Community Forecasting and the Wisdom of Crowds

[bedir than average]

Tangotiger was an old FanHome poster on the Sabermetric and Montreal Expos forums. He has a multiyear project with which FanHome has helped create data in the past called the "Fan Scouting Report"

This year he has taken the theory of the Wisdom of Crowds concept a step further and has developed a survey to provide projections for every player currently on a 40Man roster. This should be ~90% of the guys who get even a couple dozen ABs this year.

He has always made his Fan Scouting Report public and in this case he would make the results public but also study how they relate to 6 other projection systems (Marcel, ZiPS, Chone, BillJames, PECOTA, Shandler).

Here's a link to his survey for the Community Forecasting survey

Tango on Baseball

I've seen the results of six forecasting systems this year. (I'm sure some of you have seen more than that.) And all were based on some algorithm with little leeway for human interaction. Why is that? Because we can't trust any single person's opinion. But, what if we can get a consensus, a Wisdom of Crowds? Who knows more about whether Papelbon will be a starter or reliever this year: an algorithm or a Redsox fan? Who knows more about the number of games a 2006-injured Hideki Matsui will play in 2007: an algorithm or a Yankees fan? There are certain human observation elements that are critical for forecasting. That's where you can come in, and why you are here.

Go to the page of your favorite team, and put in the OPS (OBP+SLG) and ERA you expect from as many players as you feel comfortable. As well, if you can, note the number of games for the non-pitchers and the role for the pitchers. The players listed are on the 40-man roster.

Once the season starts, I'll report the Crowds' expectation, as well as those from the professional forecasting systems. Let's see who has more insight.