Synthetic shippers

[bobphilo]

Originally Posted by ultracapper

The way I understood it was the artificial surfaces theoretically would have minimum variances. I swear there are days when it seems like Golden Gate is souped up. I didn't think they could effect it much via the daily maintenance procedures, but it seems to me that GG can play awfully fast certain days, particularly at the top of the stretch.

When GG horses ship to SoCal for dirt races, I look at their races at GG just as if they were run on real dirt.

There are many factors which can vary from day to day which can give the illusion that a fair unchanging track is "speed favoring". Maybe some of the front running winners got away with a slow pace or they rode a golden rail or the closers got caught wide or were trapped in traffic or the front runners just happened to be the best horses. Let's not forget variability in performances. When one considers all these other possible reasons one is left with few if any cases that have anything to do with the "how the track is playing".

[ubercapper]

I was able to easily run a bunch of queries using Stats Race Lens on this subject. The great thing about the querying tool is when using a specific track the data set now covers five years:

The criteria was "Tampa Bay Downs, Turf," "last race different track,"last race all-weather."

315 starts covering 268 races
8% win, 7% place, 10% show, ROI -63%

Changing the surface for the race at Tampa to dirt, the results were:
448 starts covering 400 races
8% win, 8% place, 12% show, -67% ROI

I could easily break these down by sprint, route and other factors (such as track of last race) but I imagine the sample sizes are going to get small enough with additional criteria that the confidence level may be insufficient to draw different conclusions than the big set.

[usfgeology]

Originally Posted by ubercapper

I was able to easily run a bunch of queries using Stats Race Lens on this subject. The great thing about the querying tool is when using a specific track the data set now covers five years:

The criteria was "Tampa Bay Downs, Turf," "last race different track,"last race all-weather."

315 starts covering 268 races
8% win, 7% place, 10% show, ROI -63%

Changing the surface for the race at Tampa to dirt, the results were:
448 starts covering 400 races
8% win, 8% place, 12% show, -67% ROI

I could easily break these down by sprint, route and other factors (such as track of last race) but I imagine the sample sizes are going to get small enough with additional criteria that the confidence level may be insufficient to draw different conclusions than the big set.

This is awesome. Thank you. Saved me a lot of ink and headache to disprove a dead angle.