A Sox Therapy thread! Yay!

I don't think Varitek is completely done, but I expect him to be among the least productive members of the front-line talent this year. (Yeah, really going out on a limb.) And I want the transition away from him to start this year, though it doesn't have to happen starting on opening day.

Good topic. It should be interesting to check in on from time to time.

Let's not forget how valuable the observations of Jim Edmonds were last year with the Padres. Stick a fork in him already.

228/326/381

This is more general than Varitek specific but for only an .008 point jump an average that's a .022 jump in slugging percentage, which seems high to me, since presumably a majority of the hits he'd pick up would be singles. Am I just doing a terrible job of the math? That wouldn't shock me.

Let's not forget how valuable the observations of Jim Edmonds were last year with the Padres. Stick a fork in him already.

AROM, how close is Edmonds to the magic 70 WAR?

RB,

I'm not sure what you're getting at. He hit something last year and the system thinks he'll hit something else this year. It doesn't have mean he'll pick up X singles, X doubles, etc.

First, I think the projections can be usefully supplemented with observational data. I'm not sure anyone called for chucking out hte projections, but if they did, I guess I happily reject such a course of action.

Second, Varitek's projections are pretty darn terrible. 725, 787, 672... 707? A simple 5/4/3 projection would put him at a 724 OPS, and any regression to the mean factor would bump that up a little more. So, accounting for Varitek's age and position loses him several runs of projection. I think there's another several runs of projection to be lost when you account for the qualitative decline in bat speed, particularly from the left side of the plate.

Varitek had a pretty grim year last year but the disgrace that was his June was a big part of it. Even in July and September when he was bad he was (I know of no other way to put this) "normal bad" with a .581 and .586 OPS in those months. June for whatever reason he put up a .380 OPS which is just off-the-charts bad. If you take his June out his OPS was .718 for the season.

I realize the fallacy in just arbitrarily removing the worst month of the season but that month was SO bad I think it's a fair question to ask if something was going on there. Was there an injury we didn't know about? Was the divorce an issue for some reason that month more than others? Even if you substitute a July-level performance for June his OPS for the season is a more palatable .697. I just believe (maybe I'm wishing it was so) that SOMETHING was up in June beyond just a normal slump.

Yes, his projection does take a hit because of age and position, I would guess. And I think that's what those who don't buy the projection are missing. His being 37 and a catcher is already rolled into the projection itself.

I think whether we can look at Tek and determine a loss of batspeed (or whatever) that is not going to be reflected in the projection is questionable. Out of curiosity, how many is several runs. Will he be 1.1 WAR? 0.6? Let's make some predictions and see where the chips fall.

I'm not sure what you're getting at. He hit something last year and the system thinks he'll hit something else this year. It doesn't have mean he'll pick up X singles, X doubles, etc.

I meant that it seemed like the numbers were seeing a significant uptick in power for Varitek, relative to his hits in general. Maybe I'm wrong about it.

And I think that's what those who don't buy the projection are missing. His being 37 and a catcher is already rolled into the projection itself.

Right, but his being Jason Varitek, by definition, cannot be rolled into the projection. And I think it's very obviously true that the fact that he's Jason Varitek, and not someone else with certain shared characteristics, is a fact that must be taken into account every time you decide to speak about his likely future performance. What bothers me about "those who buy into only looking at the projection" is that they don't acknowledge the epistemological status of a projection.

One could make the argument that the difference between the individual Jason Varitek and the set of aggregated characteristics that are fed into a projection system is not usefully knowable in any case, or is not usefully knowable in Varitek's case. I find the former deeply problematic, and the latter at least reasonable, and then we'd be arguing on the same wavelength.

Speaking to the former, you could have a case where something catastrophic enough happens that we have a pretty good idea that the projection != Varitek. On the extreme end, he could have his arm blown off or he could die. Somewhat less extreme, but still convincing, would be if he had Tommy John surgery the week before the start of the season. Then we'd know that the stuff fed into the projection was bogus.

Where it breaks down for me is when we eschew the projections in favor amateur scouting reports about body quickness and bat speed (I know that's a cruel pairing). I just think fans are generally too biased by their frustration watching players fail to accurately predict whether they would bounce back. I would also say that we, as fans, don't know enough about how bat speed and the like come and go from season to season. Tek may well have had a slow bat last year but regain some speed on it next year.

Also, let's hear how you think he'll hit. I really think it will be a fun exercise to follow along this year and see how well the projections did vs. projections + BBTF scouting. Who knows, I might even be wrong!

Where it breaks down for me is when we eschew the projections in favor amateur scouting reports about body quickness and bat speed (I know that's a cruel pairing).

But I, at least, didn't eschew the projections. I used them as my baseline, and I made occasional and moderate adjustments in cases where I think other factors relating to the particularity of the human beings and human bodies that are the actual baseball players are observable and useful.

I would tend to think that Varitek is going to hit somewhere in the 650-680 OPS range, a season roughly equal to his last season, a bit worse. I want to watch him hit in spring, because I think it's at least conceivable that some mechanical adjustments to his swing might allow him to stave off decline in the short term.

I'm going off the top of my head here, but if I had to pick another Red Sox whom I think is over- or under-projected, I'd say Jon Lester is underprojected because of the real, qualitative improvement in his command that came from growing leg strength and general cancer recovery. I can't think of any other players right now where I think I've seen something that I'm confident enough in to bring to the table. I looked up Kevin Youkilis in case the consensus had him retaining his fluky '08 power spike, but it looks like he's projected to lose most of it.

For what it's worth, I think the people who dispute Ellsbury's projections (see the last active thread) are basically wrong, and I don't see major issues in his game what would prevent him from being a good contributor in CF, as he is projected.

I'm skeptical of Jed Lowrie's glove, but that's a case where these actually isn't any useful data to work with anyway.

EDIT: Do you think that that there are no Red Sox players for whom you have useful information to add in to adjust their projections?

I realize the fallacy in just arbitrarily removing the worst month of the season but that month was SO bad I think it's a fair question to ask if something was going on there.

I think a 580 OPS month is pretty standard variance from a guy we think has a true talent OPS in the 715 range.

I think his june OPS was actually 380. As impossible as that may appear.

EDIT: I checked it up. It was 381

Oops, I misread the post.

Still don't think it's anything special. I would guess that a month that bad would happen once every 12-15 months by chance for a guy like Varitek.

I'm going off the top of my head here, but if I had to pick another Red Sox whom I think is over- or under-projected, I'd say Jon Lester is underprojected because of the real, qualitative improvement in his command that came from growing leg strength and general cancer recovery

I think the fact that he had cancer that messed with his performance that he is completely recovered from makes his projection less useful.

I don't think you know enough to say that he is underprojected because of "real, qualitative improvement in his command that came from growing leg strength". To me that is exactly the same as saying Youkilis' power surge was for real because of his obsessive workouts at whatever facility it is that Peter Gammons loves talking about, or that Shea Hillenbrand's sessions with Dwight Evans were so enlightening that his .330/.330/.550 April is for real.

It sounds plausible, but I can imagine other explanations that sound plausible. How about the improvement was from Lester submitting to Jason Varitek's calming leadership and making changes to his approach? Remember in April he was the exact same maddening(ly lucky) pitcher he was in 2007. I'm pretty sure he didn't grow some glutes the night before his no hitter, but look at this quote from Lester talking about how he gives all the credit in the world to Jason Varitek for his great performance. How are we to choose from the infinite, plausible sounding explanations?

Also, let's hear how you think he'll hit. I really think it will be a fun exercise to follow along this year and see how well the projections did vs. projections + BBTF scouting.

Prediction: .239/.328/.377.

It sounds plausible, but I can imagine other explanations that sound plausible.

This is correct. There are various other things that could more specifically be the cause of Lester's improvement as he recovered from cancer. Lester's leg strength, unlike his command, is not something I can usefully observe. nonetheless-

How are we to choose from the infinite, plausible sounding explanations?

Through the use of our critical faculties. Lester, as your paragraph makes pretty clear, never said that Varitek was responsible for his improvement, but rather consistently gave Varitek credit for his success whenever it occured. Farrell, on the other hand, made a very specific, medically and physiologically plausible claim relating Lester's cancer recovery, leg strength, and command. I'm definitely ready to choose that as the most plausible explanation, though of course I can't really know. Hell, neither can John Farrell or even Jon Lester.

A simple 5/4/3 projection...

I realize the fallacy in just arbitrarily removing the worst month of the season but that month was SO bad I think it's a fair question to ask if something was going on there.

I'll guess most people here know this, but 543 or 321 systems are pretty crude. A good system will take into account month to month performances within seasons, especially the most recent season. If a guy's season, m to m, goes (OPS) 200 350 500 650 800 950, his projection for the following year had better be different than if his m to m went 950 800 650 500 350 200. A good system will also take into account anomolous months within a season.

If not, you've got a system that needs some serious tweaking.

I really don't think that what ark is saying is correct. I've never seen anyone show that month-to-month numbers, in the aggregate, make a useful improvement on stat projections.

Tango demonstrated that stat projections are fighting over a few percentage points of r-squared, as that's the most they can reasonably beat Marcel by, and Marcel doesn't use month to month.

AROM, how close is Edmonds to the magic 70 WAR?

Close, 66.5 as of my latest calculations. The whole system will be available on baseballprojection.com sometime soon, hopefully before the season starts. I've been working on the page designs today.

Edmonds doesn't have much chance of adding a few WAR though, since he's not getting any interest at all. Looks like he might be Loftoned.

of course I can't really know

It's ok to embrace this. A player's career is not a puzzle to be solved by finding an explanation for every blip in their record.

Tango demonstrated that stat projections are fighting over a few percentage points of r-squared, as that's the most they can reasonably beat Marcel by

That was based on a miscalculation, and he's retracted it.

I think ark is right, but I haven't seen any studies demonstrating it either. Maybe one that compared first-half and second-half numbers.

If there are two guys, first one goes 800 800 800 200 800 800, second one goes 700 700 700 700 700 700, I bet the first guy does better the following year. Of course that's an extreme case. It would be interesting to see how much better.

--

With respect to Varitek, my first question is whether the projection systems deal properly with 37-year-old catchers coming off the worst season of their career. I wouldn't be surprised at all if they systematically over-project this group. Aging curves are not well-established, selection bias is a huge problem, few catchers have ever been productive at that age, etc.

Granting for the sake of argument that the projection systems get the group right as a whole, the next question is how Varitek compares to the group. This is really unclear to me. You can point to diminished bat speed, but everyone in this group is going to have had problems. After all, they just had the worst season of their career. Some are suffering from chronic injuries, some are getting physically out of shape, you name it. I certainly don't have the confidence to say that Varitek is better or worse off than the average of these peers.

In a nutshell, projections aren't yet good enough to obviate the need for player-specific studies. Not that I know anything, but I believe the stat-oriented MLB teams spend a lot of time on player-specific studies as compared to general systems. (Of course, even if they do, that doesn't make them right.) But I'm skeptical of MCoA's claim that "decreased bat speed" adds a significant amount of information to what we already know from the statistical record. More precisely, it certainly adds information, but I have no idea whether it serves to make the "ideal statistical projection" worse or better, let alone by how much.

It's ok to embrace this. A player's career is not a puzzle to be solved by finding an explanation for every blip in their record.

Sure. What bothers me, though, is the tendency to treat whatever isn't included in aggregative projections as not worth pursuing. The fact that our knowledge will always be incomplete, always provisional and positional, should always be acknowledged, but should hardly cause us to stop learning about baseball.

There's sort of a larger thing here for me, that I think that a variety of trends in contemporary economics have pushed a notion of human society and behavior in which statistical models are misrecognized as the extent of the real - that is, what falls outside of these models is little worth pursuing - and I see an analogy here to the deep mistrust among baseball stat people of information that falls outside of aggregative statistical models. Obviously baseball and society are different, and baseball in a number of important ways is more open to this sort of modeling and projection, but I think that we need to recognize both the limits of particular kinds of tools, and that the limits of these tools is not equivalent to the limits of our knowledge.

What bothers me, though, is the tendency to treat whatever isn't included in aggregative projections as not worth pursuing.

Go read what you wrote to ark. Unless you have an enormous amount of free time on your hands, it isn't worth pursuing. Well, I suppose it's entertaining to make projections based on something beyond the numbers, in the same way it was entertaining to call Miss Cleo.

Sure. What bothers me, though, is the tendency to treat whatever isn't included in aggregative projections as not worth pursuing.

It's worth pursuing, but more and more here and elsewhere, people point to it with certainty--it's obvious to them that their own scouting, which has never been tested, is a valuable addition to the projection. Think about all the stuff we hear from actual professional scouts that turns out to be completely bogus. One that sticks out is that Blalock had a hole in his swing that would be exposed at the Major League level. But that's just one of many. And those are the pros. The rest of us, sitting at home, watching on TV, trying to decipher whether Tek's bat speed has dropped him from a 710 expected OPS to an 670 expected OPS don't stand a chance.

That, of course, is a far different thing from Lester's situation, or other cases like the ones I mention above, where there are more obvious factors that we can plainly see will affect players' abilities. On those, of course, I would have no problem adjusting my expectations.

What bothers me, though, is the tendency to treat whatever isn't included in aggregative projections as not worth pursuing.

I agree. It's easy to fall into this trap. The stereotypical stathead circa 2001 underrated fielding because there weren't reliable stats to measure it.* If you pressed him on it he would say, I know fielding is important but I can't keep track of it, while I can keep track of hitting. So I'll focus on what I know how to measure. In practice, though, this translated into undervaluation of good defense.

Surely the same scenario is going on today, with different examples. I would suggest that catchers' pitch-calling is something we all claim to understand the importance of but minimize in practice.

I think that we need to recognize both the limits of particular kinds of tools, and that the limits of these tools is not equivalent to the limits of our knowledge.

Yes. Though, if Tango's R^2 calculation had been correct, it would have meant that Marcel _was_ nearly equivalent to the limit of our knowledge. For this reason, we all should have distrusted the calculation and found the (simple) mistake much earlier. Easy for me to say now, I guess.

Edit: to agree completely with Darren in #27.

Edit 2: I forgot the asterisk. Zone rating was available in 2001, wasn't it? Dial has shown that it gets you most of the way to the fancier metrics. So maybe it was a communication issue: Dial had figured this out, but he had a hard time piercing the stathead conventional wisdom. Or maybe I'm being too unfair to the people who were around back then -- I didn't get into this stuff until 03 or so.

Dial was saying that in 01?

Wasn't there some piece about how Dial got into analysis of defense in the late 90's in order to prove that Rey Ordonez was actually a pretty good player? And how that led him to compare hitting and fielding on an apples-to-apples basis by converting ZR into runs? Then it turned out that Ordonez wasn't any good after all, but Dial had the method down. I'll see if I can find the article.

The piece I was thinking of was this one by Sam Hutcheson, which was linked from BTF at the time. I found a Dial article from 2001 laying out his method: here. You'll never guess who the first commenter was...

I realize the fallacy in just arbitrarily removing the worst month of the season but that month was SO bad I think it's a fair question to ask if something was going on there.

The stomach flu, followed by the stomach flu again.

There's sort of a larger thing here for me, that I think that a variety of trends in contemporary economics have pushed a notion of human society and behavior in which statistical models are misrecognized as the extent of the real - that is, what falls outside of these models is little worth pursuing - and I see an analogy here to the deep mistrust among baseball stat people of information that falls outside of aggregative statistical models.

Finally! Someone joins me on this side of the fence... though I wouldn't peg it to contemporary economics. People just want the world to be simpler than it is; once they have a model that appears to work, they no longer feel the need to evaluate its fitness, whether the underlying assumptions have changed, nothing. Happens all the time.

I've probably said this 100 times, but I'll say it again: if reality doesn't match expectations, it's worth considering if the expectations are wrong.

I would tend to think that Varitek is going to hit somewhere in the 650-680 OPS range, a season roughly equal to his last season, a bit worse.

This isn't intended as a reply to your projection specifically, but in general... If I recall, they're planning to use Varitek less this year. I suppose they might sit him more in cases where he's less likely to hit well. If so, his ratios might get a lift, and his counting stats will drop, from whatever one might otherwise project.

EDITed for clarity.

You'll never guess who the first commenter was...

Oh, that's grand, right down to the variation on the username.

Tek ought to only start against finesse pitchers and lefties.

once they have a model that appears to work, they no longer feel the need to evaluate its fitness, whether the underlying assumptions have changed, nothing.

If I had a projection system I'd be perfectly willing to overhaul it if you could show me I was making a mistake. If you can't prove that it's making a mistake, then I'm not going to assume that the system is perfect, but I am going to assume that you and I both have no idea what the mistakes it makes are.

I would tend to think that Varitek is going to hit somewhere in the 650-680 OPS range, a season roughly equal to his last season, a bit worse.

If Varitek is projected at .700 OPS and hits .680, that will be one of the better projections for the year. If he hits .650, it will still be far from one of the worst.

Ha ha ha. Don't I look like some fool. I guess I didn't realize it was that long ago. Did it predate UZR?

It's worth pursuing, but more and more here and elsewhere, people point to it with certainty--it's obvious to them that their own scouting, which has never been tested, is a valuable addition to the projection. Think about all the stuff we hear from actual professional scouts that turns out to be completely bogus. One that sticks out is that Blalock had a hole in his swing that would be exposed at the Major League level. But that's just one of many. And those are the pros. The rest of us, sitting at home, watching on TV, trying to decipher whether Tek's bat speed has dropped him from a 710 expected OPS to an 670 expected OPS don't stand a chance.

At the same time, there is not a single major league organization which does not use scouts in player analysis. There never has been and there never will be. Certainly our baseline, given this, must be to assume that such information has real utility.

I'd also like to know where you got this notion of "certainty". I think I've spent like 30% of my writing in these threads talking about humility and the limited nature of observation. I simply believe that statistical models for projection are also deeply limited, and so occasionally I think I have something useful to add.

Scouts are just a bunch of fat guys sitting around saying stupid stuff--haven't you read Moneyball?

I'm not arguing against scouting having value. My point was that even the real scouts are far from perfect; I would guess that us amateurs are not very good at all.

As for certainty, you always write with humility but others seem extremely sure that they've captured something with their observation that the projection misses. I think they're wrong--that's all.

Varitek sucks.

How'd Varitek's BABIP go last year? According to the Hardball Times (or some other stat site, think it was them), Castillo's was way lower than expected based on his line drive/ground ball/fly ball output.

Tek's Props was 40 points above his OPS, just like in 06 and 07. I don't think he was unlucky last year.

If there are two guys, first one goes 800 800 800 200 800 800, second one goes 700 700 700 700 700 700, I bet the first guy does better the following year. Of course that's an extreme case.

I think it really only works in extreme cases.

800 800 800 200 800 800 really looks like an 800 hitter who had an injury (or some other issue) that one month.

750 700 750 400 800 750, really just looks like a 700 hitter.

Here's Dioner Navarro's OPS+ the last 4 years:
94, 79, 70, 98.

What that doesn't show is that in May 2007 he hit .143/.167/.200, followed by .185/.264/.277
33 catchers from 2005-08 have 1000+ PAs, Dioner's OPS+ that span is 85.
He IS a better hitter than that. May-June 2007 was just off the charts bad- and really have no value in projecting him for 2009.

That's an extreme case- 97.5% of the time I do not think you can cherry pick stats like that.

For instance:
Julio Lugo had consecutive seasons of 98, 94, 105, 96
Then in May 2007 he hit .209/.241/.336, and June he hit .089/.170/.139 (!!!YIKES)
He hit .280/.322/.406 post ASB in 2007 (his OPS+ for the year ended up at 65), at the time I assumed you could throw his bad months out and he'd be back to a 95 OPS+ in 2008-
nope, he reached 78.

Tek in 2009?
I'd cap him at his 3 year average of .238/.336/.393
Given his age and position, I wouldn't be all that surprised if he couldn't reach last year's marks of .220/.313/.359

So to average those out, you think he'll repeat 08, more or less?

So, in his last two at-bats, Varitek homered off a James Shields fastball low and in, and hit a rocket to center off a Joe Nelson fastball left out over the place. Batting lefty both times.

His new, simplified swing is working!

I'm sure you're poking fun at the tendency to take opening day results too seriously, but Varitek's day was even better than the results of those two at-bats. He made good contact in the first two as well, getting the ball in the air, and not that lazilly. Also, I could be wrong, but I didn't see him take an utterly horrible swing all game, e.g. the so late the ball is already in the mitt, the weak wave at a curve outside, the eye-high fastball. I know the tendency is to only remember the terrible swings that result in strike three, but last year these 'tek specials could be found anywhere, often 2 or 3 in a PA. It really was a helluva (1) game he had.

I'm sure you're poking fun at the tendency to take opening day results too seriously,

Carlos Pena is on pace for 640 strikeouts!

but Varitek's day was even better than the results of those two at-bats. He made good contact in the first two as well, getting the ball in the air, and not that lazilly.

Agreed.

Also, I could be wrong, but I didn't see him take an utterly horrible swing all game

That was Mike Lowell a couple of times yesterday. And he seemed to run slower than usual. I'm concerned he's not quite right yet.

Varitek looks awesome.

On the other hand, Jed Lowrie looks a mess. I'm wondering how soon we'll be demanding Chris Carter in clutch situations. at the game on Wednesday, I thought they were going to Carter.

Lugo is the answer.

Of course, if that's true, it must be one horrible question.

To clarify on Varitek, I think that the beginning of the season is a time when baseball is in a bit of disequilibrium - a guy like Varitek who had a terrible swing from the left side picked up a scouting report that he can't hit a good fastball, then he comes into spring with overhauled mechanics from the left side, and he can hit a fastball as a lefty with some competence. So for a week or three, he'll be able to put up good numbers before the scouting report changes and in all likelihood, things go back to basically normal - hopefully a baseline that looks a good bit like his projections.

On Lowrie, I think it's important to recognize that this is a guy who projects to a 730-750 OPS. That's obviously better than Lugo. That's solid for a shortstop. That'd be the best shortstop the Sox have had since Nomar. But it's not a projections of particularly great offensive contributor, and I can definitely see a lefty 1B/LF with the platoon advantage being a better option in certain situations.

However, I think that Francona surely has other reasons for batting Lowrie in the clutch - with the possible upgrade to a pinch-hitter being relatively small, I think there's value in showing that you trust the young kid to be a big leaguer and hit in big league pressure situations. I just hope he starts paying it off.

On the other hand, Jed Lowrie looks a mess. I'm wondering how soon we'll be demanding Chris Carter in clutch situations. at the game on Wednesday, I thought they were going to Carter.

THREE GAMES.

However, I think that Francona surely has other reasons for batting Lowrie in the clutch - with the possible upgrade to a pinch-hitter being relatively small, I think there's value in showing that you trust the young kid to be a big leaguer and hit in big league pressure situations. I just hope he starts paying it off.

Remember that hit in the ALDS last year? Also, THREE GAMES.

Lugo is the answer.

Of course, if that's true, it must be one horrible question.

Dis made me laugh, and ponder what the question would be.

"Which baseball player do you think of when you pray at the porcelain altar?"

To clarify on Varitek, I think that the beginning of the season is a time when baseball is in a bit of disequilibrium - a guy like Varitek who had a terrible swing from the left side picked up a scouting report that he can't hit a good fastball, then he comes into spring with overhauled mechanics from the left side, and he can hit a fastball as a lefty with some competence. So for a week or three, he'll be able to put up good numbers before the scouting report changes and in all likelihood, things go back to basically normal - hopefully a baseline that looks a good bit like his projections.

I had a mild disagreement with some of this, but it occurs to me that your argument and mine end up in the same place. Never mind.

Lowrie, keep in mind, delivered an extra-inning HR in late August in Toronto, and the hit that won the ALDS. He had another game-winning hit in Bay's first game with Boston, though I wouldn't cite that weak shot as a sign of his clutch ability. He has come through in big-league pressure situations. He didn't this week, and looked bad; like you I hope it isn't a sign of things to come, but I'm not too worried about him. Not yet, at least.

Again, I'm more worried about Lowell.

Lowrie had some moments last year and I'm optimistic about him but he has been slumping since late-August of last season. He didn't hit much in the last month and a half of last season including the playoffs.

Small sample size! Small sample size!

(Though Lowrie whiffing on an 89-MPH fastball down the middle was pretty miserable.)

It is true that the Red Sox have played three games. Thanks for clearing that up.

As I said in my longer post, the issues with Lowrie are not merely ones of observation over a tiny sample - though I won't stop doing that, and I think I've made pretty clear why - but also a function of the more general fact that Lowrie is not a particularly strong hitter. And when a not particularly strong hitter hits for himself in late and clutch situations in two games in a row (and fails both times), it's a topic for conversation.

I think one point I wanted to make is that the Jed Lowrie of the projection systems is the sort of player you want to consider pinch-hitting for. And Francona's general tendency is not to pinch-hit, I would assume out of a combination of knowing that pinch-hitters see a notable decline in effectiveness from their baseline skill, and because his managerial style requires him to show faith in his players.

To be clear, I hope it's obvious I'm not saying that Lowrie is a choker, resembles a choker, is someone I'm worried is choking. I would say that we're not in Jed's head, and I think that presuming that because he has had a couple of clutch hits in his career, he doesn't need anyone to show faith in him, is not particularly convincing. Francona, as a manager, generally tries to show faith in everyone who earns their way into his lineup, whether they be rookies or 10-year veteran catchers in decline.

I'm not sure what Lowrie's platoon splits are like as a minor leaguer, but I wonder if he wouldn't be better served by just batting as a righty all the time. His admittedly very small sample size major league splits are a Varitekian lefty 307/342 v. righty 409/525. Granted, batting RH v RHP probably wouldn't do much for Lowrie's confidence, and given that he would have to learn a whole new approach to hitting RHP, is probably not something to try this year unless things go horribly wrong.

Lowrie's minor league splits are pretty much even, except for a somewhat odd extra pile of walks from the left side and a bit more power from the right side. The sample there is pretty small too, but overall, I don't think there's much statistical reason to think that Lowrie's got a notable platoon differential.

RHB: 387 AB, 284/351/465
LHB: 883 AB, 288/393/439

I got the numbers from minorleaguesplits.com - I can't find any way to create a permanent link to the data.

Francona, as a manager, generally tries to show faith in everyone who earns their way into his lineup, whether they be rookies or 10-year veteran catchers in decline.

Basically true, but I bet there are small differences based on the usual factors - veteran status, contract size, star power - that we may never really know for sure due to the small overall occurrence of pinch hitting.

Also consider that even if Francona had equal confidence in Lowell and Lowrie, he might believe that the two players' confidence in themselves would be differently affected. When Lowrie gets hit for he can tell himself that the manager thinks he is not yet a great hitter; when Lowell gets hit for, he gets the message he is no longer a great hitter. That's probably a big difference to professional athletes.

Also also consider that Lowrie probably hasn't earned his way into the lineup in Francona's mind to the extent that Lowell (and even Ellsbury) have, since Lugo's injury ended the alleged ST competition early.

In short, I think Lowrie is a lot more likely to be hit for than anyone else in the lineup, and that has nothing to do with projections (where Varitek may look worse) or temporary slumpiness scouting (where Lowell looks equally bad).

Update:

Tek:226/327/440
Castillo: 292/402/377, -2.2 UZR/150.

Update:

Tek: 235/321/429
Castillo: 304/383/370
Ortiz: 208/318/300

2 mashed potatoes for the C tonight...

and a toast to the big man! first of many for David I hope.

Why has Sox Therapy been pretty much dead all season?

Why has Sox Therapy been pretty much dead all season

Well that effing Darren went and got a life and can't continue to feed this monster....that bastard!

And Tek, 2 homers today. slg. over .500!

Someone else must have a set of keys besides Darren, right?

Why has Count the Ringzzzzzzzzzzzzz been dead for 42 years?

I wish I had posted a prediction on Vagitek in here. Despite his shockingly non-horrible start, I'm still of the opinion he is quite cooked. He may not get another hit the rest of the season.

Someone else must have a set of keys besides Darren, right?

I wish someone did. It was always nice to have Sox Therapy around, always talking baseball here, none of that other crap that's fills up the newsblog.

Why has Count the Ringzzzzzzzzzzzzz been dead for 42 years?

I asked for the keys to Count the Rings maybe 6 months ago. I didn't get a reply.

You'll get it as soon as the Great 28 simulated season ends.

Final results:

Varitek

Proj: 228/326/381, 1.6 WAR in 463 PA
Real: 209/312/391, 1.3 WAR in 423 PA

Castillo
Proj: 275/360/347, 1.2 WAR (-5 UZR) in 536 PA
Real: 304/389/347, 1.5 WAR (-11.4 UZR) in 576 PA

Whoops a day too early. Whatev.