I agree. The credibility of the proposed bidding system is crucially dependent on the points scale. In the AFL article it says:
Will the AFL's new bidding system be fairer? - AFL.com.au
"Under the new system, draft picks from 1-74 would be allocated points to indicate their respective values, with pick No.1 being worth 3000 points, pick No.2 worth 2517 and so on in a sliding scale. This scale is based on data relating to the average player salaries of each pick from the year 2000 onwards.
The decision to use average salary as the data point for determining the value attached to each draft was sensible and logical. It threw up a result that was intuitive but also scientific."
But is it? When they say 'average salaries' have they used average for each pick in the same year of contract following the pick? For example, Year 3 or Year 4 (as player salaries are set in the first 2 years). How much difference would it make to look at different years of contract? Do all players mature and develop at the same rate after their pick and what does that mean for their 'value' at the time of the pick? And have they used real salaries by adjusting for salary inflation with general salary increases over the period?
And in any case, is the simple arithmetic mean the right data to use - as they are likely to be skewed upwards by outlier highly paid players. Did they also look at the median salaries for each pick and how would the use of those influence the points scale? (My guess is that it might compress the points scale compared to the one used, but that might be wrong.)
There is so little information provided on which to judge if the points scale is 'scientific' and yet it is critical to the outcome. I hope the club has been provided with all the background assumptions and data and is getting these rigorously examined by a statistical expert.
Bookmarks