This page quickly walks through how sectionals were calculated that are visualised here.
First I downloaded a version of the race from youtube and loaded it into Kinovea, which is a free piece of video editing software for sports. A stopwatch was added to the screen, and was started as the gates opened for the Oaks. I used my past experience of calculating sectionals for the Derby and Oaks in 2013, and used paths which are consistent year on year. Two paths at roughly 5f and 8.5f were used, and then furlong poles at 10f and 11f.
The four screenshots below show the paths used, the screenshots can be found here. Video analysis of the Derby allowed me to move frame by frame clocking individual runners times over the path at 5furlongs. At 8.5furlongs, lengths were estimated using the screenshot seen below, and a 6 lengths per second scale was used to calculate times, this produced some differences between my sectionals and those of Timeform and Simon Rowlands (here), most likely due to Timeform using a more precise lengths per second scale at the sectional. Unlike the Oaks the field is stretched out which made life tougher. The two screenshots for 10furlongs and 11furlongs are unsurprisingly much tougher as runners are head on, and some are completely out of shot.
The table below shows the individual times used, or the cumulative distance (in lengths) that I estimated runners were behind the leader.
This data was then used, with the leaders time (seen in each screenshot), to calculate the different variables that are found in the actual visualisation. To calculate them I used my R package, called RcappeR. The script here calculates cumulative times, sectional times, finishing speeds, seconds behind leader, and positions at each sectional. If anyone spots any discrepencies please highlight them.
Pos | Horse | 5f | 8.5f | 10f | 11f | 12f |
{{ runner.pos }} | {{ runner.horse }} | {{ runner.c5f | number: 2 }} | {{ runner.len8_5f }} | {{ runner.len10f }} | {{ runner.c11f | number: 2 }} | {{ runner.len12f }} |