It’s definitely one of the larger changes! It’s pretty hard to avoid, however. 119 PA of merely subpar play would change our best estimate of Upton’s “true talent” very slightly, 119 PA of 167/277/216 is enough to push that probability density function to the left a bit.

There are tons of times we try to answer the question about a player “what’s he going to do the rest of the way?” For that, these are perfect (well, not perfect, but pretty good.) Zach Greinke’s off to a hot start. Statistically speaking, what do we expect the rest of the way?

Also nice for fantasy, if you can apply a playing time adjustment in your head.

I don’t like the “season total” thing, as there are rare cases you care about a player’s projected total season line, and yet it gets quoted more than it should. From re-reading azruavatar’s comment, maybe this is the piece you don’t like the most: we’re revising our projection of what a player’s season line will be? Yeah, a little revisionist. Hey look, I can nail the season projection after 162 games worth of knowledge!

Let me try to put it to you differently.

The idea of a projection is primarily to convey useful information about something. Being as accurate as possible is the means to this end, not the end itself – if a player’s statistics didn’t convey useful information about baseball, then all projection systems would be kind of pointless. It’s not trying to tinker with the crystal ball, but trying to answer a useful question – how a player will play the rest of year is useful information, so we should use the best information possible at all points.

Let me give you an analogy. Let’s say it’s payday at your office and after your paycheck is deposited, you have $2500 in your checking account. You look at your bills and you project that, after paying all your outstanding bills, you will have $1100 in your checking account. 3 days later, however, your air conditioning goes on the fritz and you have to spend $600 having it repaired. At that point, 3 days later, you don’t care whether or not your original projection of $1100 is erroneous, you care about projecting what your checking account will look like after factoring in the new information.

Projections to me aren’t important as some sort of contest to see who can be the most right, they’re trying to give people an idea of what to expect from the season. And adjusted projections are trying to give people an idea of what to expect from the rest of the season.