I guess I'm just trying to think of a way to maximize the value of the pieces we have in the infield.  I guess there really isn't any point in having this discussion until we see Everett in spring training.  If his range is the same as it was pre-injury, I buy the "his glove outweighs his bat" argument and could see slotting him in as the starter.  If his defense is merely very good as opposed to outstanding, we may have to begin thinking of him as a $3 million defensive replacement. Going by these projections, Casilla potentially represents 42 points of OBP over Everett.

Then Steven Hammond, Jamey Price, and Adam Peterson.  Peterson was awful in his age 25/26 seasons and dropped off the map after that.  I don't know much about Hammond or Price.

It's worth noting that Humber's beta is 1.28, which is fairly high, implying that PECOTA foresees a wide range of possibilities for Humber.  He also has a fairly high 29% chance of improvement coupled with a 56% chance for collapse, so PECOTA sees this as sort of a "make or break" season for Humber.  I also don't see Humber being that bad this season, but the HR rate from last year is pretty worrisome, even for the PCL.

Anyway, regarding Morneau.  His park-adjusted forecast is .273/.343/.491 and his career-to-date line is .276/.340/.498.  His raw projection is .271/.339/.465.  The Dome has been playing somewhat like a pitcher's park lately, and that knocked ~25 points off of his SLG and 1 point off of his OBP.

Conventional wisdom says guys improve from ages 22 to ages 28, have a plateau of a year or more, and then decline.

Everything about PECOTA stems from the fact that it determines a list of the players most comparable to that guy, and then seeing how those guys did the next year.  So if PECOTA isn't projecting big leaps forward, it's because the hitters that were most similar to him didn't take a big step forward.  (For most hitters, we're talking about a group of 60-100 hitters that compare well enough to be included at some level.)

Morneau's most comparable hitter is Hrbek (the second year in a row that has happened), and he had a pretty typical career arc.  If Morneau does what Hrbek did, he'll have about a 140 OPS+ next year.  Morneau's second most comparable hitter is Greg Walker, who hit .277/.345/.493 with a 122 OPS+ as a 26-year-old (very similar to Morneau's '07 campaign.)  Walker's career promptly fizzled out after age 26, though, never hitting above a 110 OPS+ again.

Those are the first two guys--PECOTA just keeps going to see what the whole collection looks like.  Overall, the group apparently didn't show much improvement over their career numbers.  I think Morneau's a decent bet to beat his career numbers, but I guess I'm skeptical that he'll be a whole lot better than his career numbers until he can show the ability to lay off a few more borderline pitches and take a few more walks.  Until he does that, I think we'll see a lot more seasons around .280/.345/.500, which is decent, but rather unspectacular for a first baseman.

As for Mauer, I'm a bit more confused, but my guess is that with the decreased playing time, and being compared most closely to catchers, he's getting compared with a bunch of guys who became injury-prone and who maybe had a flukishly good season followed by a regression to what they could really do.  (For instance, I think Mauer's capable of repeating 2006 from a value standpoint, but I'm sure there are more players who had a season like his '06 and never repeated it than there are players who had his '06 and did it over and over again.)  The full projections aren't available yet, so I don't know Mauer's similarity score.  Last year it was 17, which in and of itself is pretty low, and means that there weren't many players that similar to Mauer, and PECOTA was basically issuing a warning stating "I'm not so sure about this one."  I'd guess his similarity index is pretty low, but I'm not sure.  Mauer's such a freak for a catcher that PECOTA has to go to other positions to get comparable hitters, which probably also reduces the reliability of his forecast.

I still think those two projections are useful, as it may help make expectations for the season a little more reasonable, but I'd guess that other projection systems might be a bit more bullish on those two.  ZiPS has Mauer at .315/.404/.458 and Morneau at .283/.355/.527, for instance, and neither ZiPS or PECOTA is substantially more accurate than the other.

How come no one talks about this free agent loss?

There is NO way, NONE, that Morneau and Mauer hit that poorly this year.  I will bet the house on either of those.

Um, I hope you don't have a wife or kids, because that's certainly not a bet I would take.  I like the chances of those guys beating those lines, but there's certainly a chance that they could underperform those lines.  Morneau hit .243/.318/.384 over his last ~300 AB, so I'm certainly not betting any real estate on him beating his career average this year.  A Jackson or two?  Sure.  But not a house, and certainly not my own house.

Even if you're right and PECOTA is drastically underestimating Mauer and Morneau, that's just two hitters out of about 950 that were projected.  Two bad forecasts does not make for a useless system.

Who ran PECOTA this year?  A White Sox fan?

That is pretty rich considering that last year, White Sox fans were screaming bloody murder over how "unfair" PECOTA was being to them, and how they were going to be way better than that, and then they finished almost exactly where PECOTA projected them.

It's an algorithm designed to minimize the deviation between a player's projected performance and his actual performance in the next season.  It doesn't have feelings, and it's certainly not biased against the Twins.