The Angels made the largest relief signing of the offseason so far, committing $15.75MM over three years to right-handed sidearmer Joe Smith.  Any sizeable commitment to a reliever will be poorly received with sabermetric analysts, but did the Angels at least get the top setup man Smith's contract suggests?

Smith may have been paid based on his ERAs for the Indians in the past three seasons: 2.01, 2.96, and 2.29.  Fangraphs wins above replacement, which uses fielding independent pitching (FIP) in its calculation, does not credit Smith for those ERAs, giving him 2.0 WAR over the three seasons.  The main components of FIP are strikeouts, walks, and home runs allowed, and Smith has excelled in only one of those.  Given his strong groundball tendencies, Smith has allowed just ten home runs in 197 innings dating back to 2011.  Wins above replacement can also be calculated using runs allowed instead of FIP, and that figure credits Smith for a healthy 4.7 WAR over his last three seasons.

The Angels aren't interested in paying Smith for what he did for the Indians; he's getting $5.25MM per year from the Halos in hopes of continued sub-3.00 ERAs for 2014-16.  To see how likely that is, we typically turn to estimators like FIP, xFIP, and SIERA, which predict future ERA better than ERA does.  Using Smith's 2011-13 peripheral stats, those estimators spit out figures in the 3.33-3.68 range, well above his actual 2.42 mark.  The estimators are not crediting Smith for one potential skill, though, and that is his consistently low batting average on balls in play (BABIP).

Smith's BABIPs the last three years were .258, .253, and .282.  His career mark is .272.  Compare that to the average reliever, who was at .291 this year.  Smith seems to be better at keeping his BABIP low than other relievers, which is why he's consistently allowed fewer than eight hits per nine innings since 2008.  Smith's career BABIP against right-handed hitters is .259, versus a more normal .298 against left-handed ones.  This makes sense: he's a right-handed sidearmer, and he is able to induce weak contact against same-handed hitters.  This apparent skill has been magnified by his usage, as Smith has faced right-handed hitters two-thirds of the time in his career.

In 2013, 54 non-closer right-handed relievers pitched at least 60 innings, including Smith.  As a group, they faced right-handed hitters about 55% of the time.  In addition to the aforementioned low BABIPs, Smith has been adept at getting right-handed hitters to hit groundballs.  In 2011, Smith began the transition away from being a full-blown right-handed specialist, but he was still shielded from lefty hitters in 2011-12, magnifying his skills against righties and aiding his ERA.  Only in 2013 did Smith graduate from right-handed specialist to general setup man: he faced right-handed hitters only 50.6% of the time.  Indians manager Terry Francona let Smith face left-handed hitters 128 times, easily the most in his career.  The promotion was overdue, as he hadn't been hit too hard by southpaws since 2010.

$5.25MM a year is setup man money.  The Angels invested in Smith after he posted a 2.29 ERA in 63 innings, truly in a setup role for the first time in his career.  However, Smith's low ERA was not due to the usual factors, a low BABIP and a high groundball rate.  His .282 BABIP was his highest since 2007, and his 49.1% groundball rate was the lowest of his career (the latter owing to his facing more lefties).  Instead, a big factor in Smith's 2013 success was his left on base percentage of 86.3%.  Among relievers with at least 60 innings, Smith ranked 14th in baseball.  Almost everyone ahead of Smith on that list struck out more than 27% of batters faced, while Smith was around average at about 21%.  There's no reason to expect Smith to be much better than the relief league average LOB% of 75% going forward.

If ERA alone doesn't convince you Smith is a top setup man, then it's hard to find a particular standout skill he displayed in 2013.  He's not a strikeout guy, he doesn't have great control (especially versus left-handed hitters), and his groundball rate and BABIP weren't anything special this year.  His ERA was low because he stranded 86% of his baserunners.  The Angels probably don't have a reason to expect that to be repeated, so they're left with a guy whose only above average skill might be inducing groundballs from right-handed hitters.  They didn't need to spend $15.75MM to find a guy who can do that, with Matt Albers and Jamey Wright also on the free agent market.  That's not to suggest Albers and Wright are as good as Smith, but with limited payroll flexibility and a need for two starting pitchers, this signing was a questionable allocation of resources for the Halos.