Full season monte carlo (no draws) shot predictor model: final ladder predicitons

The Carlo Monty laptop has been running hot crunching numbers with a simplified iterative model predictor to sooth say (or out sooth say as best as possible the Footy Maths Institute!)

Adelaide still persist in coming out on top despite the large number of tight games.  GWS has a bit of a monopoly on the wooden spoon with Western Bulldogs giving them a run for their money.

Top eight certainties at the moment are: Adelaide, Hawthorn, Saints, Swans and Eagles.  Freo and Cats almost certain top 8.

Monte Carlo	Scoring shot basis (no draws)
Full season	100 iterations
Team	Win	2nd	Top 2	Top 8	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
Adelaide	24%	25%	49%	100%	50%	20%	8%	15%	7%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%
Brisbane	0%	0%	0%	63%	0%	0%	0%	0%	0%	6%	15%	42%	20%	11%	5%	1%	0%	0%	0%	0%	0%	0%
Carlton	0%	1%	1%	20%	0%	0%	0%	0%	0%	3%	7%	10%	28%	25%	17%	7%	2%	1%	0%	0%	0%	0%
Collingwood	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	1%	4%	16%	72%	6%	1%	0%	0%
Essendon	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	1%	8%	59%	32%	0%	0%
Fremantle	2%	2%	4%	91%	0%	0%	1%	4%	6%	16%	42%	22%	8%	1%	0%	0%	0%	0%	0%	0%	0%	0%
Geelong	4%	2%	6%	96%	0%	1%	1%	8%	26%	35%	16%	9%	3%	1%	0%	0%	0%	0%	0%	0%	0%	0%
Gold Coast	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	2%	33%	60%	3%	2%
GWS	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	4%	10%	86%
Hawthorn	24%	23%	47%	100%	13%	32%	25%	22%	7%	1%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%
Melbourne	0%	0%	0%	6%	0%	0%	0%	0%	0%	0%	0%	6%	17%	28%	29%	15%	5%	0%	0%	0%	0%	0%
North Melbourne	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	2%	10%	34%	37%	13%	2%	2%	0%	0%
Port Adelaide	0%	0%	0%	1%	0%	0%	0%	0%	0%	0%	0%	1%	2%	13%	13%	32%	35%	4%	0%	0%	0%	0%
Richmond	0%	0%	0%	23%	0%	0%	0%	0%	1%	8%	6%	8%	22%	19%	25%	7%	4%	0%	0%	0%	0%	0%
Saint Kilda	21%	21%	42%	100%	24%	21%	30%	16%	6%	2%	1%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%
Sydney	4%	6%	10%	100%	1%	2%	4%	11%	41%	27%	12%	2%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%
West Coast	21%	20%	41%	100%	12%	24%	31%	24%	6%	2%	1%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%
Western Bulldogs	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%	1%	87%	12%

Sydney have a big range of possible end of regular season positions but 5th or 6th is favoured at present.  North and Collingwood will not see September action (really need a big start to the season to overturn the model predicitons).

Too much number crunching for today... Will fire the laptop up again on boxing day or so

Cellar Dwellers

GWS: I do not think it will surprise many that the model predicts that GWS will most probably end up with the wooden spoon.  What is a bit surprising is how poorly Essendon perform - the round 17 clash in their "purple" patch of reasonable form.  So we will have to wait and see.

Round	Home	Away	Prediciton
1	GWS	Sydney	Sydney
2	Port Adelaide	GWS	Port Adelaide
3	GWS	Saint Kilda	Saint Kilda
4	Melbourne	GWS	Melbourne
5	GWS	Gold Coast	Gold Coast
6	Essendon	GWS	Essendon
7	GWS	Adelaide	Adelaide
8	Hawthorn	GWS	Hawthorn
9	GWS	West Coast	West Coast
10	Carlton	GWS	Carlton
11	GWS	Geelong	Geelong
12	GWS	Port Adelaide	Port Adelaide
13
14	North Melbourne	GWS	North Melbourne
15	GWS	Western Bulldogs	GWS
16	Sydney	GWS	Sydney
17	GWS	Essendon	GWS
18	Collingwood	GWS	Collingwood
19	GWS	Melbourne	Melbourne
20	Fremantle	GWS	Fremantle
21	Brisbane	GWS	Brisbane
22	GWS	Richmond	Richmond
23	Gold Coast	GWS	Gold Coast

Gold Coast are tipped to do a bit better than GWS with four wins (GWS, Bulldogs, Collingwood and GWS).

The Bulldogs are predicted to have a horror year which given their poor performance at the end of last year is not surprising as this form carries over into the model for this year.  Perhaps the bulldogs should recruit/ draft Fev - they do need a full forward and even though Fev probably would be carrying a bit too much weight and lacking in form he would draw some crowd and let us face it if the Bulldogs can not get close to the finals and suffer some of the predicted losses I see coming at least they could help their financial position by having a drawcard?

Essendon and Collingwood are bit more of a surprise for the cellar in the model.  Essendon had poor end of season form similar to the Bulldogs so perhaps Essendon in the bottom third is not too surprising - they will need to win the games against Bulldogs and GWS to have a chance of getting into the middle third.  They will also need to work hard to limit the damage to the model predictions with the 5 earmarked thrashings at the hands of the Saints, Geelong, Sydney, West Coast and Hawks (all seemingly a formality at this stage).

Collingwood could be the surprise packet from the modelling perspective (along with Adelaide as mentioned before).  Ten narrow losses and six narrow wins really determine the role the Pies will play in 2013.  The early part of the season impacts on the narrow losses/ wins a great deal.  Run through a range of limited simulations for the first six rounds it is really a toss of the coin which way the Pies go (one horror (to Pies fans: joy to everybody else) is the possibility of 12 losses in a row - that has come up a few times in the simplified Monte Carlo season run: more to come over Christmas when I can run the laptop overnight with the model).

North round out the cellar dwellers and again how the narrow win loss games pan out determine their season.  I would strongly suggest kicking for goal accuracy training over summer for both the Pies and Roos (more so Pies as 12 losses in a row is enticing - I would love to see the front page of the Collingwood Fan Newsletter: sorry Herald Sun mid year going into Round 15 against Carlton "Bucks Blue Day: Malthouse flying high?"

End of Year ladder predictions: two methods

The "first past the post" is what is used for the previous blog discussions and is the easiest of the model variations to use.  The Monte Carlo predictions which are based on the scoring shot model is more time consuming and is used in a simple form for the flexi bet option determinations.  The simple Monte Carlo distribution model works on normal score distributions rather than a point and goal distribution model (which I use for the last few weeks of the home and away and occasionally for the finals if I have the time/ determination and can be bothered to let the laptop run for a few hours crunching the numbers).  The win probabilities are added together round by round and give a narrower distribution - the result highlights the number of tight games that Adelaide are predicted to have throughout the year and bumps the Saints to top spot and gives the Hawks a home final during the first week of the finals.

Which are the easy rounds to predict in 2013?

At the bgining of the year the model is predicting that rounds 10, 4, 9, 6 and 14 are the easiest to predict the most winners.  Rounds 15, 23, 12, 17 and 11 (two of these in the bye rounds) are the hardest.  Of course the long range predictions are harder to call at the start of the season and the relative rankings will shuffle around a bit throughout the year.  Although rounds 3, 8 and 10 are looking good to play the joker in any work tipping competition or to try and win the Whopper burgers in the on line tipping (goal this year is to win more than 3 burgers with hopefully 5 perfect rounds?)