The Monty Shot predictor is based on an historical weighting of the difference between the number of shots for versus the number of shots against for a team (the focus of the initial model development was on the Hawks) and comparing to the actual value for 22 round into the future. A simple model was developed using the Hawks data from the 1930's to the end of 2012. This model prediction was used to predict each teams end of 2014 weighted scoring shot difference and then rank them.
The weighted column weights the shot predictor end of 2014 rankings with the end of 2013 shot predictor actual rankings and the Monte Carlo end of 2014 season predictions (which are erratic) and the Monte Carlo end of round 6 rankings.
As can be seen by comparing the final column with the various bookies estiamtes the top and bottom of the table do not vary much and the middle 6 shuffle around a bit. The differences are mainly my estimation that Adelaide will perform better than the bookies expect (by making/ sneaking into the 8) and that Essendon will perform quite badly. Most of the Essendon difficulties come from the poor finish to 2013 which trends the shot differences in such a manner as to make wins hard to come by - if Essendon perform as per the last two years they will reverse this trend in the first half of the season and taper off in the second half; the Bombers have certainly caused me issues with the modelling.
My analysis also leads me to believe that Adelaide and Richmond definitely have the rub of the green with respect to a potentially easier draw/ fixture with an ability to string a number of confidence winning streaks together.
Comparing the predicted shot difference numbers and comparing to the historical Hawthorn Grand Final Appearances and Wooden Spoons any weighted shot difference number as per my calculations greater than 5 puts you in contention for the flag; while a value less than minus 5 has you looking to collect a wooden spoon. By this analsyis only Hawthorn and Geelong are considered for the flag and Melbourne and GWS for the spoon.
The table and graph below shows the Hawks data:
The purple years are the spoon years, green are premierships and orange runners up. The numbers are my weighted values for shots for minus shots against. The graph below shows the values over the years (the x axis is the sequential round round number that the Hawks have played). A more detailed analysis including the other clubs will follow in a few weeks but the initial interpretation is that the forward prediction is a reasonable guide.
No comments:
Post a Comment