The Brownlow Guru speaks - 2010

[BrownlowGuru]

Hi all. I've got a system which I use to calculate Brownlow Medal probabilities. In a nutshell, it involves analysing all 176 games, determining the probability of individuals getting votes in that game, and then running a number of simulations for the whole season to get an idea of who's likely to win, place, top the club etc.

I've been doing this for a few years now, and had reasonable success punting with this (including posting here last 3 years). Here are my predictions for win and place, I'll supply some of the exotics (team, quinella, H2H) later.

Below are what I have calculated to be the real odds - if you can find odds greater than those displayed (eg Goddard, JSelwood) then they are worth a punt if you are so inclined.

WIN
Swan D (Coll) 2.218
Ablett G (Geel) 6.328
Goddard B (St K) 11.057
Judd C (Carl) 15.476
Hodge L (Haw) 17.131
Selwood J (Geel) 21.242
Pendlebury S (Coll) 25.306
Hayes L (St K) 61.617
Boyd M (WB) 63.225
Goodes A (Syd) 90.156
Dal Santo N (St K) 96.125
Harvey B (Kang) 137.93
Sandlilands A (Fre) 153.85
Montagna L (St K) 164.00
Mitchell S (Haw) 257.83
Bartel J (Geel) 328.49
Barlow M (Fre) 525.80
Thompson S (Adel) 583.26
Cross D (WB) 589.56
Cooney A (WB) 893.38
All others > 1000

PLACE (top 3)
Swan D (Coll) 1.358
Ablett G (Geel) 2.184
Goddard B (St K) 3.179
Judd C (Carl) 3.990
Hodge L (Haw) 4.046
Selwood J (Geel) 4.949
Pendlebury S (Coll) 6.217
Hayes L (St K) 10.885
Boyd M (WB) 11.464
Goodes A (Syd) 13.903
Dal Santo N (St K) 15.448
Harvey B (Kang) 20.107
Sandlilands A (Fre) 21.710
Montagna L (St K) 23.660
Mitchell S (Haw) 36.202
Bartel J (Geel) 39.438
Barlow M (Fre) 44.37
Thompson S (Adel) 60.15
Cross D (WB) 65.89
Cooney A (WB) 82.76
Swallow A (Kang) 109.47
Murphy M (Carl) 185.01
Boak T (Pt A) 197.68
All Others > 200

PLACE (top 5)
Swan D (Coll) 1.179
Ablett G (Geel) 1.579
Goddard B (St K) 2.061
Hodge L (Haw) 2.383
Judd C (Carl) 2.410
Selwood J (Geel) 2.852
Pendlebury S (Coll) 3.405
Hayes L (St K) 5.300
Boyd M (WB) 5.415
Goodes A (Syd) 6.256
Dal Santo N (St K) 7.108
Harvey B (Kang) 8.551
Sandlilands A (Fre) 9.059
Montagna L (St K) 9.711
Bartel J (Geel) 14.288
Mitchell S (Haw) 14.452
Barlow M (Fre) 14.470
Thompson S (Adel) 20.483
Cross D (WB) 24.165
Cooney A (WB) 26.815
Swallow A (Kang) 31.724
Boak T (Pt A) 54.946
Murphy M (Carl) 58.619
Lake B (WB) 69.317
Didak A (Coll) 69.901
Watson J (Ess) 91.007
All others > 100

 

[Prem87]

Who were your predicted winners in previous years?

 

[half_back_flank]

Cheers BG for that.

Last year was an anomaly for Swan you feel, i remember you had him favourite last year as well.

 

[Libo]

Who were your predicted winners in previous years?

2007: Bartel
2008: Bartel (Cooney was fourth favourite and the second best value behind Bartel)
2009: Swan (Ablett was second favourite)

 

[Lensen]

Great work as always.

Looking forward to the team markets.

 

[Dawson]

2007: Bartel
2008: Bartel (Cooney was fourth favourite and the second best value behind Bartel)
2009: Swan (Ablett was second favourite)

So a 33% strike rate

 

[Lensen]

So a 33% strike rate

Or, to look at it another way, around a 333% return given Bartel was paying about $10 the year he won it.

 

[Badesumofu]

So a 33% strike rate

No, not at all. He's trying to assess probabilities, nothing so crude as simply picking a winner. He's using a subjective notion of probability as a measure of ignorance, so it's quite hard to assess how 'right' he is. It's not accurate to say that if Dane Swan does not win the Brownlow, then Guru was wrong in his prediction.

Over a longer period, it would be possible to do a rough assessment of how accurate the odds he's calculated have been.

Some of those placing odds look pretty appealing to me, might have to pop on to Betfair.

 

[Dr Awkward]

Nice one. I prefer to bet in the teams markets or those special head to head markets that they put out. Any chance you will be letting us know of your analysis of odds for top vote getter for each team?

 

[twarby]

Brownlow King >> Brownlow Guru

 

[summerthebat]

It's a system based on probability/value, it's the only way you can ever win gambling.

This system get's my THUMBS UP (y)

In saying that, brownlow is ridiculous to wager on (with a professional attitude) considering you only get to refine your system once a year, and maybe have 50 chances at it in your entire lifetime.

 

[FRUMPY]

so like everyone else and their dog you have chosen swan.......big news.

 

[Ron The Bear]

Thanks for posting, even though some just want to see

1 Swan
2 Ablett
3 Goddard

 

[jayjones1]

good luck

 

[Eth]

Well, my tip's Ablett and Swan to tie...Goddard and Hodge next. I had Hodge on 28 too, pretty high polling from what I've gotten.

Nice system BG. Looks pretty complex

 

[BrownlowGuru]

Last year was an anomaly for Swan you feel, i remember you had him favourite last year as well.

Yeah I had Swan last year, and he polled really poorly. The model has him as a solid favourite this year and I am pretty comfortable with that. He has had a better year statistically, and (apart from last year) has generally polled very well.

To be honest, this year is the closest that the model has been to the bookies odds for the top contenders. However there are plenty of differences in the teams and groups, which I will post shortly.

 

[BrownlowGuru]

Clubs Votes:

Adel:
Thompson S 1.653
Douglas R 4.658
Johncock G 8.811
Goodwin S 35.22
Tippett K 46.26
Vince B 80.36
Others > 400

Bris:
Brown J 2.378
Rischitelli M 3.971
Black S 5.461
Brennan J 12.20
Power L 22.52
Rich D 99.00
Others > 400

Carl:
Judd C 1.215
Murphy M 11.77
Simpson K 23.51
Scotland H 32.24
Betts E 129.5
Gibbs B 156.9
Others > 500

Coll:
Swan D 1.186
Pendlebury S 6.913
Didak A 114.2
Thomas D 342.8
Others > 5000

Ess:
Watson J 1.720
Stanton B 6.208
Fletcher D 9.907
Winderlich J 19.91
Prismall B 38.23
Lovett-Murray N 39.42
Hille D 41.74
Jetta L 61.58
Ryder P 91.13
Others > 500

Freo:
Sandlilands A 2.186
Barlow M 2.457
Pavlich M 18.29
Mundy D 30.18
Duffield P 39.83
Hill S 58.10
Others > 200

Geel:
Ablett G 1.828
Selwood J 4.178
Chapman P 5.923
Bartel J 23.63
Others > 500

Haw:
Hodge L 1.416
Mitchell S 6.413
Franklin L 12.93
Rioli C 30.44
Lewis J 43.43
Others > 200

North:
Harvey B 1.726
Swallow A 3.168
Rawlings B 12.77
Wells D 40.49
Others > 1000

Melb:
Sylvia C 2.604
Green B 3.202
McDonald J 14.33
Frawley J 15.93
Bruce C 16.84
Scully T 37.06
Moloney B 37.11
Davey A 40.35
Jamar M 67.53
Others > 100

Port:
Boak T 2.372
Cornes K 5.100
Cassisi D 5.446
Pearce D 10.52
Brogan D 17.67
Rodan D 43.83
Schultz J 74.87
Others > 100

Rich:
Deledio B 3.306
Riewoldt J 3.915
Tuck S 5.358
Newman C 5.756
Martin D 24.44
Jackson D 60.32
Cousins B 95.70
Others > 100

StK:
Goddard B 1.856
Hayes L 5.127
Dal Santo N 6.564
Montagna L 8.922
Others > 1000

Syd:
Goodes A 1.752
Jack K 5.878
Malceski N 10.63
Bolton J 15.90
O'Keefe R 20.11
Mumford S 32.90
McVeigh J 89.55
Others > 100

WB:
Boyd M 2.087
Cooney A 5.686
Cross D 6.012
Lake B 11.82
Giansiracusa D 23.82
Griffen R 36.26
Hall B 43.49
Others > 1000

WCE:
Priddis M 1.899
LeCras M 8.833
Cox D 10.23
Waters B 14.12
Embley A 14.60
Selwood A 16.89
Kennedy J 30.52
Naitanui N 79.45
Others > 100

 

[Glenno23]

I reckon Thompson is value to lead the Crows

Had a great 2nd half of the season where we won a few more games and always polls well

Wonder what odds are out there for him and if they are bigger than yours BG

 

[ThomasR]

Clubs Votes:

Bris:
Rischitelli M 3.971

Freo:
Barlow M 2.457

North:
Swallow A 3.168

Melb:
Sylvia C 2.604

Port:
Boak T 2.372

Rich:
Tuck S 5.358

StK:
Dal Santo N 6.564

WB:
Boyd M 2.087

I think there's a fair bit of value in these players. Good posting

 

[DanA]

Wish you were my bookie

Pendlebury to win browlow 6.9
Pendlebury to lead club 6.9
hmmm which bet is more attractive?

$5 for Hayes to lead st Kilda in the brownlow?
That's rediculous

In fact almost all your second favourites for the club must be miles out.

 

[Brave Sir Robin]

Wish you were my bookie

Pendlebury to win browlow 6.9
Pendlebury to lead club 6.9
hmmm which bet is more attractive?

$5 for Hayes to lead st Kilda in the brownlow?
That's rediculous

In fact almost all your second favourites for the club must be miles out.

Geez, you're gonna be embarrassed when you read over the odds he posted again and realise you goofed up.

 

[ManWithNoName]

Wish you were my bookie

Pendlebury to win browlow 6.9
Pendlebury to lead club 6.9
hmmm which bet is more attractive?

$5 for Hayes to lead st Kilda in the brownlow?
That's rediculous

In fact almost all your second favourites for the club must be miles out.

He's not giving odds. He's saying if you find those players at odds higher than he's listed, that's value. If they're below his odds listed then it's not worth it.

Every year people don't understand how his system works, and every year people whinge when they lose money as a result.

 

[asanque]

Not much value out there this year.

The bookies typically offering odds well under the predicted odds.

 

[Brave Sir Robin]

He's not giving odds. He's saying if you find those players at odds higher than he's listed, that's value. If they're below his odds listed then it's not worth it.

Every year people don't understand how his system works, and every year people whinge when they lose money as a result.

No, the guy had a point. If Pendlebury was the same probability for his club as he was for the medal, then you would be right to question how well this model has been developed. Where he was wrong is that it isn't what happened.

 

[BrownlowGuru]

Not much value out there this year.

The bookies typically offering odds well under the predicted odds.

Wait until you see the groups.