THE BOOK--Playing The Percentages In Baseball

Here is Ichiro’s forecast page:
http://www.fangraphs.com/fanpdetails.aspx?playerid=1101

You will see that there are 25 ballots from Mariners fans, and 5 from non-Mariners.  And yet, look at his WAR: 4.1 from Mariner fans, and 3.9 from non-Mariner fans.

See what I’ve been saying about not needing too many fans to get these forecast things done?  Things have a tendency to settle down pretty fast.

You even see that in my quick polls on this blog, where after 20-30 answers, we already know the most popular answer.

And keep tabs on this page:
http://www.fangraphs.com/projections.aspx?pos=all&stats=bat&type=fan

So far, only Pujols and Ichiro got enough votes to appear on the “leaderboards”.  But, I will guess that within 2 days, you will see a few hundred players there.

I love it that David has taken this on…

I’d love to see some statistics of predictions. Standard deviations would be nice but histograms would be awesome.

It has been fun doing these. The leader boards are more populated now FWIW.

Fans are WAY too optimistic it seems.  We’ll have to change the baseline appropriately.

This project is gonna have tons of info.  I think one thing that might be cool is to see how fans from different divisions or leagues view players around the league.

It seems the biggest inaccuracy with fan projections will be playing time.  I would guess that most fans simply won’t regress playing time sufficiently and will assume that most starters will play 150+ games leading to inflated counting stats across the board.

At first blush, yes.  Which is why we can scale it for them, after the fact.  We can basically “force” in the numbers.