5. How Cloning Works

The basic principle we employ for determining the relative strengths of the Incoming Teams is to clone them, using the data of what our Program calculates to be the closest match-up from amongst the Staying Teams in the new Division. All Promoted Teams are cloned against the same two Staying Teams residing in the bottom half of the points table (one Staying Team for the Home venue performance data and another Staying Team for the Away venue performance data). Likewise, all Relegated Teams are similarly cloned against another set of two Staying Teams sitting somewhere in the top half of the points table.

The first step in the cloning process utilises a small algorithm, based on the following parameters, to determine which Staying Teams are to be used for cloning purposes, based on their respective Win Quotients:

how far down we consider Relegated Teams should initially sit from the top of the table, and
how far up Promoted Teams should initially sit from the bottom of the table.

The cloned data obtained in that cloning process is only in respect of the values to be utilised in respect of the following parameters:

Win Quotients,
Scoring Abilities (goals for), and
Vulnerabilities (goals against).

Having conducted a full-season ‘artificial’ predictions run, the Program then provides two different options for incorporating the artificial match results and score-lines for the Incoming Teams into an initial new-season database:

Either: Copy exactly what the previous season’s score-lines (and thus the results) were for all those matches where the Home and Away Parent Companies played the other Staying Teams.
Or: Use the artificial match results for where the Incoming Teams played the Staying Teams, except that, for where the Incoming Teams played their respective Parents, then all such matches should be shown as a Draw with a 0-0 score-line.

No matter which of the above two alternative cloning options is chosen, the final step is then to over-write the artificial results for any Incoming Team that played another Incoming Team last season with actual score-line (and thus the actual result) for that match in the previous season. This then gives us a complete set of match results and corresponding score-lines with which to commence the predictions for the first set of fixtures in the first week of play for the new season, albeit that up to a quarter of those results are artificial at the start of a new season. The content of that database can be viewed in a report titled ‘Reworked Results Matrix’, which contains the Starting Data for whichever season is being considered (i.e., Last Season or Current Season). Of course, as each season progresses, the actual results of matches played will over-write not only any previous actual match results but also any artificial match results in the starting matrix, thus gradually improving the quality and usefulness of the matrix of results.

We employ the above procedure to try to overcome any problems that could occur if we applied the actual results of just one team being cloned against as the results to be used for the Incoming Team’s artificial Home and Away results. We do that because it is inevitable that some of the Staying Teams would have had quite a few match results in the previous season that were not in line with their average level of performance. On the other hand, since the algorithms in our Program are centred predominantly on the ‘Law of Averages’, it must also be recognised that the predicted results obtained for the Incoming Teams will inevitably be somewhat Utopian, as in the real-life situation those Incoming Teams could never be expected to play to that same level of consistency for every single match. But there had to be some trade-offs to allow us to incorporate the Incoming Teams into their new Divisions, where a good number of Divisions had lost many of their former teams due to relegations and promotions for the upcoming new season.

If all that effort just to start the first week of predictions for a season sounds complicated, it is simply because it is! But our Program’s Users do not have to worry about it, as it is all taken care of with ease in the background, once the Decision Factors Input (DFI) parameters have been selected and the ‘Process DFI’ button has been clicked on. Further, our Program produces many reports that will enable the User to check each critical step of the preparation work for the upcoming season’s database of starting material, to see if changes are needed in order to obtain a better set of starting data. In fact, since the integrity of the cloning process is so vital, we have compiled 12 separate reports in all that are specifically meant for the Users to check the validity of this cloning process. This was done in the certain knowledge that, if the User did not start out the season with valid data, there would be little chance of obtaining worthwhile outputs from the Program’s weekly prediction runs.

[84/1000 to date] [for bonus subscription - see foot of Home screen]