Matter of Stats

View Original

2023 : Simulating the Final Ladder After Round 14

Tony Corke

There are now nine teams with about a 1-in-2 or better chance of playing Finals, eight teams with about a 1-in-9 or better chance of finishing Top 4, and three teams with about a 1-in-9 or better chance of finishing as Minor Premiers.

Those simulations suggest that the teams can now be roughly grouped as follows:

  1. Port Adelaide: almost certain of being finalists; 95% chance of finishing Top 4; and 50% chance of being Minor Premier

  2. Collingwood: almost certain of being finalists; 85 to 90% chance of finishing Top 4; and 30% chance of being Minor Premier

  3. Melbourne and Brisbane Lions: almost certain of being finalists; 70 to 75% chance of finishing Top 4; and about 10% chance of being Minor Premier

  4. Essendon and Western Bulldogs: about 70% chance of being finalists; about 15 to 20% chance of finishing Top 4; and very small chance of being Minor Premier

  5. St Kilda and Adelaide: about 60% chance of being finalists; 10 to 15% chance of finishing Top 4; and very small chance of being Minor Premier

  6. Geelong: about 50% chance of being finalists; about 8% chance of finishing Top 4; and virtually no chance of being Minor Premier

  7. Richmond: about 33% chance of being finalists; about 2% chance of finishing Top 4; and virtually no chance of being Minor Premier

  8. Gold Coast, Carlton, Fremantle and Sydney: about 10 to 20% chance of being a finalist; about 0.5 to 1.5% chance of finishing Top 4; and no chance of being Minor Premier

  9. GWS: about 7% chance of being a finalist and virtually no chance of finishing Top 4 or being Minor Premier

  10. Hawthorn, North Melbourne, and West Coast: little to no chance of being finalists, finishing Top 4. or being Minor Premier

(For anyone who is curious about the precise methodology used for these simulations, please refer to this blog post over on the General Probability and Statistics journal)

LADDER FINISHES

The ladder projections using the Heretical outputs appear below.

We see only small changes in Expected Win values again this week, the highest being +0.9 wins for GWS, and the lowest -1.0 wins for Fremantle.

Richmond enjoyed the largest percentage point increases in their estimated finals chances and Fremantle suffered the largest decline, but there were also significant drops for St Kilda and Gold Coast.

The range of expected wins now runs from 2.1 to 17.9 (from 2.1 to 17.4 at the end of Round 13).

TEAM AND POSITION CONCENTRATION

The HHI figures for these pre-season simulation replicates appear below. These provide us with estimates of:

  • How many ladder positions a team is currently effectively fighting for

  • How many teams are currently effectively fighting for a given ladder position

Round 14 results reduced just a little more of the uncertainty about the final home and away ladder, with that around teams falling by about 0.2 positions per team, and that around ladder positions falling by about 0.3 teams per position. Just over half of the teams are now currently effectively fighting for between 9 and 12 ladder positions, and half the ladder positions have 9.5 to 11.5 teams effectively fighting for them.

On the teams side, those with least uncertainty are West Coast, North Melbourne, and Hawthorn and, on the ladder positions side, those with least uncertainty are still 1st, 16th, 17th, and 18th.

We see that the ladder positions with the most uncertainty remain those in the middle of the table, from about 5th to 13th.

We also see that the average team is effectively competing for about 7.6 positions, and the average ladder position has about 7.6 teams competing for it.

WINS AND LADDER POSITION

The charts below investigate the relationships between ladder position and number of wins.

Firstly, we show the relationships by team, which illustrate the comments from the previous section about the teams with least uncertainty about their final home-and-away ladder finishes (viz West Coast, Hawthorn, and North Melbourne). Contrast the results for those teams with, say, those for Geelong, St Kilda, Adelaide, Richmond and Gold Coast, which are the teams with the most uncertainty according to the HHI.

In the next chart we look at the probability that each team plays Finals given that it finishes with a specified number of wins.

We see that many teams are still only about 25% chances of playing Finals if they finish on 12 wins, although the figure is closer to 50% for Adelaide, Geelong, Melbourne, and Sydney, We also see that teams have mostly better than 75% chances of playing Finals if they finish on 13 wins.

Next, let’s look at what’s required for a Top 4 finish.

Here we see more variability across the teams but find that 15 wins will still make most teams about even-money chances or better to finish Top 4. With 16 wins, the probabilities are all near to or over 75%.

Next, we show the relationships between ladder position and wins after aggregating across teams.

Based on the simulations, we can say that:

  • For teams finishing 8th in a replicate: 35% had 12 wins (up 8%), 36% had 13 wins (down 10%), and 16% had 12.5 wins (up 5%). Also, 3% had 14 wins (down 5%), and 4% had 13.5 wins (no change)

  • For teams finishing 4th in a replicate: 43% had 15 wins (down 1%), 26% had 16 wins (down 1%), and 4% had 15.5 wins (down 1%). Also 15% had 14 wins (up 1%), and 5% had 14.5 wins (no change)

  • For teams finishing win 11 wins: 2% made Finals (up 1%)

  • For teams finishing win 12 wins: 33% made Finals (up 8%)

  • For teams finishing win 13 wins: 89% made Finals (up 9%)

  • For teams finishing win 14 wins: almost 100% made Finals (up 1%)

In short, it seems that more teams might be able to progress to Finals with fewer wins after this week’s results.

LIKELY IMPORTANCE OF PERCENTAGE

Next, we’ll use the simulations to investigate the estimated likelihood that any given pair of ladder positions will be determined after Round 24 based on percentage.

So, our new estimates are that:

  • There is now slightly less than a 2-in-5 chance that 8th will be decided on percentages

  • There is now a slightly lower than 1-in-3 chance that 4th will be decided on percentages

  • There’s is now slightly less than a 1-in-5 chance that 1st will be decided on percentages

GAME IMPORTANCE

Here is an updated list of the Top 10 most important remaining games for each team. Note that not all teams still have 10 games left to play, so matches involving other teams will make it onto their list.

(For details on how this metric is calculated, see last week’s blog)

From that table we can see that the Round 15 games are likely to be of moderate importance to the Finals chances of Adelaide, Essendon, Geelong, and St Kilda, and also likely to be somewhat important to the chances of Brisbane Lions, Fremantle, and Gold Coast. They are likely to be of little to no importance to the chances of every other team.

FINALS CHANCES

Finally we take another look at team’s chances all the way through to the Grand Final, making reasonable assumptions about where Finals will be played given the participants and week of the Finals.

We now have three near equal favourites for the Flag with Melbourne, Colingwood, and Port Adelaide all assessed as being about 1-in-5 chances. Thereafter comes Brisbane Lions at 1-in-9, Geelong at 1-in-16, and then five teams all with chances of around 1-in-40 to 1-in-25.