Simulating the Finalists for 2014 : After Round 22

Normally I wouldn't bother simulating just a single round prior to the Finals, but this year it seemed like it might be an interesting exercise - even if for no other reason than to estimate just how likely it is that the Tigers will finish 9th.

I've used as the expected game margin in each contest the negative of the TAB Bookmaker handicap in the associated Line Betting market, which means that:

  • Collingwood is expected to go down to the Hawks by 48.5 points
  • Carlton is expected to go down to the Dons by 18.5 points
  • Fremantle is expected to beat Port Adelaide by 13.5 points
  • Sydney is expected to beat the Tigers by 22.5 points
  • The Kangaroos are expected to beat the Dees by 44.5 points
  • Geelong is expected to beat the Lions by 42.5 points
  • Gold Coast is expected to go down to the Eagles by 17.5 points
  • Adelaide is expected to beat the Saints by 56.5 points
  • The Western Bulldogs are expected to beat the Giants by 24.5 points

I've used these expected margins as inputs for the same process that I've been using since the end of Round 15, the outputs from 100,000 simulations of which are shown below, first in the form of manhattan-style plots for each team's projected final ladder position.

Only three teams have five possible ladder finishes in all but a vanishingly small number of simulated seasons:

  • Adelaide: 8th through 12th
  • Richmond: 7th through 11th (with a most-likely finish of 10th)
  • West Coast: 8th through 12th

The Roos and GWS are the only teams virtually assured of a particular ladder finish, the Roos all but locks for 6th and the Giants near-certainties for 16th. Other teams with high levels of certainty surrounding a particular finish are the Lions (15th), the Pies (11th), the Dons (7th), the Cats (3rd), the Hawks (2nd), the Dees (17th), the Saints (18th), and the Swans (1st).

The finer detail for each team can be gleaned from the heat-map below.

On the TAB, the only relevant Futures market still on offer is for the Top 8 and, based on these latest simulations, no team represents value in that market.

Lastly, let's look at Top 2s, Top 4s and Top 8s. 

The simulations suggest that a Sydney-Hawks finish is now about an 83% prospect, and a Hawks-Sydney finish a 9% prospect, only slightly more likely than a Sydney-Cats finish. A Hawks-Sydney finish requires that the Hawks win by a sufficient margin to boost their percentage above Sydney's, and that Sydney lose; a Sydney-Cats finish requires that Sydney win, the Hawks lose or draw, and the Cats win.

Only 10 Top 4s are now feasible according to the simulations, the most likely of which sees the final ordering as Sydney-Hawthorn-Geelong-Fremantle, which is about an even-money bet. The next-most likely ordering is Sydney-Hawthorn-Geelong-Port Adelaide, which has roughly a 30% probability. No other ordering has greater than a 7% probability.

Across the 100,000 simulations, 58 different Final 8 orderings appeared at least once, but the 10 most common orderings, shown here, account for almost 85% of the results.

The most likely Final 8, with odds of about 3/1, is:

  1. Sydney
  2. Hawthorn
  3. Geelong
  4. Fremantle
  5. Port Adelaide
  6. Kangaroos
  7. Essendon
  8. West Coast

The next-most likely Final 8, which is about a 6/1 proposition, swaps Port Adelaide and Fremantle. About equally likely is the ordering that has the same teams in positions 1 through 7 as in the most likely ordering but which swaps in Adelaide for West Coast in 8th.