SRI and DRI figures on Reports
The SRI and DRI are figures which can help indicate when a system is safe to follow.
When you create a new system the results may not be sufficient in size in order to be reliable (the sample size can be too low). But how low is too low? When is a sample size big enough to be reliable?
Sample sizes can vary depending on the type of system. If you have a system which is based on short priced favourites then you the sample size does not have to be that big in order for the system results to be reliable. But if you have a system which is based on outsiders then you should use a much bigger sample, as outsiders can have long losing streaks and you would not know if the system is on a losing streak or is just unreliable.
The DRI is message board poster Dylan's Reliability Index. This is based on the following formula:
DRI = Number of Wins * ROI%
If the DRI figure is greater than 25 then the results of the system so far should be reliable (i.e. they are not that random, due to chance etc).
The SRI figure is based on message board poster Selphie (Steve Tilley)'s Reliability Index. This figure is the sum of the expected wins of the system where the expected wins is the E part of the A/E figure. An expected win is 1 / (SP + 1) thus the SRI formula is:
SRI = SUM(1/(SP + 1)) for each horse in the system
e.g. System A has 30 results and returns a 10% profit. The odds of all those results are 10 x Evens, 10 x 2/1 and 10 x 3/1 shots.
SRI = (10 * 1/2) + (10 * 1/3) + (10 * 1/4) = 10.8
e.g. System B has 30 results and returns a 200% profit. The odds of all those results are 15 x 8/1 shots and 15 x 20/1 shots
SRI = (15 * 1/9) + (15 * 1/21) = 2.38
The SRI shows robustness when it reaches 10.0 thus System A above can be deemed to be reliable, whilst System B above should not yet be relied upon as there is insufficient data to fully assess the robustness.
More Info on SRI and DRI
The message board had a long discussion about Stats Confidence Factors. You can read the thread and how the SRI and DRI figures were calculated here: Stats Confidence Factors
Note too that these figures only indicate when a system has had sufficient runners in order to make an assessment. If you want to know if the previous results are reliable, or are more likely to be the result of random chance then look for the Archie or chi squared figures.