Thanks Fryingpan and Peakester for your suggestions.

I already had thrown a distance change filter over the above figures (
This changed the above findings to: S/R 128/370 (34.6%)
POT 30/370 (8.1%)

Peakester's filters added to this resulted in

(i) 50 % PP S/R 103/284 (36.3%)
POT 38/286 (13.2%)

(ii) POT 29/258 (11.2%)

(iii) POT 35/295 (11.8%)

(iv) all three filters S/R 63/160 (39.4%)
POT 28/160 (17.5%)


This is now 82 weeks of data so method (i) constitutes just less than 2 bets per week,
method (ii) approx 3.5 bets per week.

Is anyone able to run this over a longer scan please?