Quote:
Originally Posted by JerryBoyle
Jeff, I know you mentioned before that you pull weather data. Have you ever looked into how well weather data can be used to approximate or improve on the track condition as given by Equibase?
My initial thought was that one could get a better sense of the track's condition by looking at the weather for the last 1hr, day, week etc via some type of modeling vs just using Equibase's description. The goal would then be to develop this model and completely stop using Equibase's description. However, I'm wondering if Equibase incorporates anything that they might know about how management is handling the weather w/r/t sealing the track, etc. If Equibase does have this info, then there would still be value in their description. Just curious if you've ever looked into this..
|
Yes, I've developed a model that estimates residual moisture in the ground from weather data.
I did so to get a better understanding of track bias, to eventually create my own designations for track condition, and eventually move beyond the Equibase labels for track condition seen and used by everybody else.
The model as it currently exists does a pretty accurate job of estimating residual moisture in the ground.
But it's definitely a work in progress.
The model would be perfect if I were trying to grow alfalfa in a field.
I'm discovering an accurate estimate for residual moisture in the ground doesn't always translate to Equibase track condition.
• I've seen cases where weather data shows several inches of rain in a 24 hr period for the immediate area where a track is located. Yet somehow the track is actually fast a few days later.
Example --
The Paulick Report | 12-23-2019 | More Than A Foot Of Rain Forces Gulfstream To Close For Training, Relocate Some Horses:
https://www.paulickreport.com/news/t...e-some-horses/
Fyi, the Gulfstream dirt surface was actually fast for their Thurs 12-26-2019 opening day card. (A pretty quick turnaround given there was more than a foot of rain and video of people kayaking in the flood waters between the barns on the backside.)
My guess in cases like this is the grounds crew had ample warning and was able to seal the track ahead of the rain event.
As a result, a high percentage of the precip became runoff instead of soaking into the track surface.
• There's also the case where weather data only shows light precip for the immediate area where a track is located.
Yet, somehow the track ends up sloppy or muddy.
My guess in cases like that is maybe the square mile where the track is located got more rain than the surrounding area.
In cases like that it's also possible the grounds crew wasn't able to seal the track ahead of time, allowing much of the precip to find its way into the track surface.
• I've also seen cases where tracks label their surface as fast despite evidence to the contrary.
I'm sorry, but when thoroughbreds are getting 6f in 1:15 and change on dirt that's clearly discolored and they're still calling the surface fast:
I've learned to take that with a grain of salt.
In cases like that, I've found a residual moisture model to be really useful.
You asked if Equibase incorporates anything they might know about how management is handling weather with respect to sealing the track, etc.
Data points I've seen in the chart data:
An indicator for sealed.
Y = Sealed. Otherwise blank.
A letter code that gives you a general description of the weather that day.
A = Hot
B = Fine
C = Clear
E = Overcast
F = Foggy
H = Hazy
L = Cloudy
O = Showery
R = Rainy
S = Snowing
T = Stormy
W = Windy
And of course track condition:
FST = Fast
WF = Wet-Fast
GD = Good
SLY = Sloppy
MY = Muddy
SL = Slow
HY = Heavy
FR = Frozen
HD = Hard
FM = Firm
SF = Soft
YL = Yeilding
Along those same lines, I've seen Woodbine report depth of harrows for maintenance of their dirt course between races right on their track video.
I've also seen Woodbine report resistance from a 'going stick' or Stimpmeter for their turf courses as part of their track video.
I wish more tracks would do that.
-jp
.
