Prediction It's the Final(s) Countdown! (A realistic approach)

Remove this Banner Ad

Aug 15, 2006
13,328
3,953
.
AFL Club
Western Bulldogs


Firstly I want to say that we are not a sure thing to make finals. This thread isn't about getting ahead of ourselves, in fact it's the opposite. It's about our realistic chances of playing a final.

I'll try and update this thread each week to keep tabs of how we're tracking for a spot in the finals. If we win next week I'll starting adding more info on top 4 chances.

First up is www.matterofstats.com finals simulation after round 17:

There is a lot of interesting analysis there and I recommend clicking the link above and having a read through.
Check out the part where he ranks the remaining AFL games in importance to final ladder positions.

1437981462101


Second up is resident AFL experts Champion Data:

CK5ziR-UAAAkLYZ.png:large


Thirdly is my own analysis of what the bookies are thinking - There is a reason these guys make money.

This is a collection of odds from Sportsbet, Bet365, TAB and CrownBet and then I've eliminated the juice. Obviously they have their own agenda here but I thought it was worth considering.

SwQ037B.jpg


And finally - When I average out the MatterofStats.com, ChampionData and the bookies odds I get this final table.

Top 8 Chance Percentages
EDB5K2x.png
 
I'm a very simple man Butane. Can 9 teams have a better than 50% chance of going into 8 even if 8 and 9 are currently level on points but separated by %. There is a mathematical explanation? (my maths ain't brilliant).
 

Log in to remove this ad.

My interpretation of this is that Butane is officially saying the Lid is Off !!!!! :p

And Dogsman16 - delete that post immediately....... That is the worst form of jinxing possible.... :mad: Mods, where are the mods when you need them.....
 
I'm a very simple man Butane. Can 9 teams have a better than 50% chance of going into 8 even if 8 and 9 are currently level on points but separated by %. There is a mathematical explanation? (my maths ain't brilliant).

With a finite number of positives outcomes available, you can have multiple outcomes with a positive % chance even when a negative outcomes is guaranteed.

Say there were 4 balls in a hat 3 of them black and 1 white. 4 people reaches in to grab a ball, each has a 75% chance of getting a black ball even though there is a 100% chance that one of them will get the white ball.

4 teams chasing the final 3 spots can be looked at similarly.
 
I'm a very simple man Butane. Can 9 teams have a better than 50% chance of going into 8 even if 8 and 9 are currently level on points but separated by %. There is a mathematical explanation? (my maths ain't brilliant).

If you add up all the percentages in the final table where I've collated them you'll find it equals 800%. What this means is by the completion of round 23 you'll have 8 teams with 100% because they have all made finals. Until then we are battling it out for a larger share of the 800% pie.

(oh and 9 teams don't have 50% I just didn't order the 'implied bookies odds' :p)
 
Last edited:
With a finite number of positives outcomes available, you can have multiple outcomes with a positive % chance even when a negative outcomes is guaranteed.

Say there were 4 balls in a hat 3 of them black and 1 white. 4 people reaches in to grab a ball, each has a 75% chance of getting a black ball even though there is a 100% chance that one of them will get the white ball.

4 teams chasing the final 3 spots can be looked at similarly.
Thanks. Probabilities always did my head in.
 

(Log in to remove this ad.)

The way i see it, win the games we should and its Dogs v North at Etihad
 


Firstly I want to say that we are not a sure thing to make finals. This thread isn't about getting ahead of ourselves, in fact it's the opposite. It's about our realistic chances of playing a final.

I'll try and update this thread each week to keep tabs of how we're tracking for a spot in the finals. If we win next week I'll starting adding more info on top 4 chances.

First up is www.matterofstats.com finals simulation after round 17:

There is a lot of interesting analysis there and I recommend clicking the link above and having a read through.
Check out the part where he ranks the remaining AFL games in importance to final ladder positions.

1437981462101


Second up is resident AFL experts Champion Data:

CK5ziR-UAAAkLYZ.png:large


Thirdly is my own analysis of what the bookies are thinking - There is a reason these guys make money.

This is a collection of odds from Sportsbet, Bet365, TAB and CrownBet and then I've eliminated the juice. Obviously they have their own agenda here but I thought it was worth considering.

SwQ037B.jpg


And finally - When I average out the MatterofStats.com, ChampionData and the bookies odds I get this final table.

Top 8 Chance Percentages
EDB5K2x.png

Loved that website you linked to, bookmarked that straight away.
 
Not that anyone is getting ahead of themselves or anything, but who would we like to meet first week of the finals?

GWS or Collingwood would be nice, maybe Adelaide?
Don't think I'd want to come up against North or Geelong though.
 
I'm a very simple man Butane. Can 9 teams have a better than 50% chance of going into 8 even if 8 and 9 are currently level on points but separated by %. There is a mathematical explanation? (my maths ain't brilliant).

Even though DoggyDan provided an excellent explanation, I would like to add my own explanation via a different method:

Let's take the Champion data table where Adelaide (in 7th) have a 68% chance, GWS (in 8th) have a 52% chance and North (in 9th) have a 52% chance

You could think of that table like this: Most likely it will be either GWS or North will make the finals (with a roughly 50% chance each) where one replaces the other. But there is a 32% chance that Adelaide will miss out, and that GWS AND North will make the finals.

That is the way that GWS and North can both have higher than 50% chance - because there are a bunch of possibilities where someone else misses out on the 8. And you can think of North's chances and GWS's chances pinching a few % each (so that they end up above 50% chance) from those scenarios.

Unlike DoggyDan's perfect explanation, my explanation is technically inaccurate. But it might help you wrap your head around things a little more.
 
Not that anyone is getting ahead of themselves or anything, but who would we like to meet first week of the finals?

GWS or Collingwood would be nice, maybe Adelaide?
Don't think I'd want to come up against North or Geelong though.
i dont like the vibe of this thread...... :eek:

i'll feel much better if we beat essendon

BUT id take GWS i think. The pies would have an awesome atmosphere though
 
Even though DoggyDan provided an excellent explanation, I would like to add my own explanation via a different method:

Let's take the Champion data table where Adelaide (in 7th) have a 68% chance, GWS (in 8th) have a 52% chance and North (in 9th) have a 52% chance

You could think of that table like this: Most likely it will be either GWS or North will make the finals (with a roughly 50% chance each) where one replaces the other. But there is a 32% chance that Adelaide will miss out, and that GWS AND North will make the finals.

That is the way that GWS and North can both have higher than 50% chance - because there are a bunch of possibilities where someone else misses out on the 8. And you can think of North's chances and GWS's chances pinching a few % each (so that they end up above 50% chance) from those scenarios.

Unlike DoggyDan's perfect explanation, my explanation is technically inaccurate. But it might help you wrap your head around things a little more.

Have a further read of http://www.matterofstats.com/mafl-s...ulating-the-finalists-for-2015-after-round-17.

He discusses which games are most important and also which other teams winning/losing has the greatest ripple effect on other teams. It's really quite indepth.
 
North Melbourne in the first week of finals, I like the sound of that.

So would they I'd imagine.

I know we probably wouldn't do any major damage in the finals this year if we made it but I'd love seeing Stringer and Bont perform in a final. They seem the type made for the big games.
 
With a finite number of positives outcomes available, you can have multiple outcomes with a positive % chance even when a negative outcomes is guaranteed.

Say there were 4 balls in a hat 3 of them black and 1 white. 4 people reaches in to grab a ball, each has a 75% chance of getting a black ball even though there is a 100% chance that one of them will get the white ball.

4 teams chasing the final 3 spots can be looked at similarly.
Thanks again and one good turn deserves another so in the spirit of mutual learning and cooperation the plural subject of a regular verb (people in this case) takes reach :p.
 

Remove this Banner Ad

Back
Top