Would It Be Smart To Sign Kyle Lohse?

The offseason is starting to wind up as most of the big name free agents have signed with their new teams.  Zack Greinke is a Dodger.  Josh Hamilton is an Angel.  Michael Bourn is in Cleveland.  B.J. Upton is in Atlanta.

One name that is notably not on a team at this point is Scott Boras client Kyle Lohse.  The former St. Louis Cardinal pitcher is coming off a spectacular year.  As a matter of fact, he finished 7th place in the National League Cy Young Award vote.  His statistics were astonishing:

ERA: 2.86
WHIP: 1.090
Strikeouts: 143
Walks: 38

Those numbers indicate that he was an elite pitcher last year.  An ERA of below 3.00 is spectacular.  A WHIP of below 1.100 is spectacular.  A strikeout - walk ratio of about 3:1 is pretty good (although I wouldn't call it spectacular).  

He wasn't much different in 2011:

ERA: 3.39
WHIP: 1.168
Strikeouts: 111
Walks: 42

The numbers weren't quite as good in 2011 as they were in 2012, but they were solid.  A 3.39 ERA is good for a top two spot in a rotation for most National League teams.  His WHIP was very good and his strikeout-walk ratio was still pretty good.

Looking at his statistics over the last two years, he is probably the ace of a rotation.  Prior to 2011, however, his statistics were considerably worse.  Here are his stats from ages 22-31:

ERA: 4.79
WHIP: 1.431
Strikeouts: 984
Walks: 485

His ERA was the level of a #4 or 5 pitcher on most Major League teams, his WHIP was average to below average, and his strikeout-walk ratio was barely 2-1, which is barely passable.

So what happened?  Did Kyle Lohse figure something out or did he get extremely lucky in 2011 and 2012?

The first thing I will do is look at his expected equivalent runs against throughout his career versus his actual runs against (basically ERA without errors taken into consideration).  Here is a chart showing Kyle Lohse's actual runs against vs. the expected equivalent runs against:

 From this chart, it certainly looks like Kyle Lohse underperformed his peripherals from age 22 through age 31.  In his most extreme year, his age 31 year, one could expect his runs against to be about 5.00 but his actual runs against figure was about 7.40.  However, in the last two years, he has outperformed his peripherals by about 0.5 runs each year.  How can that be explained?

One possible explanation is that Kyle Lohse had a bad defense behind him.  ERA is not a tell-all statistic.  As I explained in a prior post, Yuniesky Betancourt had as many errors with the 2011 Brewers as Ozzie Smith had with the 1983 St. Louis Cardinals.  However, Smith got to 1.39 more balls per game than Betancourt.  If all 1.39 balls per game were an out for the 1983 Cardinals and a single for the 2011 Brewers, that would equate to more than one more run against per game.

Unfortunately, it is difficult to judge a team's defense in a given season.  The 1.39 outs that Ozzie Smith got for the 1983 Cardinals were made up for on the 2011 Brewers.  The 1.39 outs need to be made up somehow.  Perhaps the next at-bat was a groundout to second base for the 2011 Brewers.  This would have a -1.0 effect on Betancourt's range factor, but it would also have a +1.0 effect on Rickie Weeks' (or whoever was playing 2nd base) range factor.  The net range factor for the team's defense would be 0.

I will use a statistic used by www.BaseballReference.com to look at the difference in defense from the 2010 Cardinals (Lohse's team when he underperformed his peripherals badly) to the 2012 Cardinals (Lohse's team when he overperformed his peripherals.  The statistic is called Runs Above Average per 1,200 Innings:

2010 St. Louis Cardinals

Yadier Molina     +17
Albert Pujols      -3
Skip Schumaker     -5
Brendan Ryan      +14
David Freese      +10
Matt Holliday       0
Colby Rasmus      -12
Ryan Ludwick      +16

Felipe Lopez       -7
Jon Jay            -7
Randy Winn        -24
Aaron Miles        -3
Nick Stavinoha    -25

2012 St. Louis Cardinals

Yadier Molina     +15
Allen Craig        -6
Daniel Descalso    -4
Rafael Furcal      -9
David Freese       +9
Matt Holliday     -13
Jon Jay           +10
Carlos Beltran     -3

Matt Carpenter     +1
Skip Schumaker     -6
Tyler Greene      -21
Shane Robinson     -7
Tony Cruz         +14

It certainly doesn't look like the 2012 Cardinals were significantly better at defense, if at all.  While the 2010 Cardinals had a few players who were terrible defenders (Randy Winn and Nick Stavinoha), they were backups who didn't play all that much.  In the meantime, they had Brendan Ryan (+14) and Ryan Ludwick (+16), players who were replaced by Rafael Furcal (-9) and Carlos Beltran (-3), respectively.

So it doesn't look like defense really played a major role in Kyle Lohse's improved statistics.  The other two possible reasons Lohse's actual stats changed so much while his peripherals "only" changed about 1.25 runs per game are hard to separate from each other.  Perhaps Kyle Lohse got very lucky.  Or perhaps there is a such thing as a good contact pitcher (outside of the whole groundball / fly ball dynamic).  There is actually a great stat to determine if one of these two things (or both) are happening.  It is called BABIP (batting average of balls in play).  It's bascially batting average with walks and home runs out of there.  If the ball is in play and the defense is involved in a play, it's included.

Kyle Lohse's BABIP

2010: .369
2012: .267

I think we may have found something here.  The league's average BABIP over Kyle Lohse's years as a player is about .297.  An extremely high BABIP can be explained by a few things.  The first is defense, but as we see above, that doesn't seem all that likely.  The second is luck.  A weak bloop fly ball to shallow left-center field that falls for a hit is not the pitcher's fault.  A bunt single is not the pitcher's fault.  A pitcher didn't "win" if a hard hit line drive is hit directly to the right fielder; he got lucky the defense was positioned well. Generally a high BABIP implies that the pitcher is likely better than his stats.

However, could it be possible that a player controls how a ball is hit?  Can a pitcher be successful throwing fly balls by keeping all the fly balls on the infield?  Can a pitcher be unsuccessful throwing ground balls if they are all scorching ground balls?  I'm sure the answer is yes.  Still, I'm hesitant to say that this is what is happening with Kyle Lohse.  My reasoning for this is that a pitcher who can throw a pitch with good upward motion, the pitcher will likely be successful throwing fly balls.  It would be something that wouldn't change wildly lover a career.  A change in BABIP of over .102 in two years is ridiculously high.  I tend to think the vast majority of his success in lowering BABIP over time is due to luck.

So, just looking at this, I'm saying signing Kyle Lohse is a bad idea; his success over the past two years is mainly due to luck and his next contract will likely be signed based more on his last two years than his prior 9.  Still, though, it may be prudent to look at his career stats to date and see how a similar pitcher can be expected to progress in the following few years.

In my most recent post, I showed how different types of pitchers (strikeout pitchers (a.k.a. power pitchers), anti-walk pitchers (a.k.a. control pitchers), and  ground ball pitchers (a.k.a. contact pitchers)) progress over a career.  I weighted the group information I had on the prior post into an amount that matches Kyle Lohse's career stats:

Strikeout Percentage: 14.60%
Walk Percentage: 6.60%
Ground Ball / Fly ball Percentage: 0.72
Career Exp. Eq. Runs Against: 4.41

On the graph below, the thick blue line is Kyle Lohse's actual career progression.  If we want to use his career Exp. Eq. RA statistic to project the rest of his career, follow the red line.  If we want to use his strikeout percentage to project the rest of his career, follow the pink line.  If we want to use walk percentage to project the rest of his career, follow the orange line.  If we want to use groundball / fly ball percentage to project the rest of his career, follow the purple line:

 First, the good news for Kyle Lohse.  The individual statistic that seemed most consistent when determining a career projection was walk percentage.  That is consistently lower than any other statistic until age 39.  He maintains a 4.00 Exp. Eq. RA with that statistic until age 36 when there is an increase to the the 4.20 range.  That is far better than his actual stats for most of his pre-2011 career, but it's still not great (think #3 type pitcher, #2 on a team that needs more pitching).

The bad news: none of these stats show an Expected Equivalent Runs Against statistic of better than 4.00 over the remainder of his career.  His K%, GB/FB%, and overall Exp. Eq. RA statistic show a runs against statistic of closer to 4.30 in the next few years; probably not much better than a #3 or #4 for most good teams.

The big elephant in the room that also works against Lohse is that draft pick compensation is attached to him.  If a team signs him at this point, that team loses their first round draft pick to the Cardinals.  That is going to deter many teams from signing him, and the teams interested in him will take this into consideration when signing him.  But is it worth giving up a 1st round draft pick and a potential ace in the future for a #3/#4 pitcher right now, even if you can get him at a discounted rate?

The question that this post looks to answer is "Would it be smart to sign Kyle Lohse?"  Here is the information I came up with in this post:

• Kyle Lohse has underperformed his peripherals for most of his career.. but started to overperform them in 2011.
• There are three reasons this could be (defense, luck, actual pitcher's skill at contact pitching)
• Defense doesn't seem like a strong possibility; using a range-based defense stat, the 2010 Cardinals were at least as good as the 2012 Cardinals defensively and were likely better.
• Actual pitcher's skill at contact pitching may play an effect.  However, a wildly volatile BABIP from 2010 to 2012 would indicate that Lohse consistently gets batters out using more unconvential means (i.e. getting a ton of players out with the pop fly, thus maintaining a low GB/FB ratio but making up for it by getting a higher percentage of flyballs out)
• That leaves luck, and luck runs out.
• Lohse's walk percentage gives a rosier outlook than the rest of his peripheral stats.  It tells us to expect a 4.00 exp. eq. RA from Lohse over the next three years.
• Using strikeout percentage, groundball/flyball ratio, and overall exp. eq. RA, Lohse can expect about a 4.30 exp. eq. RA over the next three years.
• Lohse is also attached to draft pick compensation.

Based on all this information, I come to the conclusion that it would not be smart to sign Kyle Lohse at this point.  A team that already gave up its first round draft pick (i.e., the Angels who signed Josh Hamilton or a team that has its first round draft pick protected by being in the top ten in the draft) may make sense.  The Cardinals who would not lose a draft pick might make sense.  And it also might make sense for a  team to sign him after the draft in June so there is no draft pick compensation.  But a team that would lose a draft pick for Kyle Lohse would be absolutely foolish to sign him.