What is the theoretical reasoning that underpins your Prediction Program?
Firstly, our Program's output is premised on the basic assumption that, on average, a demonstrably better/stronger team will beat a weaker team, and so the mechanism our "Predict-A-Win" Program uses to determine the relative strengths of the teams is to assign each team a Win Quotient for both the Home and the Away venues.
Secondly, through extensive testing and analytical work, we have established that both teams' Win Quotients need to be derived by amalgamating the Home and Away performances, as follows:
- for the Home Team we use a ratio of 70:30 (H:A), and
- for the Away Team we use a ratio of 70:30 (A:H).
Thirdly, we have established that a team's past results history is a good indicator as to what it will do in the future, so we utilise a team's average score-lines over the past 20 or so matches in our Program's algorithms (the actual number being dependent on how many matches are played in a full season). However, a number of the matches we take into account will be for the "cloned" results of the Relegated and Promoted teams, where we use the outcome of "artificial" matches and the score-lines for those artificial matches in the algorithms. Our reasoning is that the outcomes of the matches for the teams that were relegated or promoted from a Division are completely irrelevant in the new season, and that the typical "averaged" expectation of a "Staying" team's performance is more solid (worthwhile) for calculation purposes.
Fourthly, the Win Quotients are used in a suite of algorithms combining the average scoring abilities of the teams plus their vulnerability (how often they let goals in). The "base data" we employ for calculation purposes is updated weekly, which allows the Program to calculate revised Win Quotients, average scoring and average vulnerability figures, and then the new predictions are based on that updated data.
Without variance we use the same algorithms in exactly the same way for each and every match prediction. This then enables us to have 100% confidence in the Reliability figures derived through running our Program.