5

I have an idea for a fantasy football experiment looking at three different areas of the draft. These three areas are the position of each pick and how often and what value are trades.

I have dug through the literature and have only found a couple of papers about this topic. I was wondering if anybody has heard of this vetting done before or any ideas on how to make this idea more concrete. If it is to vague I can explain it more later on.

Thank you for all of your help guys

UPDATE:

A little more information, the paper will look at level k play in regard to fantasy football and the implementations of this paper is to ultimately either to help companies learn how to players play and develop a better drafting system and a more ok intelligent player pool. Think of it of this way: is crowd mentality in ranking better than one's own ranking system?

user4318524
  • 51
  • 2
  • 1
    I think your assuming that your going to have someone answer who knows something about fantasy football and economics, which - given the current site size - is not likely. To get some more feedback I suggest you make it more specific what your looking for. – Thorst Feb 16 '15 at 12:14
  • 2
    Soccerman makes a good point. That said, I'm an economist (well, in training anyway) and an avid fantasy football player. One of my research projects stemmed from a fantasy football question -- comparing efficiency and other properties of assignment mechanisms (i.e. drafting, auctions, etc.) and trying to develop an alternative. You won't find a lot of published papers on fantasy football, but a lot of the questions in fantasy football are more universal and important. As it is though, your idea is not specific enough to really provide much feedback. What is the goal of your planned analysis? – Shane Feb 16 '15 at 13:44
  • I just edited the post you too. I would love your critics of it – user4318524 Feb 16 '15 at 14:35
  • 1
    That's a step in the right direction. Can you explain how you would implement a Level K model in a draft? That environment is far more complicated than the beauty contest in which I'm used to seeing Level K reasoning demonstrated. Is your ultimate goal to develop a better drafting system? Or a "more OK intelligent player pool" (I'm not sure what that means, incidentally)? – Shane Feb 16 '15 at 15:05
  • 1
    One further clarification. May I ask where you are from and in what context you hope to carry out this research? I think that's very pertinent to the design. – Shane Feb 16 '15 at 15:08
  • Definitely, I would use level k to test if after repeated tried running a draft do people still use a novice approach such as Massey and Thaler paper proved or do people start to shift towards the more advanced style of gameplay as Belchik and the Patriots, and Colts, and the Steelers use. (The War Room book) – user4318524 Feb 16 '15 at 15:09
  • And our research will be in Texas which a student subject pool of about 2000 student volunteers. – user4318524 Feb 16 '15 at 15:10
  • The Massey and Thaley paper demonstrated overconfidence (at least that was the headline result). So would more sophisticated types (i.e. higher level k) be less overconfident? If you're interested in how people change their behavior when they play a game repeatedly, there's a fairly sizeable literature on that in behavioral economics. I don't know the literature well enough to cite anything, but comparing first responses vs. more experienced play is common - I've definitely seen it for the beauty contest (where Level K was originally played with). – Shane Feb 16 '15 at 15:15
  • I realize the literature is pretty heavily there for responses in item but none is done on the effects of this effect in a sports draft environment. Also that paper says that more novice players will over confident in the drafting but a more experience player or coach will sit back and get the more picks they can. This is a theory paper and I want to make a experimental paper out of it. – user4318524 Feb 16 '15 at 15:18
  • I would not say Massey and Thaley is a theory paper. It is applied/empirical. It seems what you would be doing is trying to find through experiments how drafting behavior changes with experience. Essentially it's a learning model, whereas Massey Thaler was just a static result. In a sense that's fine. Why would you do this as an experience rather than just looking at the level of experience of Coach/GM/Owner pairings in the NFL and seeing if drafting strategies correlated with this experience? – Shane Feb 16 '15 at 15:25
  • Yeah an a learning model has never been attempted with the draft in either baseball, football or soccer. Out of curiosity what sport do you think would get the biggest response? – user4318524 Feb 16 '15 at 15:27
  • Response in what sense? In terms of interest? If this is about fantasy sports, then fantasy football is so much more popular than any other fantasy sport that it obviously would be. If it's about real drafts, I'd still say football. The NFL draft is a major event -- are any other drafts besides the NBA even televised? – Shane Feb 16 '15 at 15:31

1 Answers1

3

Your question is long and unclear. I suggest editing it to get a better response. I hope you do because I think it has good potential.

I am able to help with one element of the question.

"is crowd mentality in ranking better than one's own ranking system?" The answer to this really depends on what your own ranking system is. The Wisdom of the Crowd is pretty good at this sort of stuff. That said, I think there is definitely room in something like fantasy football to beat the crowd generated rankings.

I find when I am looking for a probabilistic outcome on something like that, a bookmakers site should be one of the first ports of call.

Jamzy
  • 3,751
  • 18
  • 48