Who'll Be There in September (Maybe)?

With just 6 rounds to go in the competition, a few articles are emerging about how the teams will fare over the remainder of the season and, most importantly, who'll finish where in the top 8.

Using the MARS Ratings from MAFL Stats I've also predicted the results for the remainder of the season, details of which appear below:

On the right of the table I've provided the game-by-game projections for each team. Green boxes denote victories, red losses, and a bracketed "H", "A" or "N" denote Home, Away and Neutral Ground venues respectively.

If these projections are accurate, there's not a lot of change that we'll see amongst the top 8. In fact, we'll see no change at all in terms of its membership and only a couple of significant changes in terms of its ordering.

Carlton, I'm forecasting, will leap 3 places and secure a top 4 spot, beating Collingwood (currently 4th), the Roos (12th), Port (9th), Melbourne (15th), and Adelaide (6th) in so doing, losing only to the 2nd-placed Geelong in Round 19. This string of results will draw Carlton level with the Lions on 56 points, but will see them grab 4th with a superior percentage.

Collingwood will surrender their top 4 spot, dropping to 6th due to losses to Carlton (7th), Adelaide (6th), and the Dogs (3rd). Victories over the Lions (5th), Richmond (14th), and Sydney (11th) won't be enough.

Port Adelaide and Hawthorn are projected to finish level with the Dons on 44 points but will miss a spot in the eight on percentages, in Hawthorn's case despite finishing the season with a three-game winning streak.

ELO Projected Ladder II

The last three weeks has had quite an effect on ELO's projected end of season ladder, as you can see in the table below.

The top 3 positions are unchanged, firmly held by the Cats, Saints and Dogs, but there's significant movement amongst the next 8.

Carlton moves from the fringes of the top 8 into 4th, dethroning the Pies who drop to 6th. The Lions and Sydney - both now projected to win two more games than thought previously - move into the top 8 at the expense of the two Adelaide teams, who both now miss out on a finals spot based on (my proxy of) percentages.

Making the final 8 now requires 11 wins, which is more in keeping with seasons past than the 10 wins that were projected previously. Also, the teams that make up the projected final 8 are the same teams that currently occupy the top 8 teams on the competition ladder.

Positions 11 through 16 are all held by the same teams as in the earlier projections, albeit that there's been some interesting but inconsequential rearrangements. Unrearranged though is Melbourne, now projected to lose all of its remaining games.

Projecting the Final Eight

This year I've adapted the ELO-based MARS Rating system so that it now provides a margin of victory as well as a tip for each game. Based on the current ratings, these margins and tips can be generated for all the remaining home and away matches and these, in turn, can be combined with the existing table to arrive at a projected end of season table.

Well, I've just done this and the results appear below.

(Because MARS provides a margin but not a score for each game, I've had to use the difference between points for and points against - rather than percentage - as my tiebreaker when teams are equal on competition points. Generally the two approaches will produce the same ordering, but not always.)

Our projections have Hawthorn and Adelaide slipping into the top 8 at the expense of the Lions and the Dons.

Adelaide is a particularly interesting case to look at. It is projected to finish 7th despite currently being 12th on the ladder and ranked only 10th on current MARS Ratings. So, how can it be projected to climb 5 ladder positions? It's on the strength of its easier draw for the remainder of the season.

Collingwood is another team that appears to benefit from its draw according to MARS Ratings, though less so than does Adelaide. The Pies are projected to finish 4th despite currently being 8th on the ladder and ranked only 5th by MARS.

In contrast to Adelaide and Collingwood, Carlton suffers, to a small degree, from a difficult remaining draw, evidenced by the fact that its projected final ladder position is below both its current ladder position and its current MARS Ranking. All other teams are projected to finish in a position that is between (or equal to) their current ladder position and current MARS Ranking.

One other notable feature of the projected ladder is that it takes only 10 wins to make the final 8 whereas, historically, it has usually taken 12 or 13 wins. This is due to the projected dominance of Geelong and St Kilda; their combined projected 43 wins doesn't leave many points to go around.