We've been here before-remember Bill's projection for Bagwell's rookie year? Yeah he didn't win the batting title but he did hit .294 and win ROTY anyway.

Wouldn't it be something if a Boston farmhand wins ROTY for the 3rd year in a row (Hanley albeit in the NL in '06). The Sox may very well have Ichiro Jr. in CF for the next 4 cost-controlled years...

remember Bill's projection for Bagwell's rookie year?

If I recall correctly, that wasn't a projection, but a MLE. I wonder if this is the same thing; it would make much more sense.

I thought of Bagwell too. .320 sounds high as a projection or MLE. The MLE spreadsheet I had for him had him at about .300 BA, including his ML time.

He'll get so many legs hits that his BA should be halfway decent as long as he avoids too many pop-ups. He's quite speedy, in case you've all forgotten.

And Timmy McCee says Ellsbury's a "good looking ballplayer." He said Dusty is a "good ballplayer", BTW.

MLE before the fact is a projection. Think about it.

But in Bagwell's case it was an MLE after the fact. Not sure about Ellsbury's.

An MLE is not a projection. An MLE is simply what a minor league (or some othjer league, like in Japan or the AFL) player "would have done" had he played in the majors, based upon lots of data from players who played in both the minors (or whatever league) and majors in the same season (or different seasons, if you adjust for age).

The key difference is regression. A projection is the mean or median of what we expect a player to do in the future based on his past performance. If that past performance is in the minors and we want to do a major league projection, we first have to do an MLE and THEN we have to use a projection algorithm which includes a regression toward the mean. So if a player has an MLE of .300 in the minors, which means literally that he "would have" hit .300 if he were playing against major league competition in major league parks, etc. He probably hit like .320 in the minors. Yet, we would NOT project him to hit .300 in the majors next year (or whenever). Not even close. Based on one year of data (assuming he had only a one-year MLE of .300), we would regress BA quite a bit for a projection. In fact, we would probably project him at something like .285 (I don't know the number off the top of my head).

Everyone's MLE's (the forecaster that is) are different of course, but in 06, I have Jacoby at .296 in 210 PA, in 07, and .292 in 07 in 481 PA. And of course he hit .353 in the majors in 127 PA. That .296 and .292 in the minors are context-neutral MLE's of course. He will hit a little higher in Fenway I assume (I assume that Fenway inflates BA as compared to the average AL park. But, my MLE's are for NL and AL players combined. Since the AL is the tougher league, he will hit less in the AL than in the NL, so I'll call the park factor and league factor a wash.

He is on the steep part of the aging curve, so you have to adjust the 06 and 07 numbers upwards for an 08 projection. Doing a quick Marcel in my head, you get around .310 for a weighted average of prior PA. Off the top of my head, I would think that you would regress that around 30% toward the mean of say .280 for a speedy 25 yo rookie with good skills. That leaves us at .301. Age adjust to his 2008 age and we have maybe .307.

The Handbook is forecasting almost mirror seasons for Josh Beckett and Daisuke Matsuzaka: same won-lost record (14-8), similar numbers in ERA (3.50 and 3.54, respectively), strikeouts (187 and 188), and walks (64 for both).

Hmmm.

Isn't it nice for a member of the Red Sox braintrust to be able to put out a set of projections that say their prospects are going to be super-great? How convenient.

I wonder if the Red Sox use James's projections while evaluating players for acquisition. That would explain why his projections seem overoptimistic about Red Sox players and prospects. Especially his pitching projections seem unrealistically good for the Red Sox staff for the last 2 years.

These aren't James's projections: they are put together by folks at Baseball Info Solutions. James's name is on the handbook as a selling point.

-- MWE

These aren't James's projections: they are put together by folks at Baseball Info Solutions. James's name is on the handbook as a selling point.

I am not sure what it means, but James' name is the listed author for the batting projections.

"The annual handbook is a collaboration between James, senior baseball operations adviser for the Sox, and Baseball Info Solutions, an industry leader in innovative statistics.

James is listed as the author of the entry on projected batting records, and as he indicates in his opening comments"

Do they use Brock6 or whatever it's called? IIRC, Szymborski based his ZiPS on that.