AFL Round 1 preview and power ratings

If you follow my writing at all – and why wouldn’t you? – you already know which teams I’m likely to tell you to expect to be 1-0 one week from now.

And if it gives you additional comfort to know that the ELO-Following Football rating system I shill for chose the same nine winners, well, so be it. It did.

So, here’s that run-down: Richmond by 42, Collingwood by four, Melbourne by about 20, Adelaide by 15 or so, Sydney by seven to 13 points (depending on who starts on both sides), West Coast by less than whatever number you’re thinking of, St Kilda by more than whatever you’re expecting me to say, then GWS and North Melbourne each by about eight.

But we’ve got more!

We’re heavily into gathering the combined opinions of experts and aficionados alike to give a consensus opinion regarding the ranking of players and teams, and occasionally the outcomes of games as well.

For example, two weeks ago we presented a preliminary average ranking of the 18 teams from the first few dozen sources we’d gathered at the time.

Today, we’re up to 71 predictions, and here’s how they average out.

1. Richmond (average position 1.66)
2. West Coast (average position 3.21)
3. Melbourne (average position 3.33)
4. Collingwood (average position 3.79)
5. Adelaide (average position 5.75)
6. GWS Giants (average position 6.26)
7. Essendon (average position 6.71)
8. Geelong (average position 8.05)
9. Hawthorn (average position 9.63)
10. Sydney (average position 10.02)
11. North Melbourne (average position 10.89)
12. Port Adelaide (average position 11.16)
13. Brisbane (average position 12.26)
14. Western Bulldogs (average position 13.84)
15. Fremantle (average position 14.00)
16. St Kilda (average position 16.15)
17. Carlton (average position 16.28)
18. Gold Coast (average position 17.81)

While not significantly different from our previous totals, there’s more defined stratification today.

For example, while positions two through four are still interchangeable, the next three – Adelaide, GWS and Essendon – have separated themselves from the three champion clubs behind them.

Tom Lynch’s new club top the predicted 2019 table. (Photo by Michael Dodge/Getty Images)

North and Port are virtually tied, Carlton and Brisbane have both drifted upwards, and Gold Coast has a stranglehold on the wooden spoon in the court of public opinion.

But remember, that’s all this is. Opinion.

When the teams start playing this week, they’re all undefeated.

Alongside those meta-predictions, I’ve got something new to share this year.

By combining eight of the top mathematical rating systems for the AFL, and then balancing them around a common mean and range, we can produce a consensus Roar Power Ranking each week on a scale of approximately 0-100, with 50 being the centre point.

Here’s our first RPR for Week Zero, replete with both rating and the average ranking of the seven systems utilized this week (one won’t start until after the first game).

The Roar Power Ranking
1. Richmond (78.98 – average position 1.4)
2. West Coast (74.94 – average position 3.4)
3. Melbourne (74.85 – average position 2.9)
4. Geelong (72.97 – average position 3.7)
5. Collingwood (71.89 – average position 4.6)
6. Greater W Sydney (67.84 – average position 6.7)
7. Adelaide (62.51 – average position 8.6)
8. Hawthorn (62.30 – average position 8.1)
9. Essendon (62.29 – average position 8.4)
10. Sydney (60.72 – average position 8.3)
11. Port Adelaide (57.40 – average position 10.6)
12. North Melbourne (53.88 – average position 11.7)
13. Brisbane (42.43 – average position 13.1)
14. Western (36.70 – average position 14.0)
15. St. Kilda (36.19 – average position 14.7)
16. Fremantle (29.05 – average position 15.9)
17. Gold Coast (14.53 – average position 17.4)
18. Carlton (12.99 – average position 17.6)

I don’t know how to convert those into game predictions yet, except to point out the obvious: Richmond vs Carlton at the MCG should be a rout.

How did Carlton end up below the Suns? Well, four of the seven metrics have them slightly below Gold Coast for some reason, that’s how.

The seven metrics we’re currently blending are Footymaths, Wooden Finger, Massey, plussixone, GRAF, The Arc (if Matt Cowgill is still updating it this season), and our own ELO-Following Football rating system.

These are subject to change if necessary: we’ll use as many mathematically justifiable systems as allow us to use them, and these seven are each excellent.

Once a game’s been played, we’ll add in the teams’ actual percentage (to be technical, we’ll take the logarithm to make it a linear progression like the rest) to the seven listed above.