Early speed, Even the computer failed.

[teddy]

I was looking at my softwares performance for the last month. It predicts hoses with an early speed advantage.

589 races were races where the computer says one horse has a big advantage, yet only 278 were first call leaders. The ones that made the lead returned 1.50 roi. The others returned 1.00. I dont have an easy ability to analyze deeper on what traits the 278 had in common. Anyone else dig deep here?

[JustRalph]

maidens included ?

[Steve 'StatMan']

separate turf & dirt as well

[dnlgfnk]

Quarter calls, seemingly a window to the early pace, can themselves be interpreted in light of 1/16ths and smaller indications of what was going on early within the quarter mile, though it must be done with informed speculation from the start call to the quarter.
In today's opener at Belmont soon to go off, Stellaluce probably figures big in software early fraction ratings. One significant tool for the trip handicapper, however, may be that for the first time in their two recent meetings, she draws outside Chaka Bwana, and could easily require more effort than in the past, even before the quarter pole reflection. When incorporating public odds into the mix (ala Benter and others suggesting it's worth), I arrive at Chaka Bwana being an early and detrimental pace factor to Stellaluce, but probably not a realistic overlay. I would downgrade Stellaluce without removing her from exotics, mostly underneath.

[teddy]

Quarter calls, seemingly a window to the early pace, can themselves be interpreted in light of 1/16ths and smaller indications of what was going on early within the quarter mile, though it must be done with informed speculation from the start call to the quarter.
In today's opener at Belmont soon to go off, Stellaluce probably figures big in software early fraction ratings. One significant tool for the trip handicapper, however, may be that for the first time in their two recent meetings, she draws outside Chaka Bwana, and could easily require more effort than in the past, even before the quarter pole reflection. When incorporating public odds into the mix (ala Benter and others suggesting it's worth), I arrive at Chaka Bwana being an early and detrimental pace factor to Stellaluce, but probably not a realistic overlay. I would downgrade Stellaluce without removing her from exotics, mostly underneath.
I have chuka much faster than stella in my software...I also considered post position might be helpful to determine who will top. What I have is strictly feet per second.
I will break down by alw, clm mdn mdnclm... and report last 30 or 40 days.

[PhantomOnTour]

Does your software account for a class dropper like the #4 who wound up as the early leader?
He came out of faster paced races than the rest, and even though he wasn't on the lead, he laid close to quick splits.
Sometimes the jock (esp a bug like Ortiz) will take the dropper and try to get the classy horse out front where he may get brave.
This guy folded up by the 1/4 pole, but the intention is what put him on the lead imo.

[teddy]

ok

predicted....... actual first calls leaders

alw 398 155
clm 835 350

mdn 249 86

mdn clm 387 153


sprints.... 894 262

mile 291 194

tf sprint 104 31

tf route 193 70

Im thinking field size will effect the ability of the software to be correct more often due to chaos..it did much better at routes...usually smaller fields and less pace pressure.

[Overlay]

As important as early speed and the ability to forecast which horses will get to the front in the initial stages of a race are, the results indicate the limitations of both the predictive power and the pari-mutuel value of any one handicapping factor, no matter how fundamental.

[teddy]

Let me check the stats on the top two speeds in races where one has a speed advantage will work out... hey maybe dutch the top 3 likely to make the lead.. who knows...

[Robert Goren]

If you really want see PP go bad at predicting early speed, wait till you track when anybody who gets to front wins easily 100% of the time and the jockeys all know it. Then it gets really hard to predict it. It amazing how fast many horses are when they are asked for it. I remember a number of years ago at MNR, they had a frozen track. Every horse who was ahead at the first call won for days. Horses that had never seen fourth place at first call were all of a sudden getting 2 or 3 length lead there.

[LottaKash]

As important as early speed and the ability to forecast which horses will get to the front in the initial stages of a race are, the results indicate the limitations of both the predictive power and the pari-mutuel value of any one handicapping factor, no matter how fundamental.

Couldn't agree more Overlay...

I believe that, despite a horse's proclivity to run early, current form and ability, and the Trainer's intentions on raceday, may have as much to do about a "go" or "no-go", early today, as anything...

best,

[classhandicapper]

If you really want see PP go bad at predicting early speed, wait till you track when anybody who gets to front wins easily 100% of the time and the jockeys all know it. Then it gets really hard to predict it. It amazing how fast many horses are when they are asked for it. I remember a number of years ago at MNR, they had a frozen track. Every horse who was ahead at the first call won for days. Horses that had never seen fourth place at first call were all of a sudden getting 2 or 3 length lead there.

Good insight. :ThmbUp:

By the way, this can also cause major problems for making pace figures and analyzing a bias.

When everyone is gunning, all the pace times will be very fast. So if you make a track variant for the pace call, you will be hard pressed to figure out how fast the track actually was to the pace calls because the races are developing so much differently than usual.

On the bias side, some speed favoring tracks are strong but not so strong that a horse can get away with anything. So if everyone is gunning, some closers will win despite the strong bias. When a few closers win, it's less obvious to a beginner than there was a speed bias.

[classhandicapper]

Does your software account for a class dropper like the #4 who wound up as the early leader?


I agree that this is an issue.

As dnlgfnk alluded to, sometimes things are going on in the first 1/16th or first 1/8th that you can't really see in the 1/4 mile time.

In higher quality races horses tend to be faster. As a soon one or more establish control, the others often kind of settle in behind them and sometimes run slower than they are capable of. When they drop in class, they are fastest and seize command and run faster fractions.