The Horse's Form Cycle

[tfm]

Originally Posted by Ted Craven

Can you tell me the next race, for each of Real Tipsy and Skewed?

Reel Tipsy -- WO_10-28-09_R6

Skewed -- WO_11-12-09_R5

[Ted Craven]

Originally Posted by tfm

Reel Tipsy -- WO_10-28-09_R6

Skewed -- WO_11-12-09_R5

TFM, I apologize that I haven't studied your posts on setups closely enough to be able to intuit more about the moving parts of your analysis, however, just relating to the 3 examples in this thread, it strikes me that all the data is there to code the observations (or at least bring them to further human attention).

1. the 'angle' you refer to (I presume meaning 'angle of n degrees slope') can be rendered numerically as change in beaten lengths (or change in velocity), but as I see it, qualified by position gain of at least n positions. In the attached RDSS screenshots, each horse shows a so-called 'powermove' (gain of 3+ positions in a segment, or got lead at 2nd or Str call and held it through finish by 3+ lengths) during the same race segment you identify in the fat-charts.

2. the context of the race shape in the previous race, where the horse made a strong move: presume that for any given race today, we have access to all (or recent) Result Charts for a given horse's past running line (so we can analyse what was happening in that previous race). Today's PP line shows the position gain and how many other horses were still ahead of our horse and how close clustered behind the leader (an indicator of pressure at that point) and where in the race the gain and clustering occurred - but it does not show how many other horses were how close to but behind our horse, and what if any effect their run had on our horse. The past Result Chart shows this clearly (and a graph of it shows it best).

What other data would be required - in general - (beside the details of intra-segment moves, path-wide, visual trip info from videos)? In general, for a race today, for all horses entered, presume that we can view the Result Charts of all past races, render them as graphs, know what the PP line of each horse in that past race would look like just as if it were running in today's race (e.g. in a Tandem) - AND, that we can know the past history of each horse in that previous race, and regress from there more or less until we don't have the data anymore. This last bit presumes someone has all PPs, all Charts for all North American races for a few years - quite a bit of data either on your own computer or accessible on demand from (say) TrackMaster/Equibase servers, but not really a logistical issue.

Does this give the raw data to begin codifying the collection of scenarios and sub-scenarios you currently study?

Ted