Two horses can finish a race in the same time and have run completely different races. One went off fast and held on. The other sat at the back and sprinted at the end. The finishing time cannot tell them apart. Sectional times can.
Sectional times in horse racing are the times a horse takes to cover each stretch of a race, usually a furlong or a few hundred metres at a time. Put them together and you get the shape of the race: where it was quick, where it slowed and who was still going at the line.
A UK race, split by split
Take a real one. On 30 September 2026 at Kempton, eleven runners went over six furlongs on the all-weather. The winner, King Cherry, took 1 minute 12.45 seconds, and this is how it got there, from the start to the line:
| Stretch | Time |
|---|---|
| Start to 5f | 14.01 |
| 5f to 4f | 11.40 |
| 4f to 3f | 11.61 |
| 3f to 2f | 11.67 |
| 2f to 1f | 11.52 |
| 1f to finish | 12.24 |
The six numbers add up to 72.45 seconds, exactly the finishing time, so nothing is missing. Two things stand out. The first furlong is slow because horses leave the stalls from a standing start. And after that the last furlong is the slowest of the lot, 12.24 seconds against about 11.5 for the middle ones. King Cherry was tiring, as most horses do at the end of a sprint.
The runner-up, Respectable Jack, was slower early, taking 14.96 for the first furlong, nearly a second behind. Then it was quicker than the winner through each of the last three furlongs: 11.44, 11.23 and 12.14 against 11.67, 11.52 and 12.24. It still finished 2½ lengths behind. That story is invisible in the result alone.
A Hong Kong race, section by section
Hong Kong divides the course into sections, up to six, rather than furlong posts. On 1 October 2026 at Sha Tin, fourteen runners went over 1200 metres on good to firm ground. The race ran in three sections and the winner took 1:08.76.
| Section 1 | Section 2 | Section 3 | Finish | |
|---|---|---|---|---|
| Winner | 24.24 | 22.50 | 22.02 | 1:08.76 |
| Third place | 25.24 | 22.26 | 21.38 | 1:08.88 |
Hong Kong also records the position of each runner at the end of every section. The winner was third, then first, then first. The horse that finished third was twelfth after the first section and thirteenth after the second, then ran the last section in 21.38, faster than the winner’s 22.02. That is a closing run you only see in the splits, and it explains why a horse that looks beaten on position may have run a very good race.
What the numbers are good for
Pace. Compare how fast the race went early with how fast it finished. A race run quickly early is commonly said to suit horses who come from behind, and a slow one the opposite. With splits you can measure whether that is what happened.
Finishing speed. The last section or furlong, on its own, is a cleaner measure of a horse’s closing ability than finishing position.
Comparison. Because the splits add up to the finishing time, you can compare two horses stretch by stretch instead of only on the total.
Our UK records also include derived speed fields, with an average speed for the early, middle and final parts of the race in miles per hour. For Kempton’s winner the first two furlongs come out at 35.42, and it checks against the splits above: 402 metres in 25.41 seconds is 35.4 mph.
The catch
Splits are only as good as their coverage. In Hong Kong nearly every runner has them from 2009. In the UK they begin in 2017 and cover only a share of runners, which our coverage by year note spells out. If you train a model on sectionals, you are training on the races that had them.
Splits describe how a race was run. That is a more modest claim than forecasting the next one, and a much more reliable one. You can see both formats in the horse racing datasets, and the API returns the same fields for UK records and Hong Kong races. For the wider picture of Hong Kong racing, read the Hong Kong racing guide.
Comments are closed