TrapMetricsTrap bias at Oxford: which traps actually win
Trap bias at Oxford measured on 4,680 races over 2 years: share of wins by trap against fair share, by distance and by grade, with the noise called noise.
Over the last 2 years at Oxford, 4,680 races, no trap has been clearly better or worse than its fair share: every difference is inside noise. The track is 397 metres round with 100 metres to the first bend, measured from our own split times.
All distances
| Trap | Runs | Wins | Win rate | Share of wins | Fair share | Bias | Read |
|---|---|---|---|---|---|---|---|
| 1 | 3,968 | 744 | 18.8% | 15.9% | 16.2% | -0.3 | noise |
| 2 | 4,427 | 861 | 19.4% | 18.4% | 18.1% | +0.3 | noise |
| 3 | 4,373 | 861 | 19.7% | 18.4% | 17.9% | +0.5 | noise |
| 4 | 4,311 | 802 | 18.6% | 17.1% | 17.6% | -0.5 | noise |
| 5 | 4,253 | 848 | 19.9% | 18.1% | 17.4% | +0.7 | noise |
| 6 | 3,133 | 564 | 18.0% | 12.1% | 12.8% | -0.8 | noise |
Trap 1 sits inside noise at -0.3 points. Trap 2 sits inside noise at +0.3 points. Trap 3 sits inside noise at +0.5 points. Trap 4 sits inside noise at -0.5 points. Trap 5 sits inside noise at +0.7 points. Trap 6 sits inside noise at -0.8 points.
By distance
Only distances with at least 1,500 runs get their own chart; a bias that shows at one trip and not another is usually the run to the first bend at work.
450 metres, 3,047 races
Trap 1 sits inside noise at +0.2 points. Trap 2 sits inside noise at +0.0 points. Trap 3 sits inside noise at +0.8 points. Trap 4 sits inside noise at +0.1 points. Trap 5 sits inside noise at +0.5 points. Trap 6 is a clear minus: 10.5% of wins against a fair share of 12.1%, -1.6 points on 1,914 runs.
253 metres, 1,372 races
No clear bias at 253 metres: every trap is inside noise on 1,372 races.
By grade
Grade letters grouped: A for graded races, D for sprints and dashes where the track uses them, S for stayers, and so on.
| Grade | Races | T1 | T2 | T3 | T4 | T5 | T6 |
|---|---|---|---|---|---|---|---|
| A | 2,659 | -0.2 | -0.3 | +1.1 | +0.2 | +0.5 | -1.4 |
| D | 1,290 | -1.5 | +1.6 | -0.4 | -1.4 | +1.7 | +0.0 |
| O | 621 | +1.5 | +0.9 | -0.2 | -1.7 | -0.6 | +0.2 |
Bias in points against fair share; * marks a clear bias.
The market and the fields
The Betfair favourite won 38.7% of 4,657 priced races here over the window. 55% of races went off with six runners; the rest had a vacant trap, which is why fair share is measured on runs and not assumed to be one sixth.
How to read this
Fair share is each trap's share of the runs, so a vacant trap 6 does not hand trap 5 a bias. A bias is called clear only when it is at least 1.5 points and at least 2.5 standard errors from fair on the runs behind it; anything smaller is a difference inside noise, and the page says so. The window is the last 2 years, refreshed whenever this page is rebuilt. The live version, with the going and our own accuracy at this track, is on the track page.
Questions this piece answers
Which trap wins most at Oxford?
Over the last 2 years no trap at Oxford is clearly better or worse than its fair share of runs: the differences are inside noise on 4,680 races.
How is fair share worked out?
Each trap's fair share is its share of the runs, not one sixth. A trap that was vacant in many races has fewer runs and so a smaller fair share; comparing its wins with a sixth would invent a bias.
How often does the favourite win at Oxford?
The Betfair market favourite won 38.7% of 4,657 races with a settled price over the last 2 years.