Lies, Damned Lies and Statistics. (And Football)

[matmat]

Edit: if you don't want to read this whole thing the questions are in capsule form in the first reply.

-----------------------------------------------------
I have several questions for those that keep their ear closer to the ground when it comes to fantasy football than I do.

In baseball the number of at bats and IP is so large and the data accumulated over the past 100 years or so is so vast that people have devised various sorts of formulas to project player stats based on various factors about the players and historical precedence for similar players. Does such a thing exist in football? or is stat projection essentially a set of completely educated guesses?

I am totally fed up with all the "ranking sites." Those are all fine and dandy if:
1) The scoring system used to make them is the same as the scoring system that you're using in your league
2) The scoring system for the ranking is actually given.
I am in several leagues that use non-basic scoring and for me stat projections are a lot more important than someone's rankings. Where do you get your stat projections? are there free sites?

Anyone have any scientific treatment of draft strategy? I have some ideas about this, but I'm wondering if anyone has done any calculations... i was mainly looking at falloffs in player production at various positions. For example, suppose you want to take a WR and RB in the next two rounds, can you quantify/estimate, for a given point system and set of projections which order will net you a larger number of points?

[matmat]

my original post is long, so here are the questions in capsule form.


1) is there a mathematical model for player projections in football?
2) where can I get free, relatively reliable stat projections?
3) is there a mathematical model for draft strategy?

[KingGhidra]

my original post is long, so here are the questions in capsule form.


1) is there a mathematical model for player projections in football?

Yes.

2) where can I get free, relatively reliable stat projections?

From someone who knows the answer to your first question and doesn't charge. I don't know any offhand that will do it free and provide free updates.

These are in line with my projections and are free but are about a month old. I'd be willing to bet a model was used to forecast those stats.

http://apps.footballguys.com/04rbproj.cfm

[34=Sweetness]

Mathmetical models do not work with human beings. There are certain intangibles that math cannot take into account. Thats just my opinion. I think footballguys.com has some nice stat projections. But im not quite so sure. BTW Im CubsFan7724 on the baseball forum matmat.

[KingGhidra]

The statistics gurus at this website that I know of are Buckychudd and CrimsonTide (who I haven't seen since December).

You might want to PM them nicely with questions.

[Acme Rockets]

Take it from a guy with a degree in math. There is no math based statistical formula that will give you any indication of how well Guy X will do.

Your best bet is to watch the guy, see the changes made to teams in the offseason and base your picks on your beliefs.

[KingGhidra]

Mathmetical models do not work with human beings. There are certain intangibles that math cannot take into account. Thats just my opinion.

Obviously, but they help you make informed decisions. If you have a choice between RB A and RB B in round two and you know that historically, guys with RB A type numbers lose a lot of value each season, you'd take RB B and you'd have history on your side.

[34=Sweetness]

I understand that, but there is no endall math model that will tell you Player A will strictly be better then Player B. All you have is patterns and trends.

[KingGhidra]

I understand that, but there is no endall math model that will tell you Player A will strictly be better then Player B.

Of course not. You have to use sense too and look at things from a football perspective.

All you have is patterns and trends.

Which is better than nothing I think.

[34=Sweetness]

I understand that, but there is no endall math model that will tell you Player A will strictly be better then Player B.

Of course not. You have to use sense too and look at things from a football perspective.

All you have is patterns and trends.

Which is better than nothing I think.

Agreed.