Hong Kong Speed Figures

[Outlander]

Today's results added.

3 newcomers win today, one heavily gambled and runs very wide at last bend and still won comfortably, Ever Luck race 2. The Group 2 as mentioned in previous posts was won by Tomodachi Kokoroe, here is the form for the 7 year old that has exceded everything before in last 2 runs for no obvious reason than addition of blinkers? Any ideas?

Racing in Thursday this week.

View attachment 161463

Hayes said one eyed blinkers transformed him, available at 7/1 skybet 10 MTP finished 3.9, absolutely smashed to pieces most of recent runs, last time another massive gamble. See the odds movements of the right on his scmp page. Time today 2nd fastest ever behind KYR in Sha Tin 1200m I think.

 

[The Hare]

Card for Thursday 8 race all weather meeting at Sha Tin.

 

[La Cressonnière]

SHA TIN AWT 30/10/2025

R1	R2	R3	R4	R5	R6	R7	R8
COURIER MAGIC	LUCKY BLESSING	HE WAS YOU	GALE SAGA	KNOW AT HEART	YODA'S CHOICE	GOOD LUCK BABE	PRECISION GOAL
NIGHT PUROSANGUE	DOUBLE BINGO	JOLLY COMPANION	SUPER SICARIO	HAPPY ACTION	GLORIOUS JOURNEY	FUN N FUN TOGETHER	HARMONY GALAXY
HARRY'S HERO	EXCELLENT DADDY	NEW POWER	KASA PAPA	GLACIATED	SPEEDY TRIDENT	THE REVOLUTION	ARIEL
FOREMOST TEDDY	ORIENTAL SURPRISE	DASH	DASHING DRAGON	FORTUNE WARRIOR	LESLIE	GIANT LEAP	MR ENERGIA
BRAVE OF FRIENDS	NOBLE GO	NO OTHER CHOICE	SUPREME AGILITY	DAILY TROPHY	LUCK IS BACK	FATAL BLOW	MOJAVE DESERT
EURO ROCKS	CHEER FOR SOUTH	HINOKAMI KAGURA	RELIABLE PROFIT	VULCANUS	CALIFORNIA STAR	SUPER JOY N FUN	TURIN MASCOT
GO GO GO	GLORY CLOUD	BEAUTY THUNDER	VIVA TASTE	WINNING MACHINE	FASHION LEGEND	DRAGON FOUR SEAS	YOUNG ACHIEVER
YEE CHEONG RAIDER	LEGEND STAR	IMPENDING LEGACY	SOARING BRONCO	GIDDY UP	MEEPMEEP	WONDER KIT	NEW FOREST
ALL EYES ON ME	HANDSOME BLOND	DR GOLDEN GENERAL	DILBAGH	FORTUNE WHISKEY	YOUTH POWER	KEMPES	NOISY BOY
PERPETUAL POWER	NOTTHESILLYONE	GIMME FIVE	SPECIAL HEDGE	LOVING VIBES	SHANGHAI STYLE	LUCY IN THE SKY	NEVER PETER OUT
SILVER KING	BEAUTY FORMULA	WATCH THIS ONE	SIGHT SUPREME	EVERYONE'S GALLOP	PRESTIGE RICKY	MUST GO	TOURBILLON PRINCE
VIEW ALL THINGS	FORTUNE RACER	MARVEL AND GOLD	YOUNG BRILLIANT	HONG KONG HALL	MAGIC COSMOS	CAMP LAND	KARMA
			MEGA FORCE		CHARMING STEED		CHILL KAKA
					SEA CHARIOT

 

[AustinDillon75]

Thinking out loud here, great post quoted from AustinDillon75 a couple of years ago on this thread.
At the time although I didn’t claim to understand the maths behind it I thought there was something very interesting in the idea he put forward of essentially an efficiency uplift for each horse. Although the i felt the maths and idea were top notch, in practice maybe it was too reliant on the pars and it didn’t pan out when I applied in to my whole database of HK results.

Recently I have created ratings for each section S1NW-S6NW, they initially are no weight ratings little bit like Mordin or Beyer, I take into account final time in my section ratings so they can be directly compared with each other and the overall speed figure SSNW.

Although ARAZI91 has confirmed not mathematically robust , I don’t think they are too bad and naively I thought I could look at them and know which horse could have won and would be helpful.

I know sections can’t be taken in isolation as a guide to the merit of horses performance but if they accounted for overall time in their composition I thought maybe it would give a useful picture.

Now I think even with overall time taken into consideration a section rating can not be used in the same way as an overall time rating.

In the example below race won by Flowing Riches his speed figure is 85.23.
The horse I want to focus on is Dr Golden General , he has finished better than most and even accounting for overall time has a final 400m rating of 86.72.
So my first thought could Dr Golden General have beaten Flowing Riches.
I watched the race and the answer is no, it’s not a fair comparison, he was off the back and finished ok and would never have got close to the winner.

View attachment 161328
View attachment 161329
Was thinking of AustinDillon75 idea of efficiency, could I use his method on my figures, no horse really runs 100% efficiently, they run with varying degrees of efficiency.
I used his formula above and calculated an EFFuplift for each horse to apply to it‘s overall speed figure (SSNW).
Dr Golden General did have the highest efficiency uplift of 4.7125 but that wouldn’t entitle him to get within 3 pts of the winner.
A final efficiency adjusted figure of 82.25 compared with 85.60 for Flowing Riches.

Think this might be the way forward efficiency uplift instead of sectional uplifts, certainly interested me enough to do more research and test with AustinDillon75 idea.

Any thoughts welcome.

Thanks for re-raising this, in the end I simply got fed up and dispirited with it all. The efficiency idea appeared to make no difference to anything, and further analysis concluded that some horses are just incapable of racing efficiently and no matter what upgrades you try and apply to them they will simply never come to pass. Often they will run more efficiently and their actual performance will be utter crap.

If I ever find time again, I might have another proper look at it all, I've developed a few coding skills and am separately trying to build prediction models and handicapping tools through Python, I'll likely pick up the HK racing at some stage again, I guess, but I have about 4 spare minutes a day these days, sadly.

 

[Outlander]

Thanks for re-raising this, in the end I simply got fed up and dispirited with it all. The efficiency idea appeared to make no difference to anything, and further analysis concluded that some horses are just incapable of racing efficiently and no matter what upgrades you try and apply to them they will simply never come to pass. Often they will run more efficiently and their actual performance will be utter crap.

If I ever find time again, I might have another proper look at it all, I've developed a few coding skills and am separately trying to build prediction models and handicapping tools through Python, I'll likely pick up the HK racing at some stage again, I guess, but I have about 4 spare minutes a day these days, sadly.

Good to hear from you again AustinDillon75 . I’m in the process of incorporating the efficiency adjustments to my sectional rating atm, I couldn’t make it work with times or pars, but looks more useful with my figures, will see how they get on when I have completed all my sheets.
Like you say might not help much but decided it’s worth doing to see.
Good luck with your models and coding, you always had good ideas and skills so hopefully will be successful with you prediction models.

 

[AustinDillon75]

Good to hear from you again AustinDillon75 . I’m in the process of incorporating the efficiency adjustments to my sectional rating atm, I couldn’t make it work with times or pars, but looks more useful with my figures, will see how they get on when I have completed all my sheets.
Like you say might not help much but decided it’s worth doing to see.
Good luck with your models and coding, you always had good ideas and skills so hopefully will be successful with you prediction models.

Thank you O Outlander. These other ideas are proving more time consuming and headscratching but the payoff could in the end prove so much greater. If I pick up Python again, and get efficiency figures, it is possible with learning to perhaps put them through Optuna tuning, and it could then tell you which efficiencies work best for each track and distance. You should also most likely then be able to obtain efficiency pars for each horse and it will reveal the "truth" over their ratings - do they need a race to be run a certain way to achieve a good time, or are they better in an evenly run race, and if you can figure how the race is likely to be run, and you have the data and coding skills to make a meaningful forecast, you can then come to a conclusion. Of course, some horses are simply too good for the race shape and pace to be a problem.

I am sure that z-scores can play a part in all this too. There are obviously statistical / coding minds better than me who can develop it further. I will defintely be taking it up again at some point.

Also good luck you guys with your continued endeavours. I hope they are getting their credit across the wider forum because so much work and effort goes into what you all do.

 

[tacker]

In my opinion sectionals shouldn't just be seen as a numbers game, far more thought needs to be given to both the race as a whole and then each individual horse involved.

In saying that some might feel there's little or no point taking note of such factors but then need to explain why connections of horses constantly use such various nuanced reasonings, you might disagree with their opinions but would need to take notice of replays with all that encompases, a lot of work involved.

 

[Outlander]

In my opinion sectionals shouldn't just be seen as a numbers game, far more thought needs to be given to both the race as a whole and then each individual horse involved.

In saying that some might feel there's little or no point taking note of such factors but then need to explain why connections of horses constantly use such various nuanced reasonings, you might disagree with their opinions but would need to take notice of replays with all that encompases, a lot of work involved.

Agree none of this is just a numbers game 100% , but how else can you keep on top of everything even with 2 meetings a week you are not going to be able assess every section or performance unless you have some kind of number, impossible for U.K. racing unless you have a way to evaluate everything efficiently. You might be able to look at a few horses a day and form an opinion without numbers, but likely to miss the ones that matter more.
The theory is that if you have a number for everything then you know where to look first, in practice it still isn’t the golden goose, I’m the first to admit.

 

[tacker]

@Arazi always strikes me as someone who has his head around this and many similar approaches but tells us he doesn’t bet, that could be a sign that he can’t make sectionals work as a stand alone approach but that’s not to say that using them alongside other strategies won’t be of benefit, same might apply with a collective strategy using the full intellectual capacity demonstrated by the members in this thread.

I would be very happy to nick anything that is worthwhile, providing I can understand it.

 

[AustinDillon75]

In my opinion sectionals shouldn't just be seen as a numbers game, far more thought needs to be given to both the race as a whole and then each individual horse involved.

In saying that some might feel there's little or no point taking note of such factors but then need to explain why connections of horses constantly use such various nuanced reasonings, you might disagree with their opinions but would need to take notice of replays with all that encompases, a lot of work involved.

Re your first paragraph, I'd already expressed caution over just using numbers because people like Rowlands just upgrade by assuming a horse can run efficiently, you might also find sectional jockey pars and styles as well that put the performance more down to the jockey riding than the horse itself. There's so many rabbit warrens you can go down but I would have thought if you can accurrately understand a horse's best speed figure you'd want to understand how it achieved it. If it managed it with a 10 second final furlong there's no point assuming it will be 9lb better if it paces more evenly, because it is rarely going to have anything left in its locker for a sprint finish. If it's done similar things off both an even pace and a sprint finish it's tactical versatility probably means you can trust it to repear its performances a bit more. A horse surely can't achieve its best figure if it runs in a manner to which it simply isn't suited.

 

[Outlander]

@Arazi always strikes me as someone who has his head around this and many similar approaches but tells us he doesn’t bet, that could be a sign that he can’t make sectionals work as a stand alone approach but that’s not to say that using them alongside other strategies won’t be of benefit, same might apply with a collective strategy using the full intellectual capacity demonstrated by the members in this thread.

I would be very happy to nick anything that is worthwhile, providing I can understand it.

T tacker I’m less interested in betting than I am in finding a good way to bet, I have limited brain power and like to do things my own way. Sections interest me as speed figures don’t tell the full picture and long thought sections could fill in the missing pieces , analysing sections on not easy , I think people cleverer that myself don‘t always know how to evaluate the full performance.
That is what I’m trying to do, had a few failed attempts using individual sections and even combinations of sections.
First I believe a section means nothing in isolation , the example I would give Usain Bolt running 1500m race , he blasts out and is well clear after 400m then the lactic acid takes over and he is tailed off barely able to raise I leg at the finish, that first 400m section looks amazing but basically is useless unless you find a way to put into context with the whole race time.
Similarly Usain could jog around and sprint the final 400m , he would still finish tailed off in the race but the final section would look amazing on paper maybe people would look and think wow look how fast he finished and note the performance , but we know the fast sections where just a product of how he ran the race in the two examples.
So my section ratings take into account the overall race time, so the fastest horse in each section won’t always have the highest rating for that section.
Now I’m thinking instead of my previous failures using individual/combinations of sections I would use AustinDillon75 efficiency upgrade idea applied to my section ratings to see if this tells me anything worthwhile. Like AustinDillon75 says he benched the idea because even if a horse hasn’t run efficiently there is no guarantee it will do so in the future, I fully agree with him because I applied his idea to be ratings sheet 2 years ago and it didn’t make universal sense, as it it didn’t always weed out the Usain Bolt type scenario.
But applying to section ratings that already take into account overall race time looks more promising , bit of work to do yet.


Here is an extreme example , there are dozens of examples like this. Race 43 HK this season.



Bull Attitude finished 11th btn 8.25lengths , he had the highest rating in section 1 , much the fastest in section 2 and one of the slowest in section 3 and 4, these numbers already factor in the overall time but what interests me if i use AustinDillon75 efficiency uplift he has the highest in the race at 6.06 and this raises his final efficiency adjusted figure to 91.76, the highest in the race, suggesting the horse could have won.
Hewitson after the race said he was obliged to make use of his mount and thought horse would be better served by being ridden in a more conservative manner.
Next time the syndicates hammered Bull Attitude in from overnight 21 to 7.1.
This is the type of horse i want to identify and bet before syndicates destroy the price, obviously we all know bookies don't let you win, so even if it turned out a winning strategy, no saying you would be allowed to bet them anyway, i could put up dozens of examples like this , but i need to get it all sorted because for all i know there could be many more examples where it doesn't pan out in future races and be another failure.
Everything tells me this is a decent method of evaluating sections, without the silly ones being highlighted.
Something like horse who have an efficiency rating in the top 4 and efficiency rank at least 3 places higher than actual finishing position is where i'm looking to trigger bets.

 

[tacker]

O Outlander & AustinDillon75

Brilliant responses from both and if your two together with others on here can’t get it to work then it’s not a winning approach.

 

[AustinDillon75]

Obviously conscious this is the Hong Kong thread but it only takes a few posts to get all the thought processes going again. I'm away for the week in Wales but our caravan has WiFi and I'm sneaking on while no ones looking! I am going to take all these sectionals at some point and run a few theories.

 

[tacker]

Been watching a few replays and noticed this race was run at a rather slow pace, efficiency E.
Point being do we see this sort of a complicated pace in hong kong ?

Client Challenge

 

[gerry]

I have put a lot of thought into this lately not sure how i started thinking about it but it has gave me a lot of interesting points to ponder and in a of beat way gave me a good way to bet horses and working really well so far.
Anyway i started thinking of one horse as we are all looking for answers wither it be sectionals or full times or just class or what ever.
And i started to think like we all do what is the best horse and it just came out FRANKEL so simple really when you think of it.
And now i think the answers can be found there i think we where unfortunate that H Cecil was unwell when Frankel was here so he didn't shout it from roof tops like he did in younger years . But he did tell us all early on it was best horse he ever had and that was some statement considering what he had before. Why did he know that so early and before it made its name for its self well he new as he did tell us it would show up well in the july cup over sprints didn't say would win but said would show up well so he new it was fast very fast.
So he new was fastest horse he ever trained as he wasn't a sprint trainer but he knew it got a mile so he knew this horse will be unbeaten at anything from 7fur to mile say unbeatable actually.
Now he didn't need to worry if it went quick first furlong or last furlong as it was so fast all it needed to do was be with them at 6fur and then go and nothing would live with it as so proved no debate about that.
If they went slow early it would have left them for dead late on and if they go quick early would still cruise along with them then go and leave them for its devastating kick later.

So i then checked all its runs one after other and there is something that tells you along its path this horse is special and it won all those races because it was simply the best, but this is where people get mistaken saying well beat same horses and how do we know it was that good.
The answer is there right in front of us H Cecil told us it was and the horse proved it was for him no doubts about it and the prove is in the form.
So what has that got to do with all the other races now that's where i started it had all the answers it was the best it was consistent it specialized over the right distance and it was unbeatable as long as it stuck to what it was a flying machine over a mile.
So now we cant find a Frankel in ever daily meeting but we can find the Frankel of the race and how do we do it there no point rating them to a Frankel but we can look at it and see what it never done like , it never run bad race, it never run out its distance gap, and most importantly its temperament was flawless. we here all the time that in betting your temperament is crucial yet horses temperament is usually up and down and all over the place so we keep saying how will it run today were never sure second guessing every day going down a lot of loop holes till it gets to the point for the normal punter you would need be scientist never mind mathematician.
There so much on VDW on here about class and ability and he even mentioned temperament and every single one keeps saying yeah if your temperament is no good then you will never win at horses. What if he wasn't talking about your temperament he was talking about the temperament of the animals and i think the answer is in Frankel never heard anyone on VDW mention that yet its so logical it won ever race it ran there was reasons why it won every race and there in view, ok was short price just about all of them but the answer was there before the race each time i believe. Who looks at golf and does not study Tiger swing, who looks at football and not look at Messa, who looks at snooker and does not study O'Sullivan.
The best clue i think i could add to this is Tiger the best now NO, is Messa the best now NO, is O'Sullivan the best now NO, Why are they not still the best two reasons age and temperament.

 

[The Hare]

Sunday at Happy Valley this week

 

[Outlander]

Been watching a few replays and noticed this race was run at a rather slow pace, efficiency E.
Point being do we see this sort of a complicated pace in hong kong ?

Client Challenge

T tacker Very few races where they mess about dawdling, there are a few but rare. Most likely bit slower in races 1800m+

 

[tacker]

T tacker Very few races where they mess about dawdling, there are a few but rare. Most likely bit slower in races 1800m+

Might it not be preferable to see such variations if we're searching for something that might stand-out ?
Is it somewhat unlikely that horses being asked to run at the wrong pace early will compensate in the later stages of the race accurately ? by that i mean run the first 3f slower that ideal doesn't always mean that the last 3f will balance out correctly in terms of efficiently.

 

[Outlander]

Might it not be preferable to see such variations if we're searching for something that might stand-out ?
Is it somewhat unlikely that horses being asked to run at the wrong pace early will compensate in the later stages of the race accurately ? by that i mean run the first 3f slower that ideal doesn't always mean that the last 3f will balance out correctly in terms of efficiently.

Certainly doesn’t always balance out perfectly, that’s why you would see some horses/races run close or better than expectation and other will still fall short even after applying an efficiency uplift, just less so, maybe more accurate picture than overall time ratings alone ?

 

[Outlander]

T tacker
i don't have numbers for UK racing , i believe it would cost £100s a month to acquire the data needed.
i just cobbled this together in 5 mins, and tried on 1 race to see if i could get any feel if an uplift could work for UK with free ATR data.
Asked grok to put data in a table and copied to excel
then did a few formulas
chose a Newcastle race at random to have a look


So what's this telling me assuming its not a load of rubbish (it could be because i tried it on one race)
Looks like the winner could have gone min 0.60seconds faster if ran more efficiently(because i don't have pars i set the most efficient horse in race Balranald at 0 (100% efficiency)).
looking at the adjusted rank the only difference is that the 3rd horse maybe could have been 2nd.
Maybe this could give you more information on a race using free info, maybe not, no idea not tested it properly.
The sheet i used is attached just in case you want to play with other races, i don't have the interest in looking at UK