Quote:
Originally Posted by Poindexter
Regarding the stats presented, how did you come up with an expected 23.5% wins? Horses won 21.23%(5/23.5) as often as they should have but had a 50% roi. That cannot be right. Even the favorite data doesn't make sense. Favorites have to win at at least a 30% clip(expected) you have them winning 13.6% clip. So given the typical roi on a horse is about .8, the roi probably should be less than .40 yet you have them at .639 roi.
|
The ROI's are $2 not to the dollar. Expected is calculated using the Gordon Pine method found here (and yes, takeout is included in the calculations):
http://www.netcapper.com/TrackTracts...e/TT010223.htm
Quote:
|
Originally Posted by bobphilo
I too am skeptical of some of these findings. I would be interested in the source of this study. Was this done by an established organization like the DRF or Equibase? DDid it have anything like the rigor of the landmark study by Prof. Quirin? What exactly was the protocol and sample selection?
Given that starting gates are designed to open fairly easily precisely for the reason to avoid hurting any horse that breaks through, I find such a dramatic affect hard to believe.
The original claim of 92% losers for breakthroughs is meaningless on the face of it since about 90% of all starters are losers to begin with. The % of breakthrough losers that are short priced favorites is bound to be lower still. It's actually possible that horses that break through my have an advantage since they may be more eager to race?
|
I searched the long comment of Equibase charts for the specific term: "broke through gate". I did that for 10 years of charts that included every t-bred race in the US in the last 10 years. You do not have to believe the results as the query is not precise. I'm sure there will be horses who broke through the gate that were not noted by the chart caller. It's imprecise science. If you actually believe horses who break through the gate are at an advantage, I'm happy to bet against you! I'm pretty sure this data is reasonably accurate.
