Community Forecasts

Also post this on Royals Review.

as well as the Fan Scouting Report.  These are great experiments and Tango runs a great blog.  I really encourage ANers to help out.  It's really simple -- takes a few minutes -- because all you project is OPS and games played (for hitters) and ERA and innings pitched (for pitchers).  No special knowledge required except being an attentive fan (which all of us are).

Is there/will there be a viewable results page?

Serves me right for cut/paste without reviewing.  In fact, I already posted in Royals Review with good results.

Everything I do that involves the community is posted.  So, just check out my site in early April.  And then check back in October when I compare them to PECOTA, ZIPS, Chone, Marcel, Bill James, Pete Palmer, and MGL.

I read the Sportsnation sports blogs for each MLB team, and I laughed/was sad when I saw no Pirates fan had responded to your request.
Is it ok to do be involved in the projections even if they are not the team we are a fan of? Or are you looking exclusively for fan projections?
I'm only a member of this A's blog, but I follow enough to post projections for your under-represented teams.
It just won't be from a Pirate/Royal fan perspective.
I'd also recommend the most recent SI, which you've probably already seen; it has at least 3-4 blogs for each ML team. That might give you a few more fan projections.

As forecasting systems start to come close to their theoretical limits for accuracy, it seems to me that any remaining progress is likely to come not from tweaking of internal parameters, but from integration with external information. What you're doing here is one very good idea - in addition to helping with playing time forecasts, which are a notable weakness of stat-based systems, fans might also be able to distinguish prospects from quad-A types. Do you have some specific issues like this that you're planning to look at, or are you just going to look at the overall results first and take it from there?

A related idea I've had would be to look at players for whom the different forecasts have the highest variance and see if there are any apparent patterns. We might find, for example, that ZIPS is better than PECOTA for one identifiable type of player, while the reverse is true for another type of player, which would allow both systems to be improved. (This one seems like an obvious idea, but I'm not aware of anyone who has done it.)

Also, one minor bug: The two lowest categories for playing time are "1-39 (Callup)" and "30-89 (Bench)." Presumably there is a typo in one of these.

The purpose of the fan forecast is to get the "inside" information that a stat-based system does not give you.  Is Mark Kotsay healthy, etc...

If someone wants to provide forecasts for other teams, they should do so only because they really follow those teams, and aren't just relying on stats (that's what PECOTA, Zips, and Marcel are for).

In the end, you'll have two methods, the stat-only, and the opinion-enhanced versions.  At the end of the season, we'll see how much one informs the other.