The Tigers had a rough pen in 2014 that looked even rougher in their brief October cameo. By park adjusted ERA, they were the fifth worst bullpen in baseball and by park adjusted FIP they were the fourth worst. Their FIP was better in 2013, but their ERA was in the bottom ten. This isn’t a new problem and while they should have a functioning Joakim Soria, and maybe a health Bruce Rondon, the Tigers offseason bullpen additions were pretty underwhelming.
They punctuated that offseason by signing Joba Chamberlain to a $1 million deal that might bump him up to a million and a half. Now, I’ll be the first to say that Chamberlain is no bullpen hero and that the team fell wildly short of my offseason expectations for what they should have done with the pen, but I did find it odd that Joba got such a small deal.
You might remember his poor second half and awful postseason, but his first half was also quite good. He threw 37 2/3 innings with a 2.63 ERA and 2.47 FIP in the first half and 25 1/3 innings in the second half with a 4.97 ERA and 4.20 FIP. On balance, that’s 63 innings of something like 0.7 to 0.9 WAR relief work, which is the kind of thing that should net you a two year deal and several million dollars.
Does Joba’s good first half vanish with his bad second half? If he had a good second half, would that mean good things for next year?
Now obviously there’s some injury history you can’t ignore, but it seems like Joba got the short end of the contract stick due to when he had his bad outings. If he had struggled out of the gate and been great down the stretch, wouldn’t he have been in higher demand? It seems like it, but I’m not so sure that makes much sense.
This got me thinking about halves. Is one more predictive than the other? Does Joba’s good first half vanish with his bad second half? If he had a good second half, would that mean good things for next year?
To investigate this, I looked at relief pitchers who threw at least 20 innings in the first and second halves in 2013 and then checked to see which half’s park adjusted FIP correlated better with 2014 (for pitchers who threw 30+ innings). That’s 94 pitchers. I didn’t expect to see big correlations in either case, but I wanted to see if there was any reason to treat April outings as less valuable than August ones when evaluating who the Tigers just re-signed.
And you know what? There was. The R-squared between first half of 2013 and all of 2014 was 0.014, which is basically as small a relationship as you can have. The R-squared for the second half of 2013 and 2014 was 0.177, which is still quite small, but it’s actually registering and is much larger than the first half number.
Now obviously, be a little cautious of the results because I just used one season, one stat, and picked my innings cut offs pretty haphazardly. Also, a .18 R-squared is very small, so it’s not like a reliever’s second half is very predictive, it’s just more predictive than their first half.
I can’t really offer a good explanation for why the second half is more predictive given what we know about reliever volatility, but it might have something to do with injuries or a survivor bias resulting from the types of pitchers that meet all of the criteria here.
But this gives a little more weight to Joba not getting bigger offers, and it also lets us look at the Tigers relievers for signs of life. Unfortunately, seven Tigers relievers had 10 or more innings last year during the second half and the best FIP of the bunch was 3.42.
So while this exercise might shine some light on why the Tigers got Joba for cheap this week, it’s also not very encouraging about his performance in the year ahead. And even the relievers who were decent for the club last year weren’t terrific in the second half. It’s just one stat and it’s a very small correlation, but it’s another reason the Tigers needed to add more bullpen help than they did.
Neil Weinberg is a Senior Analyst for TigsTown. He is also the Founder of New English D, a contributor to Gammons Daily, the Associate Managing Editor at Beyond the Box Score, and the Site Educator at FanGraphs. Follow and interact with him on Twitter @NeilWeinberg44