Certified Legendary Thread Race for the flag, in squiggly lines

[Pessimistic]

Maybe take out the hawks v brissy bears in 1988

Every last round has anomolies

 

[Final Siren]

It's probably been said (actually, after posting this I see it just was), but these graphs would be better if you excluded the Saints-Freo game. It is really distorting our position, and the circumstances surrounding that game provides a strong case for removing it.

What would the predictions have been for the Rd 22 Freo position vs Rd 22/23 Geelong position?

I just realized that my finals data was listing teams around the wrong way, so it thought the Geelong v Fremantle game was played in Perth. I've corrected the chart (here), which was making Freo look worse than it should have. SORRY ABOUT THAT DOCKERS.

Just for you, here is...

Hypothetical Freo If They'd Never Played Round 23​

This is kinda arguable, though, since Freo probably played better last week than they would have if 10 players hadn't had a rest.

Also, they move less far in this hypothetical chart than they do on the real chart, because there's a smaller gap between their actual and expected performance. That is, these things correct themselves over time.

 

[Freo74]

Interesting. Could just about draw a straight line from Hawthorn through Geelong to Fremantle.
Sydney have really fallen back to the field as well.
What's the projected score for a Hawks v Cats game if played this week? Apologies to Port but can't see them upsetting Geelong.

 

[Freo74]

Interesting. Could just about draw a straight line from Hawthorn through Geelong to Fremantle.
Sydney have really fallen back to the field as well.
What's the projected score for a Hawks v Cats game if played this week? Apologies to Port but can't see them upsetting Geelong.

 

[Roby]

Not to mention all the injuries we have to key players. Our injury list might be comparable to some other teams, but it's the players we have out, or who have retired because of injuries, not the length.

What is this subjective injury nonsense you are talking about? Has no place here.

 

[Final Siren]

Quite right, Roby.

Finals Week 2​

Dynamic Squiggle

Edit: Oh, and for next week the squiggle expects:
Fremantle 81 Sydney 62
Hawthorn 97 Geelong 87

 

[cryptor]

Edit: Oh, and for next week the squiggle expects:
Fremantle 81 Sydney 62
Hawthorn 97 Geelong 87

Does it only have Hawthorn winning by 10 due to the 2 goal home ground advantage adjustment?

 

[Final Siren]

Does it only have Hawthorn winning by 10 due to the 2 goal home ground advantage adjustment?

Yep. It would say Hawthorn 91 Geelong 93 at a neutral venue.

Bear in mind it's a really simple model and doesn't take into account ANYTHING other than venue and scores for and against. I.e. it doesn't account for Hawthorn having a week's break, or Geelong having the wood over them, or Chapman getting rubbed out for 8 weeks, etc etc. You with your superior human brain should do that.

 

[cryptor]

Yep. It would say Hawthorn 91 Geelong 93 at a neutral venue.

Bear in mind it's a really simple model and doesn't take into account ANYTHING other than venue and scores for and against. I.e. it doesn't account for Hawthorn having a week's break, or Geelong having the wood over them, or Chapman getting rubbed out for 8 weeks, etc etc. You with your superior human brain should do that.

Does the formula take the direction the teams are trending in into account at all, or just their static position at the time? (ie. Hawthorn are trending to the right while the Cats are drifting slightly down and left)

 

[Final Siren]

Does the formula take the direction the teams are trending in into account at all, or just their static position at the time? (ie. Hawthorn are trending to the right while the Cats are drifting slightly down and left)

Nope, each point is the model's best guess at where the side belongs. If there is a trend, it's up to you to spot it.

I wouldn't say Hawthorn is trending anywhere, since you need more than one data point to make a trend. But it's a nice time of the year to be heading right.

 

[pendlebury_10]

Are you the same guy that does a similar sort of thing for the Brownlow Medal?

 

[Suntory Boss Coffee]

Want the Hawks to be where the X is or up by next week

 

[Final Siren]

Are you the same guy that does a similar sort of thing for the Brownlow Medal?

No, but now I'm curious. What thing?

 

[Catfish Alley]

Does the formula take the direction the teams are trending in into account at all, or just their static position at the time? (ie. Hawthorn are trending to the right while the Cats are drifting slightly down and left)

Looking for some comfort?

 

[cryptor]

Looking for some comfort?

 

[crazywildhorse]

Does it only have Hawthorn winning by 10 due to the 2 goal home ground advantage adjustment?

I'll take it!

 

[cryptor]

I'll take it!

I don't really consider the MCG a home ground advantage for us over Geelong myself. It'd be barely worth a 2 point advantage let alone 2 goal.

 

[The Swert]

I don't really consider the MCG a home ground advantage for us over Geelong myself. It'd be barely worth a 2 point advantage let alone 2 goal.

Well it's far better than playing them at Kardinia Park. The curse would never end.

 

[crazywildhorse]

I don't really consider the MCG a home ground advantage for us over Geelong myself. It'd be barely worth a 2 point advantage let alone 2 goal.

Well at least it is a home jersey advantage!

 

[Purple_Voice]

Well it's far better than playing them at Kardinia Park. The curse would never end.

It's not as hard as everyone thinks

 

[Anoncomment]

Well it's far better than playing them at Kardinia Park. The curse would never end.

Nah, they have their own curse going on there.

 

[Bubbalicious]

Will the GST remain at 10%?

 

[Hawkk]

Well it's far better than playing them at Kardinia Park. The curse would never end.

We'd be a far better shot at Kardina Park.

We've won 10 of our 12 games at the venue!

 

[pendlebury_10]

No, but now I'm curious. What thing?

Another poster does a similar sort of thing to predict Brownlow winners each year and is very accurate.

 

[strobe00]

First of all, a big pat on the back to you for the squiggles. Some very interesting trends, and a nice way of looking at things. I like it.

But I have a couple of queries about what you said earlier... and this has nothing to do with stats; no arguments there. A straight forward approach that has been tuned using a lot of past data.

But what does this mean?

and Freo won't play "fair."

And what does this mean?

I think they can do that, since Freo is brittle.

Because if by that you mean that Freo is flaky, I think you may be allowing past Freo history to affect your perceptions of the team today. Or you haven't been watching us a lot this year.