Looking back at the 2011 season, it seems like Brian Roberts was out all year. It's surprising to see that he actually accumulated almost 200 plate appearances. Roberts missed much of the 2010 season as well, but the big injuries there (mainly his back, but also the abdomen) looked mostly resolved. In a May game this year though, a head-first slide into first-base (trying to beat out an infield single - I hate that play) resulted in a return* of concussion-like symptoms that ended up keeping him out for the rest of the season (and it's unclear if he'll ever really be back).
* Brian also missed a bit of time at the end of last season due to a concussion when he hit himself in the head with a bat, so this incident may have built off of that one.
I thought Roberts might miss a little time with lingering issues from 2010, and have some decreased performance as a result, but things turned out oh so much worse:
Maybe things would have improved with a full season - probably - but Roberts didn't really produce even when on the field. He was less patient than usual, but kept from striking out by doing a better job of putting the bat on the ball (though his contact rate has always been quite good). The low BABIP is probably partially attributable to Robert hitting the ball in the air a ton (flyballs generally go for hits less often than groundballs), but a mark that low is almost surely due to quite a bit of poor fortune.
The home run power looked pretty normal (he was on pace for 9-10 in a full season), but the extra-base hits were still down (even accounting for the lower BABIP). Only 24% of Roberts' in-play hits ended up getting stretched to doubles and triples. For his career it's 29%, and in the couple years before his injury-marred 2010 it was 34-35%.
The base-stealing efficiency was still present (6 for 7), but he seemed to run less often; stolen base attempts over singles plus walks was 19% (the lowest that rate has been in a season for him; career it's 23%). That's good in my book - the extra base is probably not worth the increased injury risk*.
* Actually (calculation unitalicized for easier reading):
If a stolen base is worth about 0.20-0.25 runs on average, and Roberts is successful 80-85% of the time, then a stolen base attempt for him is worth around 0.18 runs. Assuming Roberts is a 2 win player over 150 games, then that's ~0.013 runs per game. If on a stolen base attempt his chances of injury are as follows: 25% chance of hitting the 15 day DL, 15% chance of missing a month, 7% chance of going on the 60 day DL, and 3% chance of missing 100 games, then the run value of the injury probability is about 0.19 (just a touch higher than the value of the base).
I somewhat doubt the actual injury risks are that high (50% total), so perhaps it's OK to roll the dice.
With only 1.6 fWAR over the last two seasons and a great deal of uncertainty about his future, Roberts' contract extension now looks much worse than almost anyone could have imagined (and it wasn't expected to turn out that well to begin with). The two years and $20 M left on it probably won't provide any significant return. Obviously, that's not the biggest issue though (and I imagine insurance will cover part of the deal anyway). Brian Roberts was a major part of the Orioles for many years, and the way things have gone recently is very unfortunate. Hopefully he can recover well enough for there not to be any serious lasting effects for him, even if he never makes it back onto a baseball field.