that because of the nature of sprints you can get varied results when applied to actual races especially in handicaps however the more interesting speed figures are 08f+ and are stronger the more distance.

[Trotter]

8 hours ago, MCLARKE said:

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

A bit harsh my friend !

Remember what I said at the beginning "Be gentle with me, I'm an absolute beginner at this".

I may amend them at some stage but I'll run with them to start off with.

I think the logic may be flawed for longer distances on soft to heavy going because as the distance gets longer the impact will be greater in terms of seconds per mile compared to the shorter distances.

As an example, the Bath meeting on heavy going on the 18th October.

The longer distances are considerably slower compared to the standard than the shorter distances. This results in the sprints having much higher speed figures than longer races. 

I have a nagging doubt I'm missing something in my logic, I'm sure somebody will point it out.

This is something that I've considered over the years and my solution is to ignore it !

The logic is based on the fact that stayers are not going to be running against sprinters in the future so what I am comparing is stayers with other stayers. There's no need to compare stayers with sprinters

I guess this partly depends on how you are using your ratings. You appear to be using them to compile a list of horses to follow whereas I've always used them to study a specific race the following afternoon

So if I'm looking at a 6f sprint tomorrow generally all the ratings I have for those horses will be from other sprint races. It doesn't matter what other horses are doing in 12 furlong races ...... it's has no impact on my group of sprinters that are running in the 3.30 tomorrow

And of course vice versa if I'm looking at a 14 furlong race tomorrow

As regards calculating a going allowance ......... I've always tended to take more notice of times over sprint trips on the basis that there's less option for 'dawdling' which you might get over longer trips. If any figures at a meeting look way out of line with the other races I'll tend to ignore them for the purpose of getting an average across the meeting ...... races that get ignored would tend to be the longer ones, particularly low grade with small fields. I guess in mathematical terms I'm looking for a median not an average. So I'm happy to ignore outliers.

[MCLARKE]

1 hour ago, Zilzalian said:

My apologies i didn't for one second intend my response to seem harsh i was merely answering the two points. For what its worth i have never solved the riddle of why on average the highest speed figures are distance biased. One thing i do know is that because of the nature of sprints you can get varied results when applied to actual races especially in handicaps however the more interesting speed figures are 08f+ and are stronger the more distance.

Don't worry, I don't take things personally.

The beauty of speed figures is that we all start off with the same data but interpret it in different ways and come up with different conclusions.

The proof of it all will be if I can make any money out of them !

[MCLARKE]

1 hour ago, Trotter said:

This is something that I've considered over the years and my solution is to ignore it !

The logic is based on the fact that stayers are not going to be running against sprinters in the future so what I am comparing is stayers with other stayers. There's no need to compare stayers with sprinters

I guess this partly depends on how you are using your ratings. You appear to be using them to compile a list of horses to follow whereas I've always used them to study a specific race the following afternoon

So if I'm looking at a 6f sprint tomorrow generally all the ratings I have for those horses will be from other sprint races. It doesn't matter what other horses are doing in 12 furlong races ...... it's has no impact on my group of sprinters that are running in the 3.30 tomorrow

And of course vice versa if I'm looking at a 14 furlong race tomorrow

As regards calculating a going allowance ......... I've always tended to take more notice of times over sprint trips on the basis that there's less option for 'dawdling' which you might get over longer trips. If any figures at a meeting look way out of line with the other races I'll tend to ignore them for the purpose of getting an average across the meeting ...... races that get ignored would tend to be the longer ones, particularly low grade with small fields. I guess in mathematical terms I'm looking for a median not an average. So I'm happy to ignore outliers.

The problem is that the results of the longer races are used to calculate the going allowance. So in my example the winner of the 6f race has an exceptional speed figure due to the going allowance including the results of the longer races.

There are however solutions.

1. Ignore longer distance races when calculating the going allowance. The downside of this is that the sample size that is used to calculate the going allowance becomes even smaller.

2. Ignore the slowest race. Again the downside is the same as in point 1.

3. Ignore meetings where the going allowance is very high. I am attracted to this although I would need to decide what the cut off point would be.

4. Ignore meetings on heavy / soft going. This would be a good and easy solutions.

5. Introduce a distance adjustment (using similar logic to the class adjustment). An added complication and if it is going dependent then it might not solve the problem.

All a long way from when I first started years ago where I simply picked the second favourite !

 

[kroni]

Interesting read in here, I had a dabble with speed ratings years ago and this thread jogged my memory on some of the issues and ideas I had about them.

As @Trotter has mentioned with going allowances, judging longer races the same way as sprints is a problem, or more so comparing races round a bend(s) with races on straight courses, because rail movements making distances longer, also in dry weather and firmer going the bends can be watered to make them safe so could potentially be slower. Possible solution is to have different going allowances for straight and round course?

As for race times, there are things that make them seem better or worse than they actually are which can be hard to gauge, like head winds and tail winds, rail movements, horses getting a perfect tow into the race much like a cyclist being in the cover of the peloton and getting a lead out for a sprint finish.

One example would be Minzaal @MCLARKE mentioned above, shame it never raced again and no doubt was a good performance to win by 3.75 lengths over 6f, but everything was in its favour that day to achieve that course record. The ground was very quick especially on the straight course which led to there being 33 non-runners on the card which is a crazy amount. Minzaal got a perfect tow into the race off a very fast pace. Bit of a random one as the time was also recorded by hand which is odd and adds doubts to the accuracy of it. Also things that may of affected going allowances is the extra yardage on all the races on the round course and the fact the back straight and bends had been watered, where as the sprint course was not watered.

[Zilzalian]

Occams Razor lads, Occams Razor....

[Zilzalian]

On 4/3/2024 at 8:53 PM, MCLARKE said:

If I restrict my rating to horses that achieved a rating on good or good to firm going then the top ones are -

+ 0.03 PADDINGTON. Sandown Group 1. No surprise as this was the Coral-Eclipse. Won its next race.

-0.03 ONCE MORE FOR LOVE. Newmarket Class 2 Handicap. Ran 2.38 secs faster than the group 3 contest on the same card. Unraced since.

-0.10 HUKUM. Ascot Grade 1. Lost its next race (Arc De Triomphe).

-0.12 REGIONAL. Haydock Group 1. Unraced since.

-0.19 FOX JOURNEY. Newbury Class 4 Handicap. Ran 2 times more on turf (placed at 11/1 and 5/1).

-0.25 SHALLOW HAL. Carlisle Class 5 Apprentice Hadicap. Unplaced in next 5 runs.

-0.26 REGIONAL. Haydock Listed. Lost its next race but then won the Group 1 Betfair Cup at Haydock with an even higher speed rating at 10/1.

-0.26 AUDIENCE. Newmarket Group 3. 4 runs since, lost 3 times on soft or good to soft going but came 2nd at 20/1 on good to firm.

-0.31 FIRST SIGHT. Yarmouth Class 3 Handicap. Transferred to Meydan where it lost 2 races.

-0.33 LOVE LIES. Ayr Class 5 Maiden. Won next race at 10/11 but then lost next 2.

-0,34 WARM HEART. York Group 1. Won 2 of its next 4 runs in France and the US at odds of 3/1 and 12/5.

 

Most of these appear logical and would have shown a profit if backed afterwards.

I will be interested how REGIONAL performs on its reappearance since it recorded 2 out of the top 7 speed figures. It has an entry in the Group 2 Duke Of York Stakes at York on the 15th of May.

My next step now I am reasonably happy with my calculations is a detailed analysis of each horse in 2023.

 

 

Will 100% support the assessment of regional very unusual for a horse to return 4 consecutive numbers above 130.

[MCLARKE]

7 hours ago, kroni said:

Possible solution is to have different going allowances for straight and round course?

The number of races that are used to calculate the going allowance is very small from a statistical perspective so the calculation of the going allowance is already going to be very approximate. To reduce it from 6 to 3 would make it even more inaccurate.

 

[Trotter]

Another advantage of all weather racing for speed figures is that there are lots of meetings ........ and each course rarely goes more than a few days without racing during the winter season. So not only does the going tend to be consistent all over the track but you can compare the going with the meeting a few  days earlier to see if they're within reasonable limits of each other as the going isn't likely to change massively from meeting to meeting a few days apart

 

 

[In Play Race Reader]

It is a good time to look back and take stock, including of one aspect of all-weather racing which tends to be misrepresented. That is the nature of the surfaces on which the horses race, and in particular those surfaces’ speeds as inferred from the times recorded on them.

It is a myth that the speeds of all-weather surfaces do not vary, though they do avoid the real extremes that sometimes occur on turf.

That has not stopped every last all-weather fixture run so far this year from being described officially as “standard ” at Chelmsford, Lingfield, Newcastle, Southwell and Wolverhampton, and all 18 meetings in that time at Kempton from being described as “standard to slow”.

Time analysis paints a different picture of regular variations in implied surface speeds, some of them far from negligible.

A going allowance can be calculated which reflects the ability a horse would have to show to equal standard time, carrying weight-for-age in a well-run race, with lower indicating quicker conditions and higher the opposite.

This is how they look, summarised for all meetings in January to March inclusive in 2024. Where different time-based going allowances existed on the same card (such as following in-card track maintenance), the figure for the first race has been used.

 

Overall, there is agreement between time-based going allowances and official going descriptions 58.9% of the time, which leaves nearly half that are inaccurate as judged by such means. Wolverhampton easily leads the way in being “correct” 83.3% of the time.

Another way of looking at the data is to consider the maximum and minimum going allowances at each course, the range (the difference between maximum and minimum), and the standard deviation (which is a traditional measure of variance).

 

Wolverhampton and Lingfield perform best in terms of narrowness of range and lowness of standard deviation. They could arguably justify their unswervingly uniform approach to going description - the other courses less so.

As an illustration, a range of 50lb is equivalent to the difference between “good to soft” and “good to firm, tending to firm” on turf, or the difference between Frankel running a given time and a 90-rated handicapper doing the same.

Newcastle is a rather tricky case, in which wind sometimes plays a major part on its straight mile, which I have allowed for as much as possible. Unfortunately, British racing has ignored requests to record wind speed and direction as a matter of course as races are run, and nor does it publicise significant in-card track maintenance.

Even if you treat Newcastle as an outlier, 35% of other cards seem to have been miscategorised. Incidentally, the going allowance there on All-Weather Finals Day was 93 on my figures, or just about slap bang in the middle of “standard”, as described.

Either way, you can be sure that track conditions vary a good deal more than the official version would have you believe, and this has implications for speed, stamina, sectionals, race-positioning, distance betting, handicapping, times, in-play betting, and more besides.

Perhaps British racing will one day acknowledge that these things matter to those who bet on the sport and thereby help to keep the show on the road. But I won’t be holding my breath.

Sectional Spotlight

Sectional Spotlight Blog Archive

Dubai World Cup and all-weather surfaces

Tuesday, 2nd April 2024

The Flat is back!

Tuesday, 26th March 2024

Cheltenham 2024: Day three & day four review

Saturday, 16th March 2024

Cheltenham 2024: Day one & day two review

Thursday, 14th March 2024

Dublin Racing Festival reflections

Tuesday, 6th February 2024

Shake up in Supreme and Ryanair markets

Tuesday, 16th January 2024

The Road to Cheltenham

Tuesday, 9th January 2024

Festive round-up

Wednesday, 3rd January 2024

More blog posts

Other Blogs

Hollie Doyle

Thursday, 4th April 2024

Hong Kong Diary

Thursday, 4th April 2024

Sectional Spotlight

Tuesday, 2nd April 2024

Kevin Blake

Tuesday, 2nd April 2024

Eyecatchers

Tuesday, 2nd April 2024

Surfing The Sand

Saturday, 30th March 2024

Betfair Ambassadors

Saturday, 30th March 2024

Saturday Horses to Follow: Doncaster

Friday, 22nd March 2024

Jamie Lynch

Thursday, 21st March 2024

Declan Rix

Thursday, 21st March 2024

More Blogs

[In Play Race Reader]

Just now, In Play Race Reader said:

It is a good time to look back and take stock, including of one aspect of all-weather racing which tends to be misrepresented. That is the nature of the surfaces on which the horses race, and in particular those surfaces’ speeds as inferred from the times recorded on them.

It is a myth that the speeds of all-weather surfaces do not vary, though they do avoid the real extremes that sometimes occur on turf.

That has not stopped every last all-weather fixture run so far this year from being described officially as “standard ” at Chelmsford, Lingfield, Newcastle, Southwell and Wolverhampton, and all 18 meetings in that time at Kempton from being described as “standard to slow”.

Time analysis paints a different picture of regular variations in implied surface speeds, some of them far from negligible.

A going allowance can be calculated which reflects the ability a horse would have to show to equal standard time, carrying weight-for-age in a well-run race, with lower indicating quicker conditions and higher the opposite.

This is how they look, summarised for all meetings in January to March inclusive in 2024. Where different time-based going allowances existed on the same card (such as following in-card track maintenance), the figure for the first race has been used.

 

Overall, there is agreement between time-based going allowances and official going descriptions 58.9% of the time, which leaves nearly half that are inaccurate as judged by such means. Wolverhampton easily leads the way in being “correct” 83.3% of the time.

Another way of looking at the data is to consider the maximum and minimum going allowances at each course, the range (the difference between maximum and minimum), and the standard deviation (which is a traditional measure of variance).

 

Wolverhampton and Lingfield perform best in terms of narrowness of range and lowness of standard deviation. They could arguably justify their unswervingly uniform approach to going description - the other courses less so.

As an illustration, a range of 50lb is equivalent to the difference between “good to soft” and “good to firm, tending to firm” on turf, or the difference between Frankel running a given time and a 90-rated handicapper doing the same.

Newcastle is a rather tricky case, in which wind sometimes plays a major part on its straight mile, which I have allowed for as much as possible. Unfortunately, British racing has ignored requests to record wind speed and direction as a matter of course as races are run, and nor does it publicise significant in-card track maintenance.

Even if you treat Newcastle as an outlier, 35% of other cards seem to have been miscategorised. Incidentally, the going allowance there on All-Weather Finals Day was 93 on my figures, or just about slap bang in the middle of “standard”, as described.

Either way, you can be sure that track conditions vary a good deal more than the official version would have you believe, and this has implications for speed, stamina, sectionals, race-positioning, distance betting, handicapping, times, in-play betting, and more besides.

Perhaps British racing will one day acknowledge that these things matter to those who bet on the sport and thereby help to keep the show on the road. But I won’t be holding my breath.

Sectional Spotlight

Sectional Spotlight Blog Archive

Dubai World Cup and all-weather surfaces

Tuesday, 2nd April 2024

The Flat is back!

Tuesday, 26th March 2024

Cheltenham 2024: Day three & day four review

Saturday, 16th March 2024

Cheltenham 2024: Day one & day two review

Thursday, 14th March 2024

Dublin Racing Festival reflections

Tuesday, 6th February 2024

Shake up in Supreme and Ryanair markets

Tuesday, 16th January 2024

The Road to Cheltenham

Tuesday, 9th January 2024

Festive round-up

Wednesday, 3rd January 2024

More blog posts

Other Blogs

Hollie Doyle

Thursday, 4th April 2024

Hong Kong Diary

Thursday, 4th April 2024

Sectional Spotlight

Tuesday, 2nd April 2024

Kevin Blake

Tuesday, 2nd April 2024

Eyecatchers

Tuesday, 2nd April 2024

Surfing The Sand

Saturday, 30th March 2024

Betfair Ambassadors

Saturday, 30th March 2024

Saturday Horses to Follow: Doncaster

Friday, 22nd March 2024

Jamie Lynch

Thursday, 21st March 2024

Declan Rix

Thursday, 21st March 2024

More Blogs

Thought this was interesting by Simon Rowlands 

[Zilzalian]

10 hours ago, MCLARKE said:

The number of races that are used to calculate the going allowance is very small from a statistical perspective so the calculation of the going allowance is already going to be very approximate. To reduce it from 6 to 3 would make it even more inaccurate.

 

Run your system with the going correction and then without compare the list and see if it makes much difference.

[MCLARKE]

3 hours ago, Zilzalian said:

Run your system with the going correction and then without compare the list and see if it makes much difference.

I'll add it to my list

I've got so many angles I want to test / analyse, I did say a while back that I was ignoring racing to concentrate on my share investments but I've been dragged back in, it's much more interesting than shares

[Zilzalian]

3 hours ago, MCLARKE said:

I'll add it to my list

I've got so many angles I want to test / analyse, I did say a while back that I was ignoring racing to concentrate on my share investments but I've been dragged back in, it's much more interesting than shares

Suggestion. Royal Ascot is not really that far off so if we speed rate the same 3 races per day lets see how they compare we can also add in @Trotter if he fancy's it  and see if we can learn something from the results.

[MCLARKE]

5 hours ago, Zilzalian said:

Suggestion. Royal Ascot is not really that far off so if we speed rate the same 3 races per day lets see how they compare we can also add in @Trotter if he fancy's it  and see if we can learn something from the results.

Sounds a good test, I'm almost at a stage where I will be able to rate all races.

[MCLARKE]

First run of the numbers and a bit of a sobering conclusion.

I have analysed the 2023 flat season and recorded the results of the next run.

I have split the results between the top 50% and bottom 50%.

The top 50% show a strike rate of 13% and an AE of 0.98.

The bottom 50% show a strike rate of 6% and an AE of 1.02.

This is obviously not the way to use speed figures. In fact Nick Mordin did indicate this when he quoted "in a series of 200 races, the horse with the highest speed rating on its last run won 30% of the time but returned a 18% loss".

I need to do a bit more digging to find that gold !

 

[Zilzalian]

8 hours ago, MCLARKE said:

First run of the numbers and a bit of a sobering conclusion.

I have analysed the 2023 flat season and recorded the results of the next run.

I have split the results between the top 50% and bottom 50%.

The top 50% show a strike rate of 13% and an AE of 0.98.

The bottom 50% show a strike rate of 6% and an AE of 1.02.

This is obviously not the way to use speed figures. In fact Nick Mordin did indicate this when he quoted "in a series of 200 races, the horse with the highest speed rating on its last run won 30% of the time but returned a 18% loss".

I need to do a bit more digging to find that gold !

 

" the horse with the highest speed rating on its last run"  can you expand on this? how would you rate/bet it in a race? on its last run or its best speed figure?

[MCLARKE]

I think I would bet based on its best speed figure.

However as with most bets you cannot use speed figures without taking other factors into considerations, specifically odds, draw etc. and also the conditions of today's race, is it on the same going / distance etc. as when the speed figure was obtained.

Hopefully all will be revealed as I dig further into my data.

[harry_rag]

Must admit this thread largely flies over my head in a way that is probably karma for all that spread based waffle I come out with but I’m not surprised you can’t make a profit by just using each horse’s last run; it’s far too little data.

Each horse’s best figure would be more relevant but not in isolation. Ideally you want a rating for each horse that is based on its aggregate form, a bit like ratings for a football team based on their last x games.

I think the idea of a few of you comparing ratings for the same good quality races is a good one.