clock menu more-arrow no yes mobile

Filed under:

Breaking Down ZiPS Projections for the 2013 Baltimore Orioles Bullpen

The ZiPS projection system looks at all of baseball history to try to project player performance. What does it see for the 2013 Orioles bullpen?

The slightly-crooked-capped Pedro Strop is not projected to repeat last year's results - but then, he wasn't projected to get last year's results, either.
The slightly-crooked-capped Pedro Strop is not projected to repeat last year's results - but then, he wasn't projected to get last year's results, either.

Over the last couple of days, I've taken a look at the ZiPS projections for the likely 2013 Orioles lineup and rotation. The last part of the team that's left to take a look at is the bullpen. This was perhaps the biggest unexpected strength of last year's Orioles team, and all of the key players will be returning. What does ZiPS think about their chances to be one of the best units in the league again?

Remember, ZiPS looks at every player, including some who will come nowhere close to the big leagues this year, projecting their performance based on other players who have had similar careers. Did you know that new side-arming convert lefty Mark Hendrickson's #1 comparison is named Boom-Boom Beck? How do you think someone gets the nickname of Boom-Boom Beck?

You can still find the full set of projections for yourself on Fangraphs. There is a lot of noise in the post, such as the aforementioned Hendrickson, though that's also part of what makes it cool.

ZiPS and Kurt Vonnegut might agree that we are trapped in the amber of this moment, but baseball players might be bigger fans of Rush, because they will choose a path that's clear: they will choose free will. Though that free will could still mean they will suck, or be just OK, instead of great.


Luis Ayala

2012 Actual 75 1.267 22 2.64 160
2012 ZiPS 50.2 1.480 26 4.62 96
2013 ZiPS 62.1 1.316 28 4.04 106

What is it that makes a reliever projected to throw about 50 innings of 4.62 ERA come out and instead throw 75 innings of 2.64 ERA? One part may be Luis Ayala's poor performance with inherited runners. In 2013, he inherited 50 runners, and 22 of them came in to score, meaning they were charged to another pitcher's ERA. That's how you get fewer earned runs allowed than your projection in nearly 50% more innings than projected.

Ayala was precisely the kind of player that advanced statistics would have predicted to regress in 2012. He had a 2.09 ERA with the 2011 Yankees, with a 4.19 FIP. In 2012 that gap was "only" about a run, with a 3.67 FIP. The difference in WHIP from 2012 vs. his 2013 projection is in the walks: 1.68 BB/9 in 2012 with 2.31 projected in 2013.

So the computer probably thinks Ayala will be allowing more of his own runners to score this year, instead of someone else's. That's what I think, too.

Troy Patton

2012 Actual 55.2 1.024 15 2.43 175
2012 ZiPS 64.2 1.438 22 4.73 90
2013 ZiPS 54.2 1.189 22 3.62 118

Because Patton is left-handed and not very well known, you may be tempted to think of him as nothing more than a LOOGY (lefty one-out guy), but he had more innings pitched than games played in each of the last two seasons, so he does far more than come on to get out lefties - and in those seasons, he has done very well.

What has been his secret? Maybe the 4.18 K/BB ratio? Maybe the 50.3% ground ball rate in 2012? Maybe it's that he stranded 84.6% of baserunners - which, with a 1.024 WHIP, he wasn't allowing very many to begin with.

ZiPS actually likes a small increase in his K rate, but also a few more walks and home runs. If he defies the BB and HR predictions, we could see another very strong season from Patton, but even so, ZiPS is predicting a nice enough season, if not as great as we saw last year.

Darren O'Day

2012 Actual 67 0.940 17 2.28 185
2012 ZiPS 47.1 1.183 19 3.61 109
2013 ZiPS 53.1 1.144 21 3.54 121

O'Day and that sub-1 WHIP was acquired by the Orioles for the waiver claim fee. ZiPS already liked him well enough - in fact, his 2012 projection was the lowest ERA it projected for any Orioles pitcher. Batters hit .202/.254/.360 off O'Day, which indicates that when they did hit, it was for a little power, but most of the time they did not hit.

There were 14 pitchers with a sub-1 WHIP in at least 60 innings in all of MLB, and one of them was O'Day. Pitchers don't just come off waivers at age 29 to do that - except for when they do. Again, we have a pitcher where ZiPS likes him to have a decent season - a few more walks and a bit less luck - but any step back from last year's great bullpen is a gap the team would have to make up elsewhere to keep winning games.

ZiPS is based off of baseball history, but there's probably not many players to be found like O'Day in the ancient tomes. With the O's just giving O'Day a possible three-year contract, it's a sign they like his success to continue.

Pedro Strop

2012 Actual 66.1 1.342 18 2.44 173
2012 ZiPS 63 1.540 31 4.43 96
2013 ZiPS 62.1 1.477 28 4.04 106

There were eleven games in which Strop pitched, allowed no runs, and the Orioles won the game by one run. The world where his ERA is two runs higher is probably not one where the O's would have made the playoffs. In fairness, there were a number of one-run wins which were only such because Strop allowed a run or more. He is what he is: a fireballer who is occasionally too in love with the slider, which either gives up big hits or can't find the strike zone. But oh, that 2.44 ERA even with a 5.02 BB/9! Amazing.

Strop allowed only two home runs in all of those innings. When hitters made contact, it was a ground ball 64.3% of the time, which is also ridiculous. If he keeps up the ground ball rate, he will probably do better than ZiPS thinks again; if he ever cuts the walk rate, he'll likely be one of the best relievers in baseball.

Strop's September meltdowns against the Yankees were concerning, but then he was money for the 11th and 12th innings in game 4 of the ALDS. The pressure doesn't get much higher than extra innings of an elimination game on the road, where any runs surrendered mean the season is over.

Jim Johnson

2012 Actual 68.2 1.019 19 2.49 170
2012 ZiPS 64.2 1.283 28 3.90 109
2013 ZiPS 61.1 1.272 26 3.79 113

The difference between a 1.272 WHIP and a 1.019 WHIP is 0.253. Over the 68.2 innings, that is about 17 baserunners. Maybe 14 fewer hits and three fewer walks. Johnson strands about 75% of baserunners, meaning four more would have scored. In those 17 baserunners lay the difference between the major league leader in saves and some guy who happens to pitch the 9th inning for an MLB team.

He does not get a lot of strikeouts. He has not issued a lot of walks. The strength of Johnson is lots of ground balls, which led, generally, to easy outs, to drama-free ninth innings for a fanbase that had recently seen Kevin Gregg.

Why doesn't ZiPS like Johnson to continue this? He has had two great seasons in a row. Maybe it doesn't like the miles on his arm; 159.2 IP over two seasons is a lot for a reliever these days. Actually, it thinks his .251 BABIP must come up - it projects a .281 BABIP. Johnson faced 269 batters, struck out 41, walked 15 and hit three. Another three hit home runs. 207 batters put the ball in play for 52 hits. A .281 BABIP would mean six more hits. ZiPS also thinks he will give up six home runs instead of three.

Is that the difference between a 2.49 ERA and a 3.79 ERA? Nine hits, three of which are home runs? So many of the Orioles had such razor-thin margins for success last year. If you think they're going to do it again, you need to appreciate the unlikelihood of it all.


The rest of the bullpen will be filled out, I suspect, with names like Brian Matusz and Tommy Hunter, whose ZiPS projections think that they will still be starting pitchers. Both of those men seemed to have found new life as relievers, so you may not be crazy to ignore the 5+ ERA projections of doom for them. However, there's nothing useful to add for them to our comparisons here.


All in all, the Orioles bullpen threw 545.1 IP last year, 4th-most in the majors, in part, though not entirely, due to so many extra-inning games. They had a 3.00 ERA as a unit.

The quintet of Johnson, Strop, O'Day, Patton and Ayala is projected to throw 294 innings while giving up 125 earned runs. That is a 3.83 ERA. In 2012, the five combined for 332.2 innings pitched while giving up 91 earned runs. That is a 2.46 ERA.

With a 29-9 record in one-run games and a 16-2 record in extra innings, again, success hangs in the balance. 34 more runs allowed by the five relievers expected to pitch the most innings - in fewer innings - is the difference between unexpected playoffs and near-.500 mediocrity (or worse).

Reliever volatility from year to year is one of those accepted wisdom things in the statistics community. So is regression for teams outperforming their Pythagorean win-loss record (82-80 for the 2012 O's). So is regression for teams with marked success (or failure) in one-run games, and extra-inning games.

The trends are not in the Orioles' favor. This was also the case last year and they went ahead and made the playoffs anyway. They could find a way to do all of that again, or find other historic avenues to success and the postseason, or they could succumb to the cold, hard truths that lie within one standard deviation of the mean.

The critics are perched, waiting to have their assumptions that last year was a fluke confirmed. They are the birds and the 2013 Orioles are Hunter's hat from the 18-inning game in Seattle. They dropped their milky waste all over the team last season and they will soon be circling above, waiting for each opportune moment where they might crap some more.

Hunter picked up the win that night, pitching scoreless for the 16th and 17th innings while that glop of excrement sat on his hat. Everyone said it was good luck.