Results 1

This batch of tests was run over the period 10/06/2013 to 28/06/2013 (1 day missed).

Findings

The drop offs at the end has been traced to a bug in the App. Essentially it was in the interpretation of going for the races, in a narrow going range. The weather was relatively stable, but on the fall off days there was rain impacting the going , which stumbled into the bug. The bug meant the algorithm was being fed false data , with the anticipated results.

Beyond this, the trend in the results looked quite promising.

Tests 2, starting from 29/06/2013, will record the new trend.

Result Data

The following graphs provide a view on the performance of the Predictor, verified by the witnesses. The results are up to date as of 28/06/2013.



The above shows the percentage gain, per day, when betting as the predictor indicates.



The above shows the aggregate gain, if the initial pot amount and subsequent gains are re-invested the next day.



The above shows the aggregate gain , based upon different percentages of the pot being bet each day.