AFL Finals Dream Team 2007

Remove this Banner Ad

I have gone a very different side this week, hoping that i can make up a bit of ground. They arent consistent players though so i could easily bomb big time!

same here

i have scrapped boomer harvey and hunter and gone for jesse smith in order to get the mids i want

i only have swan in the mids but think someone like burns may score high

this week is so hard
 

Log in to remove this ad.

My football knowledge isn't the best so I used a model to pick my team, and then made a few changes after that.

The changes I made over and above the model last week made me lose about 20 points will be interesting to see how that goes this week.

Interesting. I'm quite a fan of doing that kind of analysis myself.

Care to explain how your model works?
 
Interesting. I'm quite a fan of doing that kind of analysis myself.

Care to explain how your model works?

I have set up a linear programming optimization model in SAS (Statistical Analytical Software) and used each player's average for the past 5 games as the objective equation to pick the team I wanted. Each of the constraints are fairly self explanatory from the rules. My friend was able to show that the yearly average was a better predictor than last 5 (or 4,3,2,1 for that matter) and has picked his side that way. He also uses the model to tweak his team as opposed to intuition, so he would preclude someone like Sansbury from getting picked and then let the model choose the team from there. And continue to do that until he had a team he was happy with.

The tweaking is still a guessing game I suppose but it keeps it interesting, the captain choice has not been modelled. I just scraped over 2000 last week while he got close to 2100. Interestingly it told us to stay clear of some of the players that had shockers last week like Harvey, Corey, McLeod and O'Keefe.

The regular season DT comp will be a bit harder to model as we'll have to give some subjective values (or use something other than last season's average as that is pretty well correlated to price) to all the players we are interested in to get the model to pick the team. But we're keen to try our luck again next season.
 
Thanks for the explanation

The regular season DT comp will be a bit harder to model as we'll have to give some subjective values (or use something other than last season's average as that is pretty well correlated to price) to all the players we are interested in to get the model to pick the team. But we're keen to try our luck again next season.

I got most of the way through building an Excel based model for Supercoach this year but ditched it because I found myself deviating too much to accommodate that all important subjectivity.
 
Thanks for the explanation



I got most of the way through building an Excel based model for Supercoach this year but ditched it because I found myself deviating too much to accommodate that all important subjectivity.

I think an approach that might work is to survey (subjectivity on mass) DT competitors with proven pedigree and get them to give average points per player that would be over and above what their price would indicate. Use some sort of average of these values to drive the model. In return for doing the survey I'll happily provide an individual model for each participant's own evaluation of players while at the same time producing an overall model to use based on everyone's responses. May even release the findings of the overall model to all participants and let them choose their own team from there.

Do you think something like this could work?
 
hey sorry if its already been asked, but do you only get double points this week if you register this week or is it double points for everyone?
 

(Log in to remove this ad.)

Well it defeates the purpose of doing well in round 1. A 50 point mistake is now 100 points and so on as we go further. It would make sense to just leave it as is, or only choose 11 players (double points). But when we have 22 players there is no need to double points just seems pointless. 4000 is on the cards this week. :s
 
Can anyone explain why its double points? we still have 22 players...

I think you'll find that as the weeks progress everyone's team will become more and more similar, so to give competitors some hope beyond week one, they make the latter rounds worth more. The interesting thing about this comp is that you won't win it in one round but you sure can lose it in one round.
 
Maybe part of the idea is the lower finals are worth as much while the Grand Final is worth heeps. Like in AFL playing in the grand final is worth a lot more the n week 1.

Also smmorrows idea that the teams will become more and more like each other since there is less to choose from, althoguh i think week 2 and 3 of finals has same to choose from.

Its just for fun basically, something different. Itd be boring if it was basically the same as normal DT
 
600 club??

I just had this thought:
This weekend, you might be able to become a member of the ultra-exclusive 600 club!
There aren't many dt'ers who can say that one of their players has ever scored 600 points in a game, but it could be done this week.
If your captain manages to pull out a monster (150+), then you'll be part of the 600 club this week. :cool:
150 x 2 (double points week) x 2 (captaincy bonus) = 600!
Sounds pretty neat, huh?
 
Re: 600 club??

I just had this thought:
This weekend, you might be able to become a member of the ultra-exclusive 600 club!
There aren't many dt'ers who can say that one of their players has ever scored 600 points in a game, but it could be done this week.
If your captain manages to pull out a monster (150+), then you'll be part of the 600 club this week. :cool:
150 x 2 (double points week) x 2 (captaincy bonus) = 600!
Sounds pretty neat, huh?

Quite an interesting proposition. Whose the go for captain tho? No surethings.

Priddis, Boomer, Hodge are the only ones which come to mind.
 

Remove this Banner Ad

AFL Finals Dream Team 2007

Remove this Banner Ad

Back
Top