Team Rating System, update after Finals, Week Three

[Hobbes]

Well, I've been playing with a team rating system on a spreadsheet. I've tried to make this impartial so my own opinions don't come into it. Neither does any actual knowledge of the teams, so I'll be making no attempt to judge how teams will match up against each other, or who's in and who's out. Pretty much the only data I'm inputting is the margins of the matches from this season. Difficulty of opponents is rated by the same method, iterated a few times.

None of the teams seem to take the pre-season competition seriously, so neither do I. Teams' reputations count for nothing. I do take into account home ground advantage, using a very basic algorithm I based on results from a recent season. I assume that intra-state games confer no home-ground advantage for either team, and that matches played in strange places are as if played in the state of the home team.

At the moment, there are only three rounds of data, so I expect my ratings are very rough at the moment - they'll become more accurate as the season progresses. (Ultimately, they will take into account the last 13 matches for each team). To avoid the ratings being largely determined by how much the teams are beating Essendon by, any portion of a margin of victory in excess of 50 is considered to be halved (so an 80 point win is recorded as 65). And I have a moderate opinion that good teams tend to fall over the line more often when it matters, so margins from 1-11 points are modified upward by half of the difference between the actual winning margin and 12.

Anyway, I'll post a team ranking list and a set of predictions, purely based on my ratings (and home ground advantage). As a rule, my predicted resuts will be the scoreline which doesn't alter either team's ranking.

If I don't feel trolled into the ground, I'll try to continue this through the season.

1. Sydney
2. Adelaide
3. Gold Coast
4. Western Bulldogs
5. Geelong
6. North Melbourne
7. Hawthorn
8. Port Adelaide
9. West Coast
10. GWS
11. Melbourne
12. St Kilda
13. Richmond
14. Essendon
15. Collingwood
16. Fremantle
17. Carlton
18. Brisbane

A few of these I don't expect to stay where they are. Gold Coast have had a splendid start to the season, but more than a few people would be surprised if they keep it up. Hawthorn have had only a medium start to the season - I expect them to lift at some point. Fremantle have had a particularly horrid start to the season, but I doubt they'll stay that low for long.

An now, my predictions:

West Coast +19 v Richmond
Geelong +19 v Essendon
Hawthorn +13 v St Kilda
Brisbane v Gold Coast +33
Carlton v Western Bulldogs +26
Adelaide +6 v Sydney
GWS +6 v Port Adelaide
Melbourne +8 v Collingwood
North Melbourne +30 v Fremantle

 

[Dugdale2Dwyer]

good idea, hope it works out.

 

[Kangas_11]

So like an algorithm to determine a kind of power ranking of teams? Roby... have you returned to us?

 

[iBeng]

Somewhere, Roby has sensed this and he is LIVID.

 

[exe ex machina]

Being trolled into the ground is a blessing, not a curse.

 

[Xenthor]

An now, my predictions:

West Coast +19 v Richmond
Geelong +19 v Essendon
Hawthorn +13 v St Kilda
Brisbane v Gold Coast +33
Carlton v Western Bulldogs +26
Adelaide +6 v Sydney
GWS +6 v Port Adelaide
Melbourne +8 v Collingwood
North Melbourne +30 v Fremantle[/QUOTE]
Surely Geelong beats Essendon by more than 19 points. And I would also be extremely impressed if the saints would get that close to the Hawks
Otherwise good job mate

 

[Hobbes]

An now, my predictions:

West Coast +19 v Richmond
Geelong +19 v Essendon
Hawthorn +13 v St Kilda
Brisbane v Gold Coast +33
Carlton v Western Bulldogs +26
Adelaide +6 v Sydney
GWS +6 v Port Adelaide
Melbourne +8 v Collingwood
North Melbourne +30 v Fremantle

Surely Geelong beats Essendon by more than 19 points. And I would also be extremely impressed if the saints would get that close to the Hawks
Otherwise good job mate [/QUOTE]

I'm just following the numbers. If I were making my own tips, I'd agree with you.

In one sense, you could (sort of) double my predicted margins. I'm weighting the last 13 matches on a scheme of 20-18-16-14-12-10-8-6-5-4-3-2-1. This means that, of a possible 119 when a full set of 13 matches is available for data, I'm only working with 54. So, I could multiply by 2.2 to get the magnitude right, but that would tend to overstate the weight of the evidence of recent matches.

Margins are calculated by subtracting one team's rating from the other, and then applying a modifier for home advantage. Last year the top teams tended to hover around a rating of 25-30 (and the bottom teams were a similar negative rating). At this stage, no team has a rating more than + or - 20. Home ground advantage (for interstate draws only) I apply by adding a modifier for each of the teams - Victorian teams 1, NSW 4, Queensland and South Australia 5 and Western Australia 8. So, a Queensland team playing in WA attracts a modifier of -13 for predictions, or +13 for results.

 

[boncer34]

A ratings system with the model outlined clearly?

That's not how its done.

 

[Hobbes]

For a variation, I tried using the same method but incorporating each team's last 13 matches, going back to the middle of last year's minor round.

(If I continue this calculation through the year, it will eventually converge with the other rating)

This left the following rankings:

1. Hawthorn
2. Adelaide
3. Western Bulldogs
4. Sydney
5. West Coast
6. North Melbourne
7. Geelong
8. Port Adelaide
9. Richmond (!!)
10. Gold Coast
11. Fremantle
12. GWS
13. Collingwood
14. Melbourne
15. Brisbane
16. St Kilda
17. Essendon
18. Carlton

And the following tips for this weekend:

West Coast +27 v Richmond
Geelong +40 v Essendon
Hawthorn +52 v St Kilda
Brisbane v Gold Coast +17
Carlton v Western Bulldogs +58
Adelaide +17 v Sydney
GWS v Port Adelaide +10
Melbourne v Collingwood DRAW (I calculate an advantage to Collingwood of 0.2)
North Melbourne +33 v Fremantle

And the following tips for this weekend:

 

[Inferno]

Great to see Roby's tradition of totally overrating the Crows is being continued

 

[Hobbes]

Great to see Roby's tradition of totally overrating the Crows is being continued

Well, the most recent results count for more, so teams with a very strong formline (say, Adelaide, Sydney and the Bulldogs) will be favoured. In the this-year-only version, changes can happen very quickly. Home ground advantage counts for a bit too - in both of my models, it would tip Sydney over Adelaide if they were playing at home.

 

[catstorm]

Well, the most recent results count for more, so teams with a very strong formline (say, Adelaide, Sydney and the Bulldogs) will be favoured. In the this-year-only version, changes can happen very quickly. Home ground advantage counts for a bit too - in both of my models, it would tip Sydney over Adelaide if they were playing at home.

Could you explain how Adelaide is in front of an undefeated North Melbourne having lost to them only 3 weeks ago

 

[Hobbes]

Could you explain how Adelaide is in front of an undefeated North Melbourne having lost to them only 3 weeks ago

Sure, will do.

Adelaide's three results (weighted 20-18-16) are rated at a gross of +42,+54 and -5. This takes into account the match result and away bonuses. The net results, taking into account the quality of the opposition, are 34, 57 and 2.

North Melbourne's three results are worth a gross of +8.5 (a small bonus for falling over the line in a close match), +40 and +5. The net results are +5, +21 and +19.

In layman's terms, North's failure to beat Melbourne by more than Essendon could did not really impress, and little was proved by beating Brisbane by a smaller margin than everybody else. In comparison, Adelaide had a small away loss to a strong team, and two large wins against teams rated by many as contenders. (OK, Richmond were regarded as a contender, but many are now discounting them).

 

[Bulldog Joe]

I see an issue with your small sample size. Perhaps you would get a better result if you were working on a larger list such as the last 20 or so games with an increase factor for the last 5 to take account of current form.

 

[Hobbes]

Well, the sample size will right itself as the season goes on. Obviously, a sample size of 3 matches is not enough, but I thought I'd see how I go anyway. I also posted above a version using each team's last 13 matches. As I said above, the weighting is 20-18-16-14-12-10-8-6-5-4-3-2-1 so the last three matches comprise roughly 45% of a complete rating.

(I used my rarely-exercised discretion to not include the round 23 match between Fremantle and Port, where Fremantle rested their entire team.)

 

[BEaston]

Interesting. Thanks for the info and the effort to post.

 

[Bunk Moreland]

Clearly not Roby. Roby would have Essendon #1

 

[hawkman]

How many magic beans will be put on these predictions?

 

[EGYPTIAN HORSE]

So, basically its the ladder?

 

[HairyO]

So, basically its the ladder?

Its the ladder but it ignores how well North Melbourne are playing.

 

[shintemaster]

Sure, will do.

Adelaide's three results (weighted 20-18-16) are rated at a gross of +42,+54 and -5. This takes into account the match result and away bonuses. The net results, taking into account the quality of the opposition, are 34, 57 and 2.

North Melbourne's three results are worth a gross of +8.5 (a small bonus for falling over the line in a close match), +40 and +5. The net results are +5, +21 and +19.

In layman's terms, North's failure to beat Melbourne by more than Essendon could did not really impress, and little was proved by beating Brisbane by a smaller margin than everybody else. In comparison, Adelaide had a small away loss to a strong team, and two large wins against teams rated by many as contenders. (OK, Richmond were regarded as a contender, but many are now discounting them).

Your reasoning is fine, but how can Adelaide have a small away loss to a strong team if North aren't a strong team because they only just managed to beat Melbourne and had a smaller win against Brisbane. There is a logical disconnect there.

Its the ladder but it ignores how well North Melbourne are playing.

Ah. So a Bigfooty / Media ladder. Cool.

 

[DrEvil_]

Power Ranking returns...... LOL

 

[Furn2]

Are you saying you don't take the previous Season into account ?

 

[Hobbes]

Are you saying you don't take the previous Season into account ?

In the initial post, no. About halfway down the page there's a second ladder and set of tips which does take into account the last 13 matches played by every team, which does extend back to the middle of last year.

I can post both versions as we go to compare them.

 

[Hobbes]

Your reasoning is fine, but how can Adelaide have a small away loss to a strong team if North aren't a strong team because they only just managed to beat Melbourne and had a smaller win against Brisbane. There is a logical disconnect there.



Ah. So a Bigfooty / Media ladder. Cool.

North might not be as highly rated as if they had bigger wins or played better teams, but they still have a significant positive rating.