AustinDillon75
Colt
May see him at Laytown lol
Approaching speed figures
- Thread starter AustinDillon75
- Start date
AustinDillon75
Colt
I'm repeatedly going back and forth calibrating track performances, these are the top rated horses for Kempton AW from 2013-2021
AustinDillon75
Colt
Southwell again, 2013-present. Yes the track is getting ripped up but I still want to know how horses have performed on it.
I looked at the performances of Ians Memory at the track who ran a 106. Probably some proof that its difficult to sustain a 30lb improvement in performance, albeit, they dropped him in trip for his 22 December 2015 vistory, which rates as the best performance at the track over 8 years in my ratings.
He beat Air of York by a nose, who managed a 93 and failed to win again in 22 races. Ho hum.
I looked at the performances of Ians Memory at the track who ran a 106. Probably some proof that its difficult to sustain a 30lb improvement in performance, albeit, they dropped him in trip for his 22 December 2015 vistory, which rates as the best performance at the track over 8 years in my ratings.
He beat Air of York by a nose, who managed a 93 and failed to win again in 22 races. Ho hum.
In terms of which races to put a squiggle against, I have incorporated a Rating Next Time Out feature, which can sometimes come in handy. It's just an average of comparison between consecutive ratings (ignoring any 0 ratings for falls/Pulled up etc).
So for example in this Uttoxeter race, Next Time Out they have all improved on their figure considerably. It can suggest either a slow race or a problem with underrating my figures. Comparing this with the other races on the same card can help identify which one it is.
I just use it as a guide/indicator more than anything. I'm working out the average on the total number of runners (8) rather than the number of ratings (5), basically because the formulas were easier
I think it's fair to say its not a direct correlation too, there's plenty of other variables so I don't want it to be to accurate either.
In this case Dylanseoghan is the only one that hasn't made a reappearance. When he does I'll probably use a rating of 77 as a base rather than the 57 he recorded on the day.
This is far less mathematical than what others on this thread are doing, there'll be some that will consider this approach as close to blasphemy, but I'm just posting it in the spirit of the forum of ideas and thoughts
So for example in this Uttoxeter race, Next Time Out they have all improved on their figure considerably. It can suggest either a slow race or a problem with underrating my figures. Comparing this with the other races on the same card can help identify which one it is.
I just use it as a guide/indicator more than anything. I'm working out the average on the total number of runners (8) rather than the number of ratings (5), basically because the formulas were easier
I think it's fair to say its not a direct correlation too, there's plenty of other variables so I don't want it to be to accurate either.In this case Dylanseoghan is the only one that hasn't made a reappearance. When he does I'll probably use a rating of 77 as a base rather than the 57 he recorded on the day.
| RaceID | RaceTime | Meeting | Date | Place | Horse | Weight Lbs | SF | Days Since | Recency | NTO | RNTO |
12959 | 04:15 | Uttoxeter | 01-Apr-21 | 1 | Chase The Wind | 165 | 57 | 57 | 2 | +32 | 20.00 |
12959 | 04:15 | Uttoxeter | 01-Apr-21 | 4 | Stay Out Of Court | 166 | 51 | 57 | 3 | +25 | 20.00 |
12959 | 04:15 | Uttoxeter | 01-Apr-21 | 3 | Unai | 161 | 51 | 57 | 2 | +42 | 20.00 |
12959 | 04:15 | Uttoxeter | 01-Apr-21 | 2 | Taboo | 140 | 32 | 57 | 2 | +20 | 20.00 |
12959 | 04:15 | Uttoxeter | 01-Apr-21 | 6 | Dylanseoghan | 157 | 28 | 57 | 1 | 20.00 | |
12959 | 04:15 | Uttoxeter | 01-Apr-21 | 5 | High Counsel | 149 | 28 | 57 | 4 | +41 | 20.00 |
12959 | 04:15 | Uttoxeter | 01-Apr-21 | 0 | L'Es Fremantle | 140 | 0 | 57 | 3 | 20.00 | |
12959 | 04:15 | Uttoxeter | 01-Apr-21 | 7 | Shanroe Smooch | 140 | 15 | 57 | 2 | 20.00 |
This is far less mathematical than what others on this thread are doing, there'll be some that will consider this approach as close to blasphemy, but I'm just posting it in the spirit of the forum of ideas and thoughts
AustinDillon75
Colt
If you ask me, those ideas are really good.In terms of which races to put a squiggle against, I have incorporated a Rating Next Time Out feature, which can sometimes come in handy. It's just an average of comparison between consecutive ratings (ignoring any 0 ratings for falls/Pulled up etc).
So for example in this Uttoxeter race, Next Time Out they have all improved on their figure considerably. It can suggest either a slow race or a problem with underrating my figures. Comparing this with the other races on the same card can help identify which one it is.
I just use it as a guide/indicator more than anything. I'm working out the average on the total number of runners (8) rather than the number of ratings (5), basically because the formulas were easier
In this case Dylanseoghan is the only one that hasn't made a reappearance. When he does I'll probably use a rating of 77 as a base rather than the 57 he recorded on the day.
RaceID RaceTime Meeting Date Place Horse Weight Lbs SF Days Since Recency NTO RNTO Uttoxeter 01-Apr-21 Chase The Wind Uttoxeter 01-Apr-21 Stay Out Of Court Uttoxeter 01-Apr-21 Unai Uttoxeter 01-Apr-21 Taboo Uttoxeter 01-Apr-21 Dylanseoghan Uttoxeter 01-Apr-21 High Counsel Uttoxeter 01-Apr-21 L'Es Fremantle Uttoxeter 01-Apr-21 Shanroe Smooch
This is far less mathematical than what others on this thread are doing, there'll be some that will consider this approach as close to blasphemy, but I'm just posting it in the spirit of the forum of ideas and thoughts
In previous sheets I did I'd include a column for each of the next two speed figures achieved, the issue was the excel lookups just killed the sheet and they kept freezing, so I thought it better to invest strongly in being happy with your base in the first instance so I developed ideas based on ORs, reductions for age and a formula to rate unrated horses via HRB.
One possible issue with your idea is that for first time out 2yos they are bound to achieve better figures as they mature, and you could find yourself increasing the base figure for the wrong reasons. This is especially as they will either improve at varied rates, or in some cases, just offer no substance as a viable racehorse. You then get into those dreaded WFA calculations that are a matter of preference but which on balance I choose to avoid. I do put a WFA figure next to the rating as it offers some idea of what they could achieve on maturing.
I would also observe no criticism of not being highly mathematical, there are some really simple ideas that just work. I deduct 1lb off published ORs for every two months up to the 4yo in April, so the 2yo in April rated 67 would have a base rating of 55 on my sheets. It just feels right for reasons I can't explain.
AustinDillon75
Colt
I've been getting my figures fine tuned and at a stage where I'm as happy with them as its possible to now be, there's still scope to modify and amend, but I now have a base of neary 100,000 runs rated since 2018 and can call on other ratings pre that date if necessary. Its time to start using the ratings to help inform a race.
The attached is an example of a race from 22 May at Wolverhampton. To explain it a bit, each horse is given a suggested "master" rating, with greater weight given to how recent the race was. This is under each horse's name, and a weight adjusted version is provided underneath with the race summary at the top of the sheet. At present it does not add weight to performances at a track and distance, which could be something for the future. The figures below will probably come over as somewhat selective, but its an example.
Just because a horse comes out top rated doesn't mean it wins - the idea is to take the top rated and see if it is credible based on today's conditions and opposition. Under each horse, the sheet automatically highlights races at the same track and anything within 15% either side of today's distance is also highlighted. I haven't picked the 15% for any reason, its just a gut feeling.
GA was my own calculated going allowance for that day, GA1 is usually a minus figure but the nearer to 0, the more likely the performance is a true reflection of its ability, GA2 is a plus figure and the bigger that figure, the more it may be capable of (this figure is added to the final column).
Where the speed column is highlighted in yellow, its in the best 20% of all performances in the sheet, so if recent runs are highlighted it can quickly be noted of interest. Where the RTG column is highlighted its because there is a difference between its speed rating and a suggestion of its possible capability. So Raabeh's run in November scored a 42, but could been seen as good enough to rate at 52, in other words its possible that it has figured in a falsely run race because the going allowance doesn't really explain the poor figure.
On this occasion, the favourite was Mo Henry but on my various figures there's nothing special to report, but its easy to note that he's been running over 7/8 furlongs and is being dropped in trip today. Someone somewhere presumably thought that would be the key to him and he was backed into an SP of 11/4.
Hard Solution is some way clear on LTO figures, and in his previous is a 70 over CD, the argument against is he's coming off something of a break and may be in need of the run. The truth for today could have been seen as somewhere between his master rating and his last run, but he has to at least enter considerations at 11/1 for a repeat of that 70.
Captain Ryan has performed steadily and consistently in three runs over CD, and on my algorithm has come top overall, and also has the second best LTO performance. Its becomes apparent that there are a few of these who are a bit out of form so at 16/1 you could argue he's worth a flyer.
Raabeh was second on the weight adjusted list as well. So there are probably lots of ways of using these figures. I've only done this one date so far though. I'm posting the other races for complete transparency, not the 1330 or 1500 as there was so little form to report.
In the 1400 Beat The Breeze had the best last time out figure and returned at 15/2. Top rated also won the final race as well. I just need to grab two or three weeks of results to bring the ratings up to date then we can maybe begin to start using these.
The attached is an example of a race from 22 May at Wolverhampton. To explain it a bit, each horse is given a suggested "master" rating, with greater weight given to how recent the race was. This is under each horse's name, and a weight adjusted version is provided underneath with the race summary at the top of the sheet. At present it does not add weight to performances at a track and distance, which could be something for the future. The figures below will probably come over as somewhat selective, but its an example.
Just because a horse comes out top rated doesn't mean it wins - the idea is to take the top rated and see if it is credible based on today's conditions and opposition. Under each horse, the sheet automatically highlights races at the same track and anything within 15% either side of today's distance is also highlighted. I haven't picked the 15% for any reason, its just a gut feeling.
GA was my own calculated going allowance for that day, GA1 is usually a minus figure but the nearer to 0, the more likely the performance is a true reflection of its ability, GA2 is a plus figure and the bigger that figure, the more it may be capable of (this figure is added to the final column).
Where the speed column is highlighted in yellow, its in the best 20% of all performances in the sheet, so if recent runs are highlighted it can quickly be noted of interest. Where the RTG column is highlighted its because there is a difference between its speed rating and a suggestion of its possible capability. So Raabeh's run in November scored a 42, but could been seen as good enough to rate at 52, in other words its possible that it has figured in a falsely run race because the going allowance doesn't really explain the poor figure.
On this occasion, the favourite was Mo Henry but on my various figures there's nothing special to report, but its easy to note that he's been running over 7/8 furlongs and is being dropped in trip today. Someone somewhere presumably thought that would be the key to him and he was backed into an SP of 11/4.
Hard Solution is some way clear on LTO figures, and in his previous is a 70 over CD, the argument against is he's coming off something of a break and may be in need of the run. The truth for today could have been seen as somewhere between his master rating and his last run, but he has to at least enter considerations at 11/1 for a repeat of that 70.
Captain Ryan has performed steadily and consistently in three runs over CD, and on my algorithm has come top overall, and also has the second best LTO performance. Its becomes apparent that there are a few of these who are a bit out of form so at 16/1 you could argue he's worth a flyer.
Raabeh was second on the weight adjusted list as well. So there are probably lots of ways of using these figures. I've only done this one date so far though. I'm posting the other races for complete transparency, not the 1330 or 1500 as there was so little form to report.
In the 1400 Beat The Breeze had the best last time out figure and returned at 15/2. Top rated also won the final race as well. I just need to grab two or three weeks of results to bring the ratings up to date then we can maybe begin to start using these.
AustinDillon75
Colt
The figures for the 25 May - 2 meetings. You could have took Shamshon in the last at Lingfield with minimal confidence and the top rated in the first and last at Newcastle both came in. The highlighting has gone a bit wrong in some columns but this is all still trial and error. Whatever use they end up being, its certainly fun producing them.
AustinDillon75
Colt
Here's the Wolverhampton card produced for 26 May. I've done a couple of comments under each race, save for the last, where I comment here. As a starting point, its wide open to the point where amazingly in a 9 horse race four co-favourites were sent off at 5/1.
The sheet points to Yuften as the top rated, so the way I would work is to examine his performances more closely. He has good numbers in his history, but its immediately clear its not at this course and distance and his only try here was a poor last. He isn't one of the co-favs. You could probably discount him.
If I was backing in this race I'd probably not have seperated International Law over the eventual winner, Whatwouldyouknow, but its a tough race to get involved in. One thing I have noticed is that the first five home all have a performance on the board which was one of the top 20% recorded in its last 10 races (SPD column), the last four home could not manage any speed rating highlighted in yellow. Its also noticeable that over CD, the winner had two of the best three performances, a 71 and 68 (International Law producing a 70 LTO).
Ultimately the numbers are there to help guide. It's going to be fun doing them for real. I will try and put a card up in advance of the next AW meeting as all the ratings are up to date and I've now completed the course calibrations too, so I'm pretty content. I'd also appreciate any observations, and questions about the PDF card, and what columns may improve it further (thinking draw number possibly which would be easy to do, inclusion of my HRB Master rating to compare my figures alongside, and confirmation of the 20% figure (which would have been 65 for the last race).
The sheet points to Yuften as the top rated, so the way I would work is to examine his performances more closely. He has good numbers in his history, but its immediately clear its not at this course and distance and his only try here was a poor last. He isn't one of the co-favs. You could probably discount him.
If I was backing in this race I'd probably not have seperated International Law over the eventual winner, Whatwouldyouknow, but its a tough race to get involved in. One thing I have noticed is that the first five home all have a performance on the board which was one of the top 20% recorded in its last 10 races (SPD column), the last four home could not manage any speed rating highlighted in yellow. Its also noticeable that over CD, the winner had two of the best three performances, a 71 and 68 (International Law producing a 70 LTO).
Ultimately the numbers are there to help guide. It's going to be fun doing them for real. I will try and put a card up in advance of the next AW meeting as all the ratings are up to date and I've now completed the course calibrations too, so I'm pretty content. I'd also appreciate any observations, and questions about the PDF card, and what columns may improve it further (thinking draw number possibly which would be easy to do, inclusion of my HRB Master rating to compare my figures alongside, and confirmation of the 20% figure (which would have been 65 for the last race).
AustinDillon75
Colt
Chelmsford 31 May 2021. No top rated winners but a couple going close beated a head and short head respectively, and in fairness, the winner of the last Expert Opinion was 9/1 SP and comes out of the ratings quite well.
AustinDillon75
Colt
First go at a card, these are some suggested figures for today's meeting at Lingfield.
Great work, nice presentation, well done with the 16/1 winner
AustinDillon75 , hopefully more to come for you after the hard work.
AustinDillon75
Colt
As mentioned, for now these are just for interest, no money behind them and the lists of weight/non-weight adjusted are simply to give food for thought. Would anyone really have backed Bailarico on those figures, questionable I guess.Great work, nice presentation, well done with the 16/1 winner
I'll put up speed figures for that meeting tomorrow, and put some more sheets up in advance of Chelmsford City on Wednesday.
AustinDillon75
Colt
Lingfield performances from yesterday. Passion and Glory ran a very impressive race.
AustinDillon75
Colt
Ratings for Chelmsford City tomorrow. Apologies as there are still one or two formatting issues but the figures are as they should be. I've now included my own personal HRB master ratings as well, so they can be put against my speed figures for comparison.
EDIT - had to revise, all kinds of formatting errors that I cannot understand for the life of me.
EDIT - had to revise, all kinds of formatting errors that I cannot understand for the life of me.
Last edited:
AustinDillon75
Colt
Not a great deal to shout about there, we were denied by a late head bob with Bright Start in the 1953 (got the first two home) and we got the winner of the last with Aramis Grey. Got the first three home there as well which would have been a decent boxed trifecta. I doubt I'll have the time to produce cards for tomorrow's AW but we'll see. Its a busy days work and I am going into hospital later in the day too.
markfinn
Sire
Good luck with the later appoint
AustinDillon75
Colt
How my figures rated them yesterday.
AustinDillon75
Colt
Cards and ratings for today.
Stevet1859
Yearling
Extraordinarily good results at Chelmsford, sadly didn't back it. Less spectacular, but Ryal Dynasty looked very solid "across all measures"in the 2:10 and duely won. Eagle swooped in for good measure too.....thank you for posting. Have read the threads and process you went through, and looks like all that effort paying off. (Hope all well with appointment too.....this must be good medicine.) All best