JohnCharles
10-01-2015, 08:33 PM
Hi,
I am currently working on a regression model which is, unsurprisingly, based largely around speed/pace factors.
I have a velocity calculation that returns a figure relevant to the horses performance in 99% of cases. However, I have hit a snag with horses that do not finish the race in their first outing; either establishing a scientific way to assign a figure commensurate with its (poor) performance - which in such a case is obviously subjective - or to skip/ignore the trip.
I am dealing with an extensive dataset so video analysis is not an option for me.
Clearly there are merits for and against including/excluding it in the data (the obvious one being how can you assign it any value if it didn't complete the trip!) - and variables which make this more or less relevant on its next outing (e.g if distance has changed/surface has switched, etc). But my 'punting mind' tells me a horse unable to complete a race is significant.
Where a horse does not finish the race and is pulled up, when calculating a speed figure based on lghts behind the winner, I was wondering what others would do.
Being a relative newbie to US racing, does anyone know how Beyer/Bris/DRF/Moss deal with such situations?
I am currently working on a regression model which is, unsurprisingly, based largely around speed/pace factors.
I have a velocity calculation that returns a figure relevant to the horses performance in 99% of cases. However, I have hit a snag with horses that do not finish the race in their first outing; either establishing a scientific way to assign a figure commensurate with its (poor) performance - which in such a case is obviously subjective - or to skip/ignore the trip.
I am dealing with an extensive dataset so video analysis is not an option for me.
Clearly there are merits for and against including/excluding it in the data (the obvious one being how can you assign it any value if it didn't complete the trip!) - and variables which make this more or less relevant on its next outing (e.g if distance has changed/surface has switched, etc). But my 'punting mind' tells me a horse unable to complete a race is significant.
Where a horse does not finish the race and is pulled up, when calculating a speed figure based on lghts behind the winner, I was wondering what others would do.
Being a relative newbie to US racing, does anyone know how Beyer/Bris/DRF/Moss deal with such situations?