Thursday, 18 February 2016


In most endeavours, you can only really know if your actions and efforts are true if you first set out what what you expect the outcome of them to be.

So as per years past, this blog post will state the targets and objectives of the FMI blog Season 2016. A few common goals, and some new metrics are tabled below.


As per last year, the approach remains the same in 2016. Blog layout and data presentation is retained, with reduced commentary and opinion pieces. Time restrictions make anything more very difficult. There will be occasional posts with depth, perhaps one a month should be expected.
Contributions are, of course, always welcome.

As per 2015, the blog will done from Tokyo for a week or three as a vacation is planned. There should be minimal delays expected to the blog output (and also for FMITL'ers and FMICL'ers you will also notice an impact).
Efforts will be made to be punctual to a point, and your patience is appreciated.

As per 2015, there are other plans to contribute to another blog site. These will hopefully be fun and interesting, even though 'niche' in nature. With an 11 game membership this year (for the first time in 10 years), writing a bit about actual football could be an interesting diversion.
And much like last year's plans, they might not eventuate too.

2016 GOALS

Over the off-season, the tipping calculations have been tested and adjusted. Altering factors and comparing outcomes over the last 20 years confirmed the 'sweet spot'used in the past is still relevant. It also highlighted room for improvement in a number of areas.
Additional work and testing on changes to those have yielded benefits as here;
Over the 20 seasons, there was cumulative season on season correct tipping percent increases, as well as improved MAE.
Using the Monash competition methodology, game average point scores were also up overall, which reflected in more top-10 finishes in that competition.
Of course there were some seasons where the return was nominally worse or on par with using the 2105 system, but as there had been nett improvement on a number of scales, the new system has been implemented for season 2016.
The new calculations also impact the team ratings, and a the bottom of this post is a table that has the new season starting rankings points compared to the previous system.


Another correct tipping return over 145 (70%) should be expected, and if 150 (72%) is breached, even better.

The always elusive 9 from 9 correct tips is another target. In 2016, the target is to achieve 2 perfect rounds. A target missed in 2015.

The MAE (mean absolute error) is a measure of accuracy of predicted margin to actual margin. As was described to me by a regular reader and twitter correspondent (@SgtButane), a value under 30 would be a very good return.
Under 31 a realistic goal.

Tipping exactly the right margin is a highlight through the year, but is extremely difficult given the variety of score outputs possible.
A better target would be to look at margins tipped within the 2 goal range. Experience shows a target of 29% or more of games tipped in the 0-12 point should be achievable.
That is, on average, just under 5 games per fortnight (22%) tipped to within 2 goals of actual margin.

....And three exacts will be nice too.


Under the 2015 system a top 10 finish was achieved as the FMI rounded out in 9th after a terrible start to the season. Target 2016 is to again be top 10 minimum and top 5 would be ideal.

In terms of the total Monash score, 1850 is the minimum to meet KPI, and nudging past 1900 will be a great return.

TEAMPTS(Fmr System)
2West Coast1436201416
5North Melbourne1255101245
6Port Adelaide122431221
9W Bulldogs/FFC1159101149
14St Kilda8525847
16Gold Coast8381837
18Brisbane Lions781-2783

No comments:

Post a comment