Looking for bias in the K9 Lottery!

Hi all.. Just thought I would share a little project I have been working on over the past month or so with greyhound betting.
I wanted to see if I could look at any trends with regards to trap, location and time bias that may occur when betting on the dogs.
The heart of a system is a trigger that will place a lay bet on a trap based on certain calculations regarding wom and volume % etc... normal kind of stuff.
The results of this trigger is then fed into a database which, over time, will look at which location/trap/time is doing well and which isnt.
Once enough data is obtained we can start to see a trend and have a new "Live betting" trigger which is actually placing back/lay bets and a varying stake depending on whats worked well previously.
Here is some stats... you can see in the graph, the profit with the learning system in place compared to straight £1 lays
You can also see the daily profit, note the profit is increasing as the days go by as the system "learns" the bias.

Ill post the images below.....

Here you can see the graph of the last month of so.... system compared to straight, blind lays.

Here you can see the daily profit

The above results are based on nearly 5000 races