Matter of Stats

View Original

2021 : Simulating the Final Ladder After Round 16

The latest simulation results come with something of an asterisk given the uncertainly still surrounding the fixturing for the remainder of the season. For the purposes of these simulations, I’ve assumed that all remaining matches are played at the venues as currently set down, although I realise that this mightn’t be true even for the Round 17 games.

That said, the likely impacts of any variation to the gazetted fixturing is likely to be small, and the competition is at such an interesting point, that I think some simulations are better than none at all.

(I’ll update it, time permitting later in the week once the venues for Round 17 are locked in)

So, here goes …

Based on the latest Standard Methodology simulations, we still have 10 teams that have a better than 1-in-3 shot at playing Finals, and eight that have about a 1-in-2 or better shot. There are also five teams that have roughly a 1-in-2 shot or better at the Top 4.

Under the Heretical Methodology, there are also 10 teams that have a better than 1-in-3 shot at playing Finals, but only six that have about a 1-in-2 or better shot. There are also five teams that have a roughly 1-in-2 shot or better at the Top 4.

(For details about the methodologies I’ve used, see this earlier blog. Note that I have assumed no change in fixturing for the remainder of the season except for switching the future Swans v Saints game to be a home game for the Saints)

LADDER FINISHES

Team Expected Wins under the Standard Methodology grew by 0.9 wins this week for GWS, and fell by the same amount for Richmond and Melbourne.

The only significant moves in terms of Finals chances were GWS’s increases of around 27% points, Sydney’s increase of around 19% points, Richmond’s decrease by about 26% points, and West Coast’s decrease by about 18% points. Melbourne saw their Top 4 chances decrease by about 12% points to 83%.

Also, according to both the Standard and the Heretical Methodology, the Dogs are around 60% to 65% chances for the Minor Premiership now, while the Dees and Lions are around 15% chances.

(Note that the comparative moves in the table for the Heretical simulations below is for Round 14 versus Round 16, while that for the Standard simulations in for Round 15 versus Round 16).

STANDARD METHODOLOGY

HERETICAL METHODOLOGY

The range of Expected Wins using the Standard Methodology across the 18 teams moved to about 12.1 wins this week, while using the Heretical Methodology it came in at 13.1 wins.

The animation below shows how each team’s ladder probabilities have varied across the season so far.

Across the 50,000 Standard Methodology simulations there are 18,143 different sets of teams playing Finals, treating different orderings of the same eight teams as a different set. No ordering appears in more than 86 replicates. If we ignore ordering, there are still 203 different sets of eight Finalists.

Amongst the teams currently in the Top 8, the estimated probability that any 4 through 8 of them play Finals is:

  • All 8: 39%

  • Only 7: 53%

  • Only 6: 8%

  • Only 5: 0.1%

Put another way, there’s about a 61% estimated chance that the teams playing Finals won’t be the current Top 8 teams.

TEAM AND POSITION CONCENTRATION

The HHI figures for the most recent Standard Methodology simulation replicates appear below.

(For information about the HHI, also see that previous blog linked to earlier).

We see a reduction of over one full spot in final ladder position uncertainty this week only for Hawthorn and Sydney, and an increase of over one full spot for Gold Coast and Melbourne.

Carlton, St Kilda, Essendon, and Fremantle are the teams with greatest uncertainty about the positions in which they’ll finish on the ladder, each of them effectively competing now for about 9 ladder spots. North Melbourne, Western Bulldogs, Sydney, and Hawthorn have least uncertainty, and are all effectively competing for only about 1.5 to 4 spots.

After 16 rounds, the average team is still effectively competing for about 6 different ladder positions.

Last week, unusually, saw a range of increases and decreases in the uncertainty associated with various ladder positions, with the largest increase for 2nd, and the largest decrease for 6th.

Positions 7th through 14th are now the ones that are associated with most uncertainty, and some of the top and tail positions - 1st through 6th, and 17th and 18th - with least uncertainty.

There are now, effectively, only about 2.4 teams vying for the Minor Premiership, and just 1.7 battling for the Spoon.

IMPORTANCE OF PERCENTAGE

The Standard Methodology simulations suggest that percentages are relatively unlikely to decide key ladder positions.

There is now an estimated 39% chance that 8th and 9th will finish equal on competition points, a 14% chance that 8th and 10th will, and a 4% chance that 8th and 11th will.

There’s also still about a 32% chance that the double-chance in the Finals will be determined by percentage.

WINS TO MAKE The 8 AND THE 4

Based on the Standard Methodology, the estimates of team chances of playing Finals should they finish on only 11 wins now mostly lie in the 20% to 30% range, although they are a little lower for Fremantle, Adelaide, and St Kilda, and a little higher for Brisbane Lions, Port Adelaide, and Richmond. Team chances still also mostly lie in the 75% to 90% range should a team record 12 wins, although they are also a little lower for Fremantle, St Kilda and Adelaide, and higher for Western Bulldogs, Brisbane Lions, Port Adelaide, and Richmond.

Looking next at making the Top 4, we see that teams’ chances range from about 20% to 60% with 14 wins, and 65% to 95% with 15 wins, with the Lions and Dogs still faring best, and Sydney and West Coast still faring worst.

HOW MANY WINS TO FINISH WHERE

Now a look at what the Standard Methodology shows for the joint distribution of number of wins and final ladder position for each team.

WINS FOR A PARTICULAR LADDER POSITION

And, finally, an update on the distribution of wins associated with each ladder position.

It’s now the case that 11 wins rather than 12 is more likely to be associated with the team finishing 8th, but 15 wins is still more likely than 14 or 16 wins to be associated with the team finishing 4th.

GAME IMPORTANCE

Finally, let’s take a look at how the Standard Methodology estimates the importance of each of the remaining 62 games (see this blog for details about how these are calculated).

This week, again, I’m going to present the importance information by showing the 10 most important games for each team (except for North Melbourne and Western Bulldogs, for whom no single game result can budge their finals chances from 0 and 1 respectively)

Round 17, then, includes:

  • Likely very important games for: Essendon, Carlton, Fremantle, GWS, Richmond, and West Coast

  • Likely moderately important games for: St Kilda and Sydney