Kyle Farnsworth‘s free agent decision is expected to come soon, with the big right-hander having narrowed his decision down to the Tampa Bay Rays and the ever-present Mystery Team. Farnsworth lost his closer’s spot in Tampa Bay to the phenomenon that was Fernando Rodney, and his injury-limited 2012 season was somewhat of a disappointment: in 27 innings, Farnsworth allowed a 4.00 ERA and a 3.39 FIP; his 4.67 BB/9 was his highest since 2001.

A ballooning walk rate and decreasing velocity — his average fastball velocity fell from 96.7 MPH to 95.4 MPH* — gives ample reason for pessimism and has likely limited Farnsworth’s options in free agency. But, at least by the numbers, there’s still reason to believe The Professor can be a sharp contributor to a bullpen in 2013.

**All PITCHf/x data in this post from BrooksBaseball.net*

As R.J. Anderson pointed out at The Process Report, most of Farnsworth’s poor results came from one disaster outing, magnified by his shortened season:

Although his overall numbers were mediocre (a 4.00 ERA and 1.79 strikeout-to-walk ratio in 27 innings) much of the damage occurred in one outing; a July 4 appearance against the Yankees. Stripping away any playerâ€™s lowest moment is admittedly a little cheap (not to mention often mendacious), but in this case one rotten apple is spoiling the bunch.

In the 32 appearances following Farnsworthâ€™s meltdown, he completed 25 2/3 innings, struck out 22 batters, walked nine (two intentionally, which makes for a strikeout-to-unintentional walk rate of more than 3.00), and compiled a 3.16 ERA.

The July 4th outing in question was just Farnsworth’s second of the season; he walked four of the five batters he faced (he struck out the fifth) and he was charged with three earned runs. As Anderson mentions, his K/BB looks much prettier in his subsequent outings, and the particularly concerning walk rate comes down to a downright solid 2.45 per nine innings.

But the true reason to believe isn’t in 27 innings of performance, whether mediocre or slightly above average. It’s in the sinker Farnsworth added to his arsenal in late 2009, now the driver behind his excellent 55.6 percent ground ball rate and the pitch he’ll have to ride throughout the rest of his late 30s.

In many respects, Farnsworth’s sinker was at its best in 2012. Its .219 batting average and .250 slugging percentage allowed were its best marks ever. It induced 70 percent ground balls, the highest rate since he introduced it in 2009 (77 percent). Most remarkably, only 31 percent of Farnsworth’s sinkers went for balls — his previous low was 35 percent in 2010.

Increased control with the sinker and more ground balls when it is put in play go a long way in explaining how Farnsworth posted a ground ball rate nearly five percent better than his previous career high. It’s only 27 innings, but there’s enough here to believe it’s at least partially real.

First of all, Farnsworth picked up his arm angle (perhaps as a result of the injury) and the result was more vertical movement on all of his pitches. But the gap widened between his four-seam fastball and his sinker — the largest since he introduced the pitch on a significant level four years ago:

Secondly, although 145 pitches doesn’t sound like many, it’s enough to get a good read on his control from a balls and strikes perspective. The following formula gives us the standard deviation for a binary statistic — like the rate at which a pitch is a ball:

In this case “B” is the ball rate on Farnsworth’s fastball. With 145 thrown, the standard deviation comes out to 3.8 percent with a mean of 31.0 percent; for his career, Farnsworth has thrown 1,175 sinkers with an average of 36.7 percent balls, resulting in a standard deviation of just 1.4 percent. A one-tailed significance test suggests over 90 percent statistical significance (p = 0.0796). Perhaps some of this is the Jose Molina impact behind the plate, but even in 27 innings there is somewhat strong evidence Farnsworth can spot his sinker for strikes better than ever (and at the least as well as ever).

As Farnsworth ages, he’ll need to continue locating his sinker consistently. His fastball velocity is likely to keep declining, and he’ll need other weapons aside from a fastball capable of touching 98 or 99 MPH. The sinker doesn’t have the blaze the rest of his arsenal features, but it provides him a solid, consistent way to keep hitters off balance, the ball out of the air, and the count in his favor. He showed this ability last season, and it should give any signing team confidence in his ability to perform in the later innings.