Speed Figure calculation and usage

[EC1]

Was wondering what you made of it EC. Given the fact it was a handicap, and they've not really finished that well strung out, and that the field were 5 from 24 in handicaps coming into the race, how much do you want to stick your neck on the block and suggest the first 6 are all well ahead of their marks.

Looking at the first few - some did have higher OHR's at one point - a couple could be improving..it wouldn't initially strike me as a race to follow..but it makes the other races very slow if its the only true run race on the card.

its a very fast time comparitively - even compared to the other 7f race.

I'm struggling to use Pru's formula to see which were true run- i can't see the furlong markers in the straight..if I use the path that crosses the track its might be too far out..3.5f ish. I might have a go at it to see if it does give some more info re the times

 

[J Alfred Prufrock]

If you treat the end of the path as being 3.33f from home, you get the following for the leaders at the sectional and the leaders (i.e. winner) at the line:

Bushman 106.5%
Tartan Gigha 102.2%
Fame And Glory 106.0%
Fiery Lad 103.4%
Snow Fairy 108.7%
Shakespearean 104.1%
Cansili Star 99.2%

I have sectionals from that path going back to the 1960s for 1½m races.

 

[EC1]

If you treat the end of the path as being 3.33f from home, you get the following for the leaders at the sectional and the leaders (i.e. winner) at the line:

Bushman 106.5%
Tartan Gigha 102.2%
Fame And Glory 106.0%
Fiery Lad 103.4%
Snow Fairy 108.7%
Shakespearean 104.1%
Cansili Star 99.2%

I have sectionals from that path going back to the 1960s for 1½m races.

many thanks Pru - just been trying to do it off RUK - not got full 3.33f for all races or the thing sticks on one race

just looking at the race I would have thought F&G's race was truly run..but 106% is looking slow?

using final times to make figures i suspect the 12f times don't fit with the rest..this has been the case other years.

looking at the %'s would suggest that Cansili's is the only true run race there..Tartans' not bad

at what % would you say a race is just a bit slow early? ..102?..103?

then very slow 105+ ?

 

[Harbinger]

Looks a fast race alright. Using the RP Standard Times with a basic adjustment for the extra distance caused by the dolling out, and on the assumption that Cansili Star ran to a form rating of about 99, I would expect him to run 6lbs slower than standard. Instead, he's run about 11lbs faster than standard; so 17lbs fast overall - comfortably the fastest race of the day.

 

[J Alfred Prufrock]

I have the following timefigures:

Bushman 94
Tartan Gigha 104
Fame And Glory 117
Fiery Lad 102
Snow Fairy 106
Shakespearean 96
Cansili Star 101

And that final-race time is indeed notably good for the calibre of horse. I would have expected a 97 maximum.

The individual sectionals of Derby winners over the years suggest something around 105.4% is optimum for the trip. It will be different at different distances and very different at different distances on a track with a configuration like Epsom's. The 3.33f is an estimate, of course.

Interestingly, Snow Fairy's closing sectional (as far as you can engineer it from visual margins and the leaders' sectionals) is as fast as any filly in the race since the distance was changed to 12f 10 yds in 1992: Alexandrova clocked a very similar closing time. Snow Fairy was showing pace at least as much as stamina to come from where she did.

 

[EC1]

Looks a fast race alright. Using the RP Standard Times with a basic adjustment for the extra distance caused by the dolling out, and on the assumption that Cansili Star ran to a form rating of about 99, I would expect him to run 6lbs slower than standard. Instead, he's run about 11lbs faster than standard; so 17lbs faster than expected - comfortably the fastest race of the day.

the %'s are showing the other races to be pretty slow though bar Tartan..Tartan appears..just on final times..to have run a par race..which makes Cansili's very fast..102% for that race could make it a few lbs slow..but that still makes Cansili very fast race

i'm surprised that F&G's race shows a slow %..it doesn't look slow by eye..the front 3 seemed to go a solid pace

Pru..is there a correlation between % and lbs slow?..say 106% = 10lb slow..or something like that

 

[EC1]

I have the following timefigures:

Bushman 94
Tartan Gigha 104
Fame And Glory 117
Fiery Lad 102
Snow Fairy 106
Shakespearean 96
Cansili Star 101

And that final-race time is indeed notably good for the calibre of horse. I would have expected a 97 maximum.

The individual sectionals of Derby winners over the years suggest something around 105.4% is optimum for the trip. It will be different at different distances and very different at different distances on a track with a configuration like Epsom's. The 3.33f is an estimate, of course.

Interestingly, Snow Fairy's closing sectional (as far as you can engineer it from visual margins and the leaders' sectionals) is as fast as any filly in the race since the distance was changed to 12f 10 yds in 1992: Alexandrova clocked a very similar closing time. Snow Fairy was showing pace at least as much as stamina to come from where she did.

so you can't say that 100% is uniform..it varies from track to track?

I would imagine at more traditional tracks its a fair figure?

 

[J Alfred Prufrock]

so you can't say that 100% is uniform..it varies from track to track?

I would imagine at more traditional tracks its a fair figure?

Optimum finishing speeds for final 2f at Kempton (where we actually had electronic sectionals for some time) are between 100.1 and 101.8 for races at up to 8f, depending on the distance of the race. While Kempton is flat, it has bends and two separate courses.

Pru..is there a correlation between % and lbs slow?..say 106% = 10lb slow..or something like that

There is, but it depends on the length of the sectional compared to the length of the race. Square the difference between actual sectional and optimum sectional first and you are along the right tracks.

 

[EC1]

Optimum finishing speeds for final 2f at Kempton (where we actually had electronic sectionals for some time) are between 100.1 and 101.8 for races at up to 8f, depending on the distance of the race. While Kempton is flat, it has bends and two separate courses.



There is, but it depends on the length of the sectional compared to the length of the race. Square the difference between actual sectional and optimum sectional first and you are along the right tracks.

cheers Pru

its all about what the final 2f should be ..compared to verall time..in a true run race at that course..ie Epsom is probably different to most courses in that respect

 

[Desert Orchid]

I've done my time ratings for yesterday. They appear to come to similar conclusions, bearing in mind that I don't do sectionals.

Race 1 - Bushman: time rating 101 (cf OR 109)
Race 2 - Tartan Gigha 110 (100)
Race 3 - Fame And glory 115 (128)
Race 4 - Fiery Lad 107 (105)
Race 5 - Snow Fairy 97 (+15wfa = 112)(107)
Race 6 - Shakespearian 99 (+ 10wfa = 109)(107)
Race 7 - Cansili Star 99 (+ 10wfa = 109)(85)

Had that last race not been there, I'd have said the figures look especially good in terms of pointing to a card of true-run races. If it's the case then the last race should throw up a hatful of future winners. On the other hand, it might just be an indication that we should treat race times with a degree of caution.

 

[EC1]

I've done my time ratings for yesterday. They appear to come to similar conclusions, bearing in mind that I don't do sectionals.

Race 1 - Bushman: time rating 101 (cf OR 109)
Race 2 - Tartan Gigha 110 (100)
Race 3 - Fame And glory 115 (128)
Race 4 - Fiery Lad 107 (105)
Race 5 - Snow Fairy 97 (+15wfa = 112)(107)
Race 6 - Shakespearian 99 (+ 10wfa = 109)(107)
Race 7 - Cansili Star 99 (+ 10wfa = 109)(85)

Had that last race not been there, I'd have said the figures look especially good in terms of pointing to a card of true-run races. If it's the case then the last race should throw up a hatful of future winners. On the other hand, it might just be an indication that we should treat race times with a degree of caution.

I think using final times and calculating the allowance worked out pretty well DO..even without sectionals..and we have all concluded the last race was very fast. The key race was Tartan Gigha's..the sectionals will tell if that was run evenly..everything points to it being so.

When i did the sectional analysis for the Derby last year I calculated the % time that should be spent up to that path over the 12f distance here to signify even pace...is 73.50%...I know Pru will also have done this..his formula is just another way of showing this really.

an even pace would point to a horse spending 73.50% of its time in that first slice.

Yesterday F&G spent 73.70% of his time in that first slice..and he was right on the pace....which would suggest his final time yesterday wasn't too far off the optimum he is capable of achieving. That race was pretty evenly run..which points to the 12f allowance yesterday being different from the other distances..I doubt very much that F&G has run much below his very best speed figure..but we all get that race about 11-13lb slow.

I've noticed this before at the Derby Distance...the allowancediffers as do 5f races here I don't think the 12f times yesterday should be connected to the others..F&G's race was not slow run..so there is no way that race is a stone slow..but that doesn't affect how we view the 3yo handicap anyway.

If Pru could post up what his expected % figure for the 8.5f distance is..we can then know how truly run Tartan Gigha's race was..is 102% near the expected for that distance basically?....I'll guess its near the optimum % for a true run race. That is where the sectionals can really confirm if your allowance is correct or not.

what would be the optimal % for the 7f trip Pru?

just re wfa DO - i still can't see giving that many lbs look right to me..wfa is a handicapping tool and i don't see how it fits in with time analysis in that way. For instance..I add just 5 speed points/lbs at this time of year to 3yo...becasue if you check..over a decent sized sample..3yo's on the clock are only 0.35 seconds per mile shy of maturity at this time of year.

wfa re speed figures is another talking point i think

 

[Desert Orchid]

Yes, while I include wfa I don't allow it to dominate my thinking, but I wouldn't want to go down the road of making an alternative allowance. It's the same with 2yos. I still think the reason some juveniles put up very fast times is that they're just more mature. It would stand to reason a similar argument could be made regarding 3yos.

The other interesting aspect is the difference between F&G's time and Snow Fairy's. If F&G's time was on a par with its form, you'd have to say Snow Fairy is its equal at weight for sex. I find that hard to believe.

I think the really top class horses don't necessarily run really fast times because they very seldom need to or very seldom have races run conducively.

 

[EC1]

Yes, while I include wfa I don't allow it to dominate my thinking, but I wouldn't want to go down the road of making an alternative allowance. It's the same with 2yos. I still think the reason some juveniles put up very fast times is that they're just more mature. It would stand to reason a similar argument could be made regarding 3yos.

its just the adding of such a lot of lbs DO that is out of kilter imo.

at this time of year for instance..I get 2yo's being 1.6 seconds per mile shy..and 3yo's 0.35 shy of maturity.

but the wfa scale would mean adding far more lbs back than that to the speed figure..it just doesn't correlate imo..and i think adding the actual wfa scale gives 2yo's and 3yo's over inflated speed figures.

wfa poundage is a handicapping tool imo..speed figure wfa is different game imo

 

[EC1]

you say Snow Fairy wouldn't look right if F&G was a true run race..the reason for that is that adding wfa as you do is overinflating Snow Fairy...you are adding 10lb more than the clock says is possible..it should be 5lb wfa speed lbs.

I can't see that F&G's race wasn't that evenly run..73.7% is about right for even pace..just a whisker slow. Using my wfa method in this instance..and assuming F&G has run to a 123-125 speed figure..just a bit less than OHR..would put Snow Fairy on a wfa included rating on my way of dong wfa of 105-107..which imo is about right.

Snow Fairy's race sectional to the path was 74.5%..which means she wouldn't be capable of running a speed figure in line with her OHR in a race run at that pace..but it wouldn't trash the speed figure totally because thats not a pedestrian pace..but it is slower than even pace.

the sectionals are a vital key to all this..far more insight than bare times

 

[Desert Orchid]

at this time of year for instance..I get 2yo's being 1.6 seconds per mile shy..

One difficulty I have with this kind of approach is that 2yos aren't asked to run a mile at this stage. They don't run beyond 6f at this time and if they did they'd run a fair bit faster than they could at a mile. Even then, 1.6s at 5f is about 29lbs. The wfa table wouldn't be far off that.

 

[EC1]

One difficulty I have with this kind of approach is that 2yos aren't asked to run a mile at this stage. They don't run beyond 6f at this time and if they did they'd run a fair bit faster than they could at a mile. Even then, 1.6s at 5f is about 29lbs. The wfa table wouldn't be far off that.

its not 1.6 at 5f though..its 1.6 per mile...at 5f its 1 second..derived from [1.6 X 0.625]..5f is 0.625 th of a mile.

wfa in June at 5f is 32lbs for a 2yo..on the clock though in reality its only 1 second..hence adding 32lbs is overrating a 2yo on the clock by way too much.

per mile times don't involve horses having to run a mile.

you will find that per mile calcs..work out exactly the same as treating each distance differently re poundage..it just makes calcs so much easier

 

[Desert Orchid]

But they run at different speeds at different distances.

 

[EC1]

But they run at different speeds at different distances.

i'm not seeing how per mile calcs are flawed..you will have to enlighten me DO

try this - compare 2yo races where they run over the same c/d as an older race..equate them class wise ..and by the month ..and you will find i'm not far out.

according to the wfa handicapping scale a 2yo of same abilty over 5 furlongs as an older horse would be 2 seconds slower..its a full second wrong

you are adding lbs the horse hasn't earned or ever likely to earn.

wfa scale is for handicapping..speed figures has to treat it differently

to prove its wrong - check your top 30 speed figures over the last 5 years..they will in the main be 2yo & 3yo's..there is a reason for that

 

[J Alfred Prufrock]

The sectional %s above were for the leaders and can be used as some sort of indication of the pace of the race (if such a thing can be quantified meaningfully from the leaders alone).

The sectional %s of the individual horses are what affects how fast those individual horses are capable of getting from the start of a race to the finish. I make them:

Bushman 107.2%
Tartan Gigha 104.2%
Fame And Glory 106.0%
Fiery Lad 106.6%
Snow Fairy 112.7% (!)
Shakespearean 104.1%
Cansili Star 100.5%

This is based on ONE SECTIONAL, however, and while it is a significant sectional it does not tell the whole story of a race, especially over longer distances. I make the optimum finishing speed for 1½m races at Epsom - based on 1½m races there over the years that have been run in times close to the winners' abilities - 104.3%.

I suspect that if you could break yesterday's Coronation Cup down further you would find that it was run at a sound pace for much of the way but that the pace steadied somewhat mid-race. Fame And Glory went from being 7 lengths clear of the chasing pack to being only a couple clear in the space of a little over two furlongs. If he was going about as fast as he "should" then the others were doing quite a bit too much at that stage, and I doubt they were.

Still, I would not be upgrading the overall time much in the circumstances, so you may have a point about separate allowances being required at the longer trip.

Unfortunately, I dropped my work on sectionals at Epsom, other than the 1½m races, some years back, so I can't help with the optimum 7f sectional. Given the nature of the track at Epsom I would have thought it would be between 102 and 104, suggesting that Cansili Star's overall time could be upgraded a few pounds further. But that's speculation.

 

[J Alfred Prufrock]

i'm not seeing how per mile calcs are flawed..you will have to enlighten me DO

For what it's worth, Bob Wilkins in his book "Bioenergetics and Racehorse Ratings" has the following to say about seconds-per-furlong (or per mile) adjustments compared to pounds-based adjustments:

"...the key result is that for non-standard going, represented by a fixed value of Fg [going coefficient], seconds per furlong is not constant, but increases significantly as the race distance increases...the use of a fixed value of SPF for all distances results in a significant error in the correction required to the standard time...if ratings are to be calculated in pounds this method [a constant weight correction, as in my spreadsheets examples] of correcting for the going is preferable to the seconds per furlong method" (pages 40 and 41)

 

[EC1]

The sectional %s above were for the leaders and can be used as some sort of indication of the pace of the race (if such a thing can be quantified meaningfully from the leaders alone).

The sectional %s of the individual horses are what affects how fast those individual horses are capable of getting from the start of a race to the finish. I make them:

Bushman 107.2%
Tartan Gigha 104.2%
Fame And Glory 106.0%
Fiery Lad 106.6%
Snow Fairy 112.7% (!)
Shakespearean 104.1%
Cansili Star 100.5%

This is based on ONE SECTIONAL, however, and while it is a significant sectional it does not tell the whole story of a race, especially over longer distances. I make the optimum finishing speed for 1½m races at Epsom - based on 1½m races there over the years that have been run in times close to the winners' abilities - 104.3%.

I suspect that if you could break yesterday's Coronation Cup down further you would find that it was run at a sound pace for much of the way but that the pace steadied somewhat mid-race. Fame And Glory went from being 7 lengths clear of the chasing pack to being only a couple clear in the space of a little over two furlongs. If he was going about as fast as he "should" then the others were doing quite a bit too much at that stage, and I doubt they were.

Still, I would not be upgrading the overall time much in the circumstances, so you may have a point about separate allowances being required at the longer trip.

Unfortunately, I dropped my work on sectionals at Epsom, other than the 1½m races, some years back, so I can't help with the optimum 7f sectional. Given the nature of the track at Epsom I would have thought it would be between 102 and 104, suggesting that Cansili Star's overall time could be upgraded a few pounds further. But that's speculation.

Hi Pru

If you check Motivators Derby - his time and Crow Woods are out of line with the rest of the card..in 2002 High Chapparal + Salim Toto times don't correlate to a lesser degree either..also Lammtarra's year..both 12f races out of kilter with rest of card

the nature of each distance at Epsom makes every race thats run is run on a different part of the course isn't it?..chutes for this and that etc


i thought 102-104 would be fair..tartan gigha looks a good marker - the first two in the 3yo handicap certainly look interesting..they did put 2 lengths betwen themselves and the rest

maybe its one of those races where the first 5 will all come out and win

 

[EC1]

For what it's worth, Bob Wilkins in his book "Bioenergetics and Racehorse Ratings" has the following to say about seconds-per-furlong (or per mile) adjustments compared to pounds-based adjustments:

"...the key result is that for non-standard going, represented by a fixed value of Fg [going coefficient], seconds per furlong is not constant, but increases significantly as the race distance increases...the use of a fixed value of SPF for all distances results in a significant error in the correction required to the standard time...if ratings are to be calculated in pounds this method [a constant weight correction, as in my spreadsheets examples] of correcting for the going is preferable to the seconds per furlong method" (pages 40 and 41)

not really grasping that tbh - if races are being slowed down or speeded up - you are going to express that in poundage..well I do anyway..I calculate what figure in poundage a horse earns - and then what it should earn..the difference that goes to working out the allowance is always in poundage ..how i work. Thats why I express the going in poundage as for this card we been looking at

 

[J Alfred Prufrock]

The time of The Derby suggests the going allowance at 1½m does not need to be slower.....well, the time for the winner anyway.....!

I have only Galileo's closing sectional faster in this race, though Nijinsky was close.

 

[J Alfred Prufrock]

Correction: I was forgetting last year's steadily run race which saw Sea The Stars the second-fastest closing sectional behind Galileo. This goes in at number three.

 

[Harbinger]

How much of the race time did the part before the closing sectional count for today, compared to those others you mentioned?