SABR-Toothed Triber: Josh Tomlin & Rainy-Day Parades

Two days ago Josh Tomlin made his debut for the Tribe against the New York Yankees—the same Yankees who lead the majors in wins (64), runs (545), OBP (.354), and wOBA (.351).  In other words, a 25 year old rookie was going up against the best offense in baseball.

And Tomlin was awesome.  He retired the first nine batters he saw.  He faced the minimum through six innings.  He didn’t walk a single batter against the team with the highest on-base percentage in baseball. The only run he “allowed” scored from first base after Tomlin had exited the game.  The rookie recorded his first win by going 7+ innings and walking off the mound to a standing ovation.

It was a remarkable debut.  Which makes what I’m about to do a real bummer.  Sorry ‘bout that.

Anyway, here’s the line from the other night:

Like I said, impressive stuff.  The best thing I see there is that he didn’t walk anyone.  Walks kill pitchers, and sometimes I don’t think we realize how crucial free passes are to winning and losing ballgames.  This season, when the Indians walk 3 or fewer batters, they’re record is 26-24, but when walking more than three, it’s 16-35.  In their 42 wins, they’ve averaged 3.00 walks per game, but in their 59 losses, they’ve average 4.24.  Free passes correlate fairly well with winning, and Tomlin’s ability to limit walks is a huge plus.

Let’s keep looking, though.  Judging by the dearth of strikeouts, it looks like Tomlin might be a contact pitcher—and there’s nothing wrong with that per se.  However, to survive in the majors, it helps to keep that “contact” primarily in the form of groundballs; otherwise, the flyballs will eventually become home runs at a pretty consistent rate.  Keeping the ball on the ground is how pitchers like Jake Westbrook and Fausto Carmona survive: they both have groundball rates around 55%.  But Tomlin allowed nearly 60% of his batted balls in the air; that rate jumps to two-thirds if you include line drives.  So only one-third of the batted balls he allowed were on the ground.

Which brings me to a pretty fundamental assumption when evaluating pitching performances, and it has to do with BABiP, or “batting average on balls in play.”  Typically, groundballers have slightly higher BABiPs than flyball guys (usually .290-.300 for flyball pitchers and .300-.310 for groundball pitchers).  This makes sense, because flyballs are caught for outs more often than groundballs are.  But the tiny advantage that flyballers have in BABiP is usually lost over the long run, because they give up more home runs (home runs are not counted in BABiP because they are not “balls in play”).

And because flyball pitchers give up so many more home runs than groundballers, you almost never see a successful flyball pitcher who can’t strike people out at a decent clip.  Let’s look at some of the pitchers who are the most flyball-prone in the AL this season:

You’ll notice that each of these pitchers is above average at striking batters out (average is 7.03 K/9).  The worst is Javy Vazquez, and you see that despite an unnaturally low BABiP, he still has the worst ERA of the bunch.  The best strikeout guy here is Jared Weaver, and his ERA is accordingly the lowest on the list.  Basically, a flyball pitcher’s success correlates fairly well with his ability to strike batters out.

And this brings me back to Josh Tomlin.  You’ll notice that in his debut, not only was his BABiP incredibly low (.145), but none of his flyballs left the park (0.00 HR/9).  For a guy with only two strikeouts on the night, we might say he got a little lucky.

Let’s look at his career AAA numbers to see if any patterns emerge:

The nice thing about larger sample sizes is that a player’s profile is more likely to regress to his true ability.  So what do we see here?

First, he’s a strike-thrower.  For a guy who doesn’t strike out many batters—6.53 K/9 in AAA is pretty unremarkable—he still has a healthy 2.44 K/BB ratio, meaning he’s particularly stingy with the free pass.  Remember, that’s more important than people often give credit for: walks become runs (and losses) at an alarming rate.  Furthermore, while he’s certainly not a groundball pitcher (38.5%), Tomlin’s not quite the flyball guy that he appeared to be against the Yankees.  His 36.9% flyball percentage would be the highest on the Indians’ starting staff, but it’s still pretty close to the MLB average.  And accordingly, his 1.02 HR/9 is right in line with the major league average, which makes sense: a pitcher who lets up an average number of flyballs is likely to give up an average number of home runs.

So what do I see in these numbers?  I see an average pitcher.  A Mitch Talbot type.  Not a lot of strikeouts, not a lot of walks, not particularly dominating, but not half-bad either.  A golf announcer would call him “useful” with a British accent.

But Jon, you’re saying, look at that ERA!  2.76 ERAs are NOT average!  They’re great!  That’s more than “useful”!

And this is where I become a real Debbie Downer.  That 2.76 ERA is great.  It’s phenomenal.  Unfortunately, it’s probably not all that sustainable.  Tomlin’s BABiP in AAA was .238, which is particularly low, and it’s not likely to stay there.

Why do I say this?  Well, a couple of reasons.

First, let’s look at Tomlin’s BABiP by level:

I don’t see anything there that suggests he has some ability to keep batted balls from dropping in for hits at better rate than anyone else.  Those minor league BABiPs are typical, and they make this season’s .238 look pretty unsustainable.  And Tuesday night’s .145 BABiP is certainly lucky.

But more importantly (and theoretically), no pitcher can control his BABiP.  Some percentage of batted balls will become hits, and the pitcher can’t control how many.  We see evidence of this randomness every time we watch a game.  For the season, Fausto’s been lucky with batted balls (.269 BABiP in 2010 compared to .300 for his career), but last night, he got hammered.  Regardless, with a big enough sample, these things tend to even out.  Don’t trust me still?  Here are the career BABiPs for some current and former Indian starters:

Scott said yesterday, “The internet always wins.”  Let me add this: regression always wins too.  With a big enough sample, all pitchers will have a BABiP around .300.

And eventually Tomlin will too.  That’s not to say we shouldn’t celebrate his performance from Tuesday—it was a great debut that had me cheering from my couch.  But when we forecast a player like Tomlin, he looks a lot more like Mitch Talbot than Greg Maddux: low strikeouts, low walks, and average everywhere else.

Not that there’s anything wrong with that–like I said, Mitch is useful to have around, and the Shoppach trade looks great in hindsight.  But to me, Tomlin’s not a “front-end of the rotation guy”—he’s just “another guy”.  Kinda like Mitch.

Oh, and one more thing.  There’s no Easter Bunny.  Sorry.