Think fast. The deadline to tender unsigned players a contract is approaching. A key member of your bullpen is coming off a down season. What will you do?

If you're Pittsburgh Pirates GM Neil Huntington, such decisions are part of your job. 

If you're talking about the Matt Capps of '07 or '08, that would be very, very difficult to replace. He's probably not somebody we non-tender. The second half of '08 and into '09 ... it's not that hard to replace a reliever with a 5.00 or 6.00 ERA.

With Matt Capps, it seemed to come down to marketability. Seems that the market was impacted by some news leaks.

In that event, Huntington was asked, why was Capps not traded before that, given that the team surely had its internal valuations on him well before the past week or so: "We were working on it, and we had multiple conversations that disappeared when the media report came out.

Huntington was more explicit in his reasoning during a recent MLB.com chat:

The decision to not tender Matt Capps was a difficult decision. While it is obvious Matt's 2009 performance was not his best, we did feel there were indicators that he would have a better season in 2010. The arbitration process aggressively rewards saves, home runs and wins while not always properly accounting for the metrics behind those numbers. Despite wanting to retain Matt and making an aggressive offer that we believed would be at or near his free-agent value prior to the tender deadline, we felt that the risk of an arbitration award at a substantially higher amount was not a good business decision for us. We may be right or wrong on Matt's free-agent value (and his 2010 performance), and we may be right or wrong on the performance of the pitcher(s) and/or player(s) we re-allocate the money toward, but we felt that it was the right move for us. Obviously, we would have preferred to get something of value in trade for Matt, but given his track record beginning in the second half of 2008, his trade value was limited throughout the summer and again this offseason. We wish the best for Matt and are certainly open to retaining him if we are able to find a common ground with his agent.

Emphasis my own. Note Huntington avoids putting blame on the media this time around. At least directly.

I think they made a mistake, and Capps will be valued relief pitcher somewhere in 2010. Why am I so confident that a pitcher the Pirates had no use for is in line for a good job with another club? First, the list of teams in on Capps is quite long. Second, he's not pitching any better or worse now than he was a couple years ago.

PITCHf/x Indicates Little Has Changed

I realize Capps gave up far more walks and home runs than expected in 2009. Results are results. But what is reasonable to expect going forward? I'm not going to project Capps' 2010 line, but I am going to show you that, beneath the surface, the later vintage Capps (using Huntington's parameters) was just as effective as the earlier version.

PITCHf/x gives us pitch-by-pitch data on what the pitcher threw, how fast it went, where it went and what the batter did with it. Further, within Gameday (which includes the PITCHf/x data), we also get batted ball types and a lot more. PITCHf/x started in 2007, but wasn't in every big league park until 2008. Pittsburgh's system came on line in September 2007. This leaves some holes in the data, so keep that in mind. For the most part, I'm not all that interested in the individual pitches. But my analysis starts there and rolls-up into the top line.

The Basics: Capps is, essentially, a three-pitch guy. Fastball, slider and change-up. He throws mostly four-seam fastballs, but also throws a two-seamer. As noted earlier, he also has a cutter.

For the purposes of this discussion, I'll go with three pitches, lumping fastballs together and cutters with sliders:

Fastball: 92 mph; Slider: 85; Change-up: 87

Nothing earth shattering, a fairly typical arsenal featuring reasonable, but not spectacular, power.

Now that we know what he's got, let's split it up. Huntington alluded to a second half decline in 2008. Capps was out of action for a few weeks after the All Star Break, so we can conveniently split his PITCHf/x data (which starts in 2007, and is incomplete for that year) into "before" and "after", and line-up close enough in terms of sample sizes.

Most of these stats are somewhat self-explanatory, but I'll run them down just in case ... strikes are pitches in a two foot wide zone, based on each hitter's own top/bottom zones. Whiffs are misses per swings, chases are swings out of the zone, watches are takes in it, B:CS are umpire calls, GB/LD/FB/PU are Gameday stringer assigned batted ball types, HR/F+L is simply homers per flies+liners, rv100 are run values vs MLB average (2007-2009) per 100 pitches (negative values being better for pitchers), rv100E is the same thing but using batted ball types instead of actual hits/outs - takes away some fielding and park luck, along with contact quality unfortunatly. Neither rv stat I use is park adjusted at the moment, and AL/NL are not separated so apply a mental penalty to NL pitchers. 

One could argue Capps was a better pitcher, with poorer luck, in the supposed down swing of the "After" period.

In case you're wondering, fastball velocity was 93.5 "before" and 94.1 "after". I also split Capps' data into 2009 only and 2007/2008. In that case, the sample sizes were nearly identical, as were the performances.

Where To?

The latest news, as of this writing, is about Capps narrowing down the field of suitors:

[R]ighty reliever Matt Capps' agent, who found himself fielding multiple suitors, said Capps will narrow the list to the five most-serious.

By "serious" I'm assuming multi-year offers and opportunities to close. So far, the Nationals are on the list. Given my own loyalties, I would not be displeased with Capps working set-up innings for the Cubs for the next two years, if the price is right.