Thanks for posting this
Jackform
Instinctively I thought this was a non-runner and decided to have a look at actual results.
I set up 2 systems

Which I think cover the selection process.
I then downloaded the qualifiers from both systems and combined them sorting by trainer and race date.
I ended up with 16,390 selections from 1st January 2003 through to 4th February 2016
I then applied the staking plan.
The results are interesting!
Highest Bank +16,294 on 22nd October 2013
Lowest Bank -18,353 on 30th April 2010
The big issue is with the staking plan, particuarly with bit highlighted in Yellow
This results in a potential huge increase in stakes when a trainer has a longish losing run followed by a short odds winner and another losing run.
An example is shown below for PD Evans
The first column is the stake the sceond column is the resultant profit or loss from the bet and the final column is the running PL for the trainer.
So for the second bet shown a profit of 6.27 points is achieved but the running loss for Evans is now -27.73, so the next stake is half the current loss +1 which is 14.865, not a huge jump but look at the stake following the next winning bet (assuming I have interpreted the staking plan correctly)
