©2018 by Fat Stats Blog. Proudly created with Wix.com

Search
  • Liam Webb

Why Patrick Cripps can (probably) not win the Brownlow in 2018.

There has been a lot of deserved hype about Patrick Cripps this year. As of round 19, he leads the league in contested disposals, he is third in clearances and along with Charlie Curnow has been a shining light in a dismal Carlton season. Unfortunately for him, he may just be the best player in the league this year, but he would need to defy history and numbers to win the Brownlow medal. At the time of writing, Cripps is widely second favourite with bookies at around 7-8 dollars, behind Hawthorn ball magnet Tom Mitchell. Our brownlow predictive model chaRlie has Cripps 24th with an estimated 12 votes. Even the most optimistic of its model iterations, also allowing for the small historical negative bias the model has given him, has him getting a very generous 18 votes. If the season were to finish here, it seems possible he wouldn’t make the top 10, let alone potentially win it. So what is causing this discrepancy between statistics and the media/punters? (example). There is one major reason - Carlton’s season has been historically bad. At the time of writing, they have only won two games - incredibly against Essendon in round 8, and not as incredibly against the also struggling Gold Coast Suns in round 19. You have to look back to the Suns and Giants in their inaugural seasons or the pre-merger Fitzroy Lions in the mid-nineties to find a team as disappointing as the 2018 Blues.


Unfortunately for Cripps, teams in losing sides have a much lower historical chance of obtaining votes. Since 2005, players in losing sides have only received 2409 votes out of a total of the 14580 available (19.7%). Of those votes, 37% of single vote allocations go to a player on the losing side, 27.7% of two votes and only 10.5% of three votes. These statistics include games where the score was close. Carlton has had many games this year where they were not close. To start off on a positive note however, it’s worth looking at the games where Carlton actually won.


Round 8 v Essendon

chaRlie calculated game rankings (probabilities) for Carlton v Essendon, R8, 2018

Cripps was good, but was he good enough to get a vote? chaRlie has Cripps as 4th best on field, behind Matthew Kreuzer, Ed Curnow and Dale Thomas. The official AFL match report has Cripps best on ground, but its a statistically contentious summary. Ed Curnow had 24 touches, 13 tackles, 5 inside 50s and two goals. The only statistical category where Cripps is significantly superior is contested disposals (a game high 19, the same amount as Kreuzer). Kreuzer added 31 hitouts, 8 clearances and 7 tackles himself to have a stand out game. Dale Thomas, the third ranked player in chaRlie had 23 uncontested disposals including 9 rebound 50s, 14 marks and 11 one percenters to also put pressure on Cripps. It's possible that Cripps might get three votes in this game, but he could easily get 0 or 1 as well.


Round 19 v Gold Coast

chaRlie calculated game rankings (probabilities) for Carlton v Gold Coast, R19, 2018

There were two stand out performers in this game - Cripps and captain Mark Murphy. Most commentators agree with chaRlie that Cripps will get the three, but even in an upset there is daylight between those two and the third ranked player so a minimum of two votes seems a lock.

Losing Games: The official AFL game predictor currently has Patrick Cripps winning the Brownlow on 21 votes and has given Cripps 4 three vote games so far - Rounds 1, 5, 6 and 10. Carlton lost those games by 26,10,21 and 28 points respectively. As a rule of thumb, the bigger the margin, the lower the chance of a player on the losing side getting 3 votes.

Patrick Cripps official AFL brownlow predictor as of round 21 (equal first with Tom Mitchell)

At 10-20 points the historical likelihood of a player getting any votes on the losing team drops to 14.4%, and for losses between 20-30 points that probability drops again to 8.3%. Based on these historical statistics, the overall likelihood of Cripps achieving what the AFL predictor suggests is approximately 1 in 12000. The AFL predictor also gives Cripps 2 votes in Carlton's 64 point loss to St Kilda in R17. Cripps had a brilliant game - 35 disposals, 23 contested possessions and 10 clearances. Despite this, he is going to have to defy history to get votes; less than 1% of votes have been allocated to the losing team in losses of greater than 60 points.


Percentage of votes given to a player in the losing team.

What about Gary Ablett in 2013? Quite often Gazza's successful Brownlow campaign of 2013 is brought up as analogous to Cripp's this year. There is one significant difference - Gold Coast won 8 games and finished 14th in 2013. In those 8 wins, Gary Ablett polled 19 votes out of his eventual 28 votes. Even the greatest ever vote-getter in losing teams only averages just under half a vote a game. To put this in perspective, using Gary's average in Carlton's season would leave him with at most 17 votes. In addition, you also can’t underestimate the vote attracting power of arguably the greatest player of the modern game dragging a team of kids through a season on his back. Cripps is a jet, but he isn’t Gary Ablett Jnr just yet.


How many votes can he get?


The above plot shows chaRlie’s round by round predictions by round for Cripps. chaRlie predicts votes in rounds 1, 5 and 6, however unlike the AFL chaRlie does not predict 3 votes in any of these rounds. In the case of round 1 it seems more likely he will get a single vote. In round 10 ( also a 3 vote prediction from the AFL) Cripps doesn’t get votes from chaRlie in any of the 100 iterations. chaRlie has Cripps with a chance of votes in 7 games, with his best chances in the Gold Coast victory in round 19 and an average of 2.4 votes in the 20 point loss to Collingwood (Cripps had 28 disposals, 19 contested and 3 goals).


There is no doubt that Cripps has had an outstanding year, which will likely be reflected in the coaches voting and the Carlton best and fairest. If Carlton wins their remaining two games against the Bulldogs and Adelaide, then Cripps may scrape into the 15-20 vote range. Unfortunately for Patrick, history has taught us that the umpires are much more likely to give votes to players on the winning team, seemingly regardless of influence on the game. In the probable event they don’t win another game for the year, a total vote count of around 15 seems likely - we have no doubt media hype will contribute some votes that chaRlie misses. Of course, if he is on 10 votes after 6 rounds, we will be deleting this blog post and avoiding anyone who still identifies as a Carlton supporter.

862 views