SPEED RATINGS

[MCLARKE]

Today's top rated

Hamilton 7.50

SACRED FALLS

[MCLARKE]

I continue to meander down several winding roads with this, the final destination is still some way off. I suppose if it was easy everybody would do it.

I have now analysed all turf races for 2018 - 2022 for horses that have achieved a speed rating of 67 or above.

The top criteria I have then applied are 

Number of runners in qualifying race > 7. It seems logical that bigger fields will produce more truely run races.

Qualifying course not Doncaster. Not sure of the logic of this other than the first meeting of the year is at Doncaster and these early season types may not produce the goods later on. The AE is 0.93.

Horse not bred in USA. Again not sure of the logic of this but it is a strong statistic. The AE is 0.87.

Distance not more than 2 lengths longer than the qualifying race.

Speed rating in last race not more than 24 higher than previous race. This would indicate that horses with a big improvement in the last race are overbet.

Forecast odds < 41. These horses have an AE less than 1 and a very low strike rate so it is sensible to exclude these.

This leaves 7,320 winners from 52,623 runners with a LSP of 2,096 points. The AE is 1.061.

I have then applied these criteria to the test data, 2023. The results were 2,189 winners from 16,519 runners. The AE was 1.042 and the ROI to proportional stakes was 2.3%. The only slight disappointment was the LSP at - 14 points.

I have reanalysed the data for 2019 - 2023 to use for 2024 races. The top 2 criteria are the same along with the distance stat. There are some changes elsewhere with the USA statistic no longer being key. New criteria are

Draw < 15.

Forecast odds > 5/4 and < 22/1.

I will now start betting these selections (to small stakes).

[Zilzalian]

Have you looked at royal ascots over these years and  if so do they rate/return the same or similar for all your criteria?

[MCLARKE]

1 hour ago, Zilzalian said:

Have you looked at royal ascots over these years and  if so do they rate/return the same or similar for all your criteria?

Good point. I'll have a look. The current analysis just takes data for one season so doesn't carry over races from a previous season, that is something else I need to do.

[MCLARKE]

Today's top rated

Nottingham 3.45

GREAT BLASKET 28/10 LADBROKES

[MCLARKE]

First bets for the "system"

[MCLARKE]

1.58 point on the opening day.

[MCLARKE]

Current loss - 8.02

[MCLARKE]

On 6/12/2024 at 8:09 PM, Zilzalian said:

Have you looked at royal ascots over these years and  if so do they rate/return the same or similar for all your criteria?

The results for 2023 were very bad with an AE of 0.76 and a ROI of -27% with proportional staking.

The returns from handicaps were good with 9 wins from 171 runs and a profit of 197 points. The AE was 1.18.

I'll have a look at a Royal Ascot only system and see what the returns from that are.

[Zilzalian]

29 minutes ago, MCLARKE said:

The results for 2023 were very bad with an AE of 0.76 and a ROI of -27% with proportional staking.

The returns from handicaps were good with 9 wins from 171 runs and a profit of 197 points. The AE was 1.18.

I'll have a look at a Royal Ascot only system and see what the returns from that are.

Be careful dont go back to before they relaid the track. In my opinion Royal Ascot is a good test for the accuracy of your speed figures, like Cheltenham they are all triers.

Edited June 16 by Zilzalian
omission

[MCLARKE]

I think the current system may be a bit too broadbrush.

It is always a bit of a compromise, an overall system will have a very large data size compared to a more specific analysis which will have a much smaller data size.

[MCLARKE]

Current loss - 3.81

[MCLARKE]

On 5/23/2024 at 7:46 PM, MCLARKE said:

Regional is due to run in the 2.30 at Curragh on Saturday, available at 6/1and also entered in 2 sprints at Royal Ascot

Runs in the 3.45 tomorrow at Ascot.

9/2 in a field of 17, not sure whether that is a value price.

[harry_rag]

I understand the desire to have a more or less automated system as I tend to prefer that approach myself but I can’t shake the feeling that you’re trying to systemise something that lends itself to having a degree of subjectivity in the final decision. For instance could you end up with a top rated 5/1 selection with a 33/1 shot rated 1 point lower being ignored?

If we were looking at that based on our “experiment” approach we’d probably have leant towards the 33/1 shot being the better value.

I did have a stab at creating a value rating. I’ll remind myself how that worked and see if it’s something that could potentially be applied to your system. Maybe backing the best “value” horse in the race would produce better returns.

[richard-westwood]

52 minutes ago, MCLARKE said:

Runs in the 3.45 tomorrow at Ascot.

9/2 in a field of 17, not sure whether that is a value price.

I have regional top rAted ....if you look at his form he tends to need the first run then runs 10 PTS faster on his 2nd ....so I think now is the time to catch him 

[MCLARKE]

1 hour ago, harry_rag said:

I understand the desire to have a more or less automated system as I tend to prefer that approach myself but I can’t shake the feeling that you’re trying to systemise something that lends itself to having a degree of subjectivity in the final decision. For instance could you end up with a top rated 5/1 selection with a 33/1 shot rated 1 point lower being ignored?

If we were looking at that based on our “experiment” approach we’d probably have leant towards the 33/1 shot being the better value.

I did have a stab at creating a value rating. I’ll remind myself how that worked and see if it’s something that could potentially be applied to your system. Maybe backing the best “value” horse in the race would produce better returns.

There can be several horses in a race. Based on last years results horses in the forecast range 7/1 - 8/1 have provided the best results based on the p-score.

I agree though that something doesn't quite feel right at the moment, I think because there are so many selections. On my to do list are :-

Split between handicap / non-handicap races

Adjust for weight

Systems based on horses with bigger speed figures

Create a rating for each horse using all variables, this should ensure that just one horse is clear rated in each race

[Zilzalian]

One of the biggest problems i have encountered over the years is the sheer amount of horses that run on the flat...

General Estimate

The number of flat horses running in a given year can vary significantly. Here are some rough estimates based on available data:

• Great Britain: Around 14,000 - 16,000 individual flat horses typically run in a year.
• Ireland: Approximately 8,000 - 10,000 flat horses run annually.
• France: Around 10,000 - 12,000 flat horses participate in races each year.

When you consider this then you must conclude that you need to specialise somewhat which means you have to ditch something and the obvious answer is handicaps and low grade races that by logic are the most inconsistent beasts.

Now contrary to this one might want to just concentrate on the all weather or specific courses or ground conditions etc etc.

In my experience it pays to concentrate on the top tier ie group races. lets be honest here there is more than enough of them and they are the most consistent.

Estimates and Typical Data:

While exact numbers for a specific year would require accessing these reports, here are general insights:

• Britain: Typically, around 1,000 to 1,200 horses might participate in Pattern races annually.
• Ireland: Approximately 600 to 800 horses could be involved in these races each year.
• France: Around 800 to 1,000 horses are expected to run in Pattern races annually

Edited June 17 by Zilzalian
update

[MCLARKE]

A good day yesterday with MOSTAR DREAMS winning at 13.68 from the 2 selections

Current profit 7.62 points

[MCLARKE]

Current profit 1.87 points

[MCLARKE]

Current profit 2.87 points

[MCLARKE]

Current loss -2.23 points

[Zilzalian]

1 hour ago, MCLARKE said:

Current loss -2.23 points

that looks better than yesterday to my eyes. although i have tawfeeg close top with mountain breeze and fairy godmother.

Edited Friday at 01:27 PM by Zilzalian

[MCLARKE]

Saved by Pilgrim in the last

Simmering ran well but no match for the winner

[MCLARKE]

Current profit + 9.17 points

[MCLARKE]

I escaped unscathed from my first festival with an encouraging profit of 16.42 points.

Particularly pleased with the 3.45 yesterday where my 4 selections filled the first 4 places. The SF paid £193, the Tricast £1,065, perhaps I should take @Zilzalian's advice and try this bet in the futute.

Current profit + 24.04 points.

 

[Zilzalian]

2 hours ago, MCLARKE said:

I escaped unscathed from my first festival with an encouraging profit of 16.42 points.

Particularly pleased with the 3.45 yesterday where my 4 selections filled the first 4 places. The SF paid £193, the Tricast £1,065, perhaps I should take @Zilzalian's advice and try this bet in the futute.

Current profit + 24.04 points.

 

Yep you do the work you use the product but only where justified and that's the criteria you have to work out, personally i do 50p forecasts an 25p tricasts irrespective of my single but not if the prices are to short. It's a discussion. 

[Villa Chris]

Skybet have also stopped the forecast/tricast 10% bonus on top 

[LeMale]

4 hours ago, MCLARKE said:

I escaped unscathed from my first festival with an encouraging profit of 16.42 points.

Particularly pleased with the 3.45 yesterday where my 4 selections filled the first 4 places. The SF paid £193, the Tricast £1,065, perhaps I should take @Zilzalian's advice and try this bet in the futute.

Current profit + 24.04 points.

 

He's got me into them. I had a huge payout on my 80-1 nap comp winner with the 80-1 2nd. (which i can't find, too far back i think) And this one. All good fun for small stakes.

[LeMale]

Just now, LeMale said:

He's got me into them. I had a huge payout on my 80-1 nap comp winner with the 80-1 2nd. (which i can't find, too far back i think) And this one. All good fun for small stakes.

[LeMale]

Just now, LeMale said:

Chopped the end off, might need a new nick-name. But as @Zilzalian says, has to be done with bigger priced runners.