H32: early jumping error then back into it

The notebook

In order: what the lab saw, what it asked, what it proposed and why, what the tests said, and what an investigator concluded. Numbers are aggregates on the discovery data unless a step says otherwise.

1. Hypothesised15 Oct, 23:10

   H32: early jumping error then back into it KILLED amber

   because An early jumping error last time (no slow start), got back into the race, beaten 5 lengths or less.

   last_early_setback = 1 AND last_recovered = 1 AND last_jumping_error = 1 AND last_slow_start = 0 AND last_position > 1 AND last_beaten_lengths <= 5 AND jumping_errors_last3 <= 1

   made by import

2. Tested15 Oct, 23:10

   broad: A/E 1.15 on 136 bets, by period 1.10 · 1.29. info

   Betfair SP return +22.0%. Every period above 1.00 (spread 0.19). Zone: all. On H32: early jumping error then back into it

   By period

   betsA/E0.71.01.3

   Earlier half911.10

   Later half451.29

3. Tested15 Oct, 23:10

   chase: A/E 1.41 on 35 bets. info

   Betfair SP return +35.5%. Zone: all. On H32: early jumping error then back into it

   By period

   betsA/E0.71.01.3

   Earlier half27–

   Later half8–

4. Tested15 Oct, 23:10

   hurdle: A/E 1.17 on 73 bets, by period 1.22. info

   Betfair SP return +20.3%. Every period above 1.00 (spread 0.00). Zone: all. On H32: early jumping error then back into it

   By period

   betsA/E0.71.01.3

   Earlier half481.22

   Later half25–

5. Killed15 Oct, 23:10

   H32: early jumping error then back into it Discovery → KILLEDimport: tier C: rejected, do not test again unless new data changes the idea; best [chase] AMBER A/E 1.405 on 35 bets (broad AMBER)