Another day, another implosion from the Twins rotation. Last night it was Scott Baker. On any given night, Nick Blackburn or Kevin Slowey are putting the Twins in an early hole. So naturally, many of us have pushed for a trade to shore up the starting rotation. My question is simple, now that the consensus #1 trade target, Cliff Lee, is off the table, how much can the Twins improve themselves with a deadline trade? To answer this, we must first understand how bad our existing guys are performing, what sort of internal options we have in house, and how they compare to some of the options out there.
First, I'm only looking at this from a starting pitching perspective. At this point, I don't see any glaring holes in the lineup, especially after Justin Morneau comes back from his concussion. Third base has been the biggest question mark much of the season, but Danny Valencia has performed better than I expected, good enough for Gardy to sit Nick Punto on a regular basis. And in the bullpen, there are many options, including Anthony Slama, Kyle Waldrop and Pat Neshek. So I don't see a trade there either.
Second, I'm assuming that Nick Blackburn is going to be sent down to Rochester by the deadline, replaced in the rotation by Brian Duensing, with Jeff Manship taking over the long relief role. So any consideration of trades will involve replacing either Scott Baker or Kevin Slowey, who would likely move to the DL to accommodate a trade.
Much more after the jump.
Step One: How Bad Are Our Current Guys?
Let's just say that Baker, Blackburn and Slowey have not been very good over the past couple months. Over their past 10 starts apiece:
|Pitcher||IP||H||R||ER||HR||BB||SO||Team W-L||Pitcher W-L|
These are some pretty ugly numbers, especially for Blackburn, who has a 9.38 ERA over his last 10 starts. It's not as if Baker (5.40) or Slowey (5.36) have been very good either. But they've been much better than Blackburn. Frankly, I find it surprising that the Twins have even managed to win 11 of the 30 games these guys have started. It's good to have some offense. But I digress.
Let's assume for the sake of argument that we've already replaced Blackburn with Duensing. That leaves Baker or Slowey who would potentially be replaced. You may remember that a while back I wrote an article that focused on "Making Win-Loss Record a More Useful Stat", using what I call "Expected Wins" (eW) for pitchers and teams based on the number of innings and runs allowed by a pitcher, as well as the run support he receives, normalized across the league. Looking at the same three pitchers gives us a good idea of the pecking order so far this season:
|Pitcher||GS||Tm W-L||eW-L (Pitcher)||Exp W-L (T)||eWPCT||e Tm Wins|
|Scott Baker||19||8-11||5.92 - 7.97||8.91 - 10.09||.469||76.0|
|Kevin Slowey||18||10-8||5.01 - 7.59||7.55 - 10.45||.419||68.0|
|Nick Blackburn||18||10-8||5.17 - 8.84||6.98 - 11.02||.388||62.8|
Based solely on the number of innings pitched and runs allowed (i.e., not considering run support), Baker has been the best of the three (.469 expected team winning percentage, projects t0 76 wins over a full season), but has the worst actual team record (8-11) to show for it. This is not surprising, considering that Baker is often matched up against the opponents' #1 starters. Blackburn has clearly been the worst, with Slowey somewhere in between. Since I'm already replacing Blackburn with Duensing, it would appear that Slowey is the prime candidate to be replaced. These numbers are in line with xFIP (3.84, 4.68, 5.11 for Baker, Slowey, Blackburn), so again nothing surprising. But I'm going to use that .419 winning percentage to determine how much the various trade options might improve the team.
Step Two: What Do the Twins Have In House?
Here's where it gets a bit difficult to compare apples to apples. The two obvious candidates on the 40-man roster are Jeff Manship and Glen Perkins. But neither has pitched very well in Rochester (5.11 ERA, 109 H / 81 IP for Manship and 7.00 ERA, 124 H / 91.1 IP for Perkins), so I have little confidence in either being an improvement over Slowey at this point. Ryan Mullins (4.60 ERA, 106 H / 90.0 IP) has been the best starter in Rochester, and he hasn't been great either, nor do I think he would be added to the 40-man roster at this point. Short answer: other than Duensing, the Twins have nothing in house that would help.
Step Three: What is Out There?
For the purposes of this discussion, I'm going to leave contracts and cost out of the equation and look simply at how much a number of pitchers could potentially improve the Twins. Also, due to my potentially being struck by lightning for even considering it, I'm not going to include Carlos Silva in the discussion.I stuck to some of the most common trade targets across the league and our comment boards.
|Pitcher||GS||IP||H||HR||BB||SO||xFIP||Exp W-L (T)||eWPCT|
|Roy Oswalt||19||124.0||100||11||33||117||3.42||9.79 - 9.21||.515|
|Brett Myers||19||129.0||126||11||39||93||4.00||10.22 - 8.78||.538|
|Ted Lilly||16||104.0||92||18||25||75||4.62||7.34 - 8.66||.459|
|Dan Haren||20||135.0||155||21||27||133||3.42||8.96 - 11.04||.448|
|Ben Sheets||19||119.1||123||18||43||84||4.48||7.41 - 11.59||.390|
|Jake Westbrook||19||113.2||123||13||37||67||4.35||8.37 - 10.63||.441|
No surprise, Oswalt and Haren are the two studs according to xFIP, which is generally considered to be a better predictor of future performance than ERA or win-loss record. Looking at xFIP and ERA splits, one would expect Lilly (4.07 ERA) and Myers (3.35 ERA) to regress at some point, while Haren (4.60 ERA) should see some improvement when fly balls finally stop leaving the yard at such a high rate (13.0% HR/FB).
Step Four: How Much Would They Help?
Here is where I am going to make a leap of logic and assume that expected team winning percentage correlates relatively well with ERA for these pitchers. For the nine pitchers discussed so far in this article, there is a solid -0.79 correlation between eWPCT and ERA, which makes sense since eWPCT is based on number of runs allowed and innings pitched, same as ERA. And we know that xFIP tends to be a better predictor of future performance than past ERA. Using a best fit line for a sample of 20 pitchers across the league, I determined a rough formula to convert xFIP into expected team winning percentage, which I will call "Derived Win Percentage", or dWPCT.
Remember, at the trade deadline, the Twins will have played 105 of 162 games, leaving only 57 games until the end of the season. Assuming that any starter we trade for would pitch one of the first couple games in August, that's only 12 starts, or around a third of a season. The "Wins Added" in the table above shows the number of additional wins the Twins could expect based on the difference in expected winning percentages. The bottom line is that none of the trade options identified above would on average be expected to improve the Twins by even one total win. Sure, a guy could catch fire and run off 10 straight wins, but it's not likely for any of these guys. Also note that I didn't even account for the reduction most pitchers see in performance when going from the NL to the AL, a much stronger hitters league. So I would consider the +0.9 wins as a ceiling for the mean.
When one considers the massive contract for a Roy Oswalt or Dan Haren, or the cost it would take in prospects to trade for many of these guys, we need to consider whether it is worth it for an extra win. Yes, it could be the difference between making it to the playoffs in a tight three team race, but with three starters sucking right now, would it get us there? And it's also a decent chance that Slowey picks it up at some point and improves down the stretch. If we were talking about replacing Nick Blackburn with a trade target, we'd likely see a greater improvement. But we're replacing a good starter who has been struggling of late. I say stick with Slowey and Baker, ride it out and if we manage to make it to the playoffs, see how far Francisco Liriano and Carl Pavano can take us.
It would sure help if Baker and Slowey grew mustaches...if they can even grow facial hair.