They then dropped him to the bottom of the order for 9 of his next 10 starts (7 games hitting 9th and 1 each hitting 7th and 8th). He hit .429 / .459 / .629 in those 10 games (and two PA off the bench), while going 0/4, HBP in his sole game hitting leadoff.
They then returned Ellsbury to the leadoff spot, and he hit .300 / .333 /.438 the rest of the way.
Ellsbury's performance when dropped down in the order was precisely as I had predicted, but there was weak evidence of a changed overall approach upon his return to the top of the lineup. He was better, but not so much better that I've ever been motivated to test the improvement for significance.
Nevertheless, I felt convinced at the time that this interpretation of his performance and power upside was correct, and I looked forward to the day when he would start attacking the ball whenever he was in a pitcher's count, and blossom as a home run hitter.
Again, I have no idea how these e-mails actually fit into Ellsbury's history. At one extreme, it's possible that they tried getting him to change his approach immediately, and couldn't convince him to try it for two years. At the other extreme, the e-mails may have been neglected and forgotten (although I had been thanked and told they were interesting), even when they finally noticed for themselves that Ellsbury's approach in most hitter's counts was criminally defensive.
What we do know is that he started hitting home runs in 2011. If you look at his 2011 splits, you'll see that he hit .388 / .512 / .711 when ahead in the count, an OPS relative to league (sOPS+) of 154. In 2009, he had hit .319 / .455 / .476 when ahead, which sounds good, but is actually well below average (sOPS+ of 86). His sOPS when behind in the count went up from just 170 to 180.
And you know what? It was incredibly obvious. He was sitting on pitches and hammering them, and he'd never done that before. It caught me eye almost immediately and made me very happy. I think even Jim Rice noticed it and pointed to it at the cause of his improvement.
What's truly interesting is Ellsbury's percentage of balls hit in the air. Previous to 2011: .475, .483, .499, .507 ... and the next term in that series is .570. That, folks, is best explained as the product of a somewhat different swing. A swing that a guy who believed he might be able to hit more home runs by sitting on more pitches might adopt. So we have good reason to believe that Ellsbury changed his approach in two different, complementary ways in 2011.
And then he suffered a serious shoulder injury seven games into 2012, and was out until the All-Star break. He ended up with .533 balls in the air, easily his second highest career mark, but a decline in HR/FB from .167 to .047. In other words, the shoulder injury killed his power, but his swing was still close to his 2011 version, and since he wasn't hitting the ball on the ground, where BABIP thrives, he ended with a dismal slash line of .271 / .313 / .370.
His 2013 divides into two very interesting chunks. On July 3 he was hitting .298 / .361 / .404, with a career-low .474 balls in the air, and a microscopic .013 HR/FB. This is a guy who recognizes his shoulder is still hurting him, has gone back to hitting the ball on the ground, and is having success with the style.
From July 4th on, Ellsbury hit .298 / .346 / .458, with a .518 balls in the air and, more tellingly, a .140 HR/FB. Now, there's no question in my mind that his 2011 numbers were greatly boosted by the reluctance of opposing pitchers to take him seriously as a power threat (that's just based on watching every game and often saying "I can't believe they threw him that pitch" while grinning). I think his 2013 second-chunk .140 HR/FB is precisely the same pop as his .167 from 2011; the difference is tougher pitches.
So the next question when we try to project Ellsbury's power going forward is his balls-in-air percentage. I don't think it's too easy or even desirable to change your swing too much mid-season, so I think it's quite possible that his .474 to .518 change in 2013 doesn't represent what he's capable of, if he begins a season feeling his shoulder is 100% (as he ought to in 2014) and is in a home-run hitting frame of mind (ditto). I think .540 or .550 is reasonable.
Finally, there's Yankee Stadium. In 2011, it would have turned his 15 Fenway homers into 17 (adding five to right and subtracting three from left), but in 2013, it would have turned 4 into 11. It's hard to derive a prediction from that, and no one has HR/FB Park Factors by batter handedness. But my most recent edition of the Bill James Handbook tells me that LHB hit about 75% more HRs in Yankee Stadium than in Fenway.
So, what happens if Ellsbury maintains his second-half HR/FB, improves his balls-in-air percentage to .550, and plays 120 games for the Yankees? He hits 25 HR. If he plays 140 games, he hits 30.
I think he's going to be worth the money.
Do I wonder whether I helped him earn even a tiny part of it? I suppose I do.