24 September 2014

Leaving Kevin Gausman Out of the Playoff Rotation Would Be a Mistake

Kavin Gausman (photo via Keith Allison)
With less than a week left in the regular season and the American League East wrapped up, the Orioles find themselves in the enviable position of being able to rest their starters and line up their starting rotation for the playoffs.  Due to the increased number of days off during the postseason, a team will typically only use their best 4 starting pitchers during a playoff run, sending the 5th starter to the bullpen as a long man/extra reliever.  Essentially, the extra days off in the schedule allow the 4 starters to pitch on regular (or even extra) rest, despite one less slot being filled in the rotation. This presents a (good) problem for Baltimore since all of their starting pitchers not named Ubaldo Jimenez have been pitching really well during the second half of the season.  According to most metrics, the starters have been better than the average major league rotation since the All-Star break.

Orioles Starting Rotation Since All-Star Break
Other than the ERA, these numbers are by no means outstanding, but it’s a good sign from a group that had many doubters before and during the season (including me).  So who gets sent to the bullpen?  My initial thought was that the obvious choice would be Miguel Gonzalez, until this tweet scurried across my timeline on Sunday (followed by a very similar tweet on Monday)…

It was surprising to see that Kubatko sounded so sure that Gonzalez was going to be a part of the postseason rotation instead of Kevin Gausman.  I’m not doubting him, he’s way more tuned in to what the team is likely to do than I am, as that’s his job.  However, I was surprised since I think an argument could be made that Gausman has been good enough to possibly start game 1 of the Division Series.  Having said that, let’s take a closer look at why Gausman is a better choice to start than Gonzalez.

First, let’s see how they compare over the course of the entire 2014 season.

Kevin Gausman vs Miguel Gonzalez - 2014 Season
There are a few differences in how the two pitched this year, but overall, they’ve been fairly similar according to ERA and strikeout/walk rates.  Aside from innings pitched (Gausman made one start in May, but only joined the team full-time in June), the two start to diverge with FIP and fWAR.  Gausman’s lead in fWAR is quite sizeable, especially when you consider that he’s pitched almost 44 innings less than Gonzalez.   This can be attributed (partly) to the differences in their BABIP’s and LOB%.  Compared to 2014 starting pitcher league averages of .296 (BABIP) and 72.4% (LOB%), these numbers would indicate that Gonzalez has been the benefactor of good luck, while Gausman has not (although he hasn’t necessarily had bad luck either).

However, Gonzalez has shown an ability over the past 3 years to consistently post low BABIP’s relative to the league average, which helps him outperform his FIP and generally pitch better than expected.  It’s believed that some (although few) pitchers have a skill that allows them to accomplish this feat on a consistent basis.  Matt Cain, who outperformed his FIP for 6 straight years before failing to do so in 2013 is a prime example.  I’m not saying Gonzalez has this skill, but I wouldn’t rule it out either.  Interestingly enough, Gonzalez’s BABIP in 2014 is actually higher than his career level of .264.  The same can’t be said for his LOB% though, as he’s leaving runners stranded at a 5% higher rate than his career average, which I believe is safe to assume is driven by good luck.

Just on the quality of their 2014 seasons, I would still go with Gausman based on his results and peripherals.  However, maybe the Orioles should choose the “hot hand”.  This end point is arbitrary, but let’s see how each of the two have performed over the last month.

Kevin Gausman vs Miguel Gonzalez - Last 30  Days
As you can see, both have been excellent over the last 30 days, with Gausman showing greater improvement in both strikeout and walk rates.  However, what really sticks out is the extreme BABIP and LOB% of Gonzalez.  Even for him, a BABIP of .226 and a strand rate of 94.1% are very extreme, and he can’t be expected to sustain those levels.  In contrast, Gausman has pitched even better over the last 30 days than he has over the course of the season, despite being slightly unluckier.  Additionally, Gausman has shown consistent improvement over the course of the year when it comes to striking batters out and limiting free passes.

Kevin Gausman K% and BB% by Month (2014)
The Orioles would be doing themselves a disservice by putting Kevin Gausman in the bullpen instead of Miguel Gonzalez.  In fact, an argument could probably be made that Kevin Gausman is Baltimore’s best starting pitching option right now, although that is not within the scope of this post.  There’s no doubt that the Orioles can succeed in the postseason without Kevin Gausman in their rotation, but if that’s the case, they’re not giving themselves their best chance at it.


Erik said...

Gausman is a better pitcher. The question is, is Gausman still a 2-pitch pitcher or is he now a 3-pitch pitcher? How is he doing on the third time through the lineup relative to Gonzalez?

If his value is in the first time through, or the first and second time through, then he belongs in the bullpen.

Matt Perez said...

Gausman has thrown 153 innings this year after throwing only 129 last year.

They're putting him in the bullpen (or leaving him off the postseason roster) because it gives them a chance to limit his innings.

Suppose the average starter pitches eighteen innings every fifteen regular season games. Most relievers aren't asked to throw 3 days in a row. Presuming no off days an above-average reliever could throw two innings per three games or about ten innings per fifteen games. Even considering leverage you'd rather have a good pitcher in the rotation then in the bullpen.

But the playoff schedule is such that there aren't three games in a row for the division series.

It seems likely that only the #1 starter will have the potentially to have two starts and the other three starters will get one. That means that average starter will pitch six innings per playoff series while the top-of-the-line reliever could pitch four or five innings.

It's very possible that a top reliever would be more valuable than an average starter depending on the situations he pitched in for the ALDS.

Matt Perez said...

Gausman is considerably better his first time through the lineup rather than the second or third. If he knows he's only going through once then he may be even better.

For example, Britton threw 92 as a starter and 95 as a reliever. Hunter threw 92 as a starter and 96 as a reliever.

Can Gausman throw 95 as a starter and 98 as a reliever?

Matt Kremnitzer said...

I'd be stunned if Gausman's innings total has much to do with him being left out of the playoff rotation. We're talking about the possibility of one start if they don't make it out of the first round.

But Gausman could have an important role if he's needed out of the bullpen if one of the starters gets shelled early in the series.

Unknown said...

It wasn't included in the article, but I looked at the OPS for each pitcher by times through the lineup in 2014.

1st - .585
2nd - .790
3rd - .795

1st - .792
2nd - .696
3rd - .873

If he's in the bullpen, it would depend on how they plan on using him. If he's a high leverage weapon, he could be valuable, but may not pitch much considering the Orioles have at least 3 other high leverage weapons in the bullpen. If he's used as a long-man/mop up, then I don't think he provides much value and he should probably just be in the rotation to begin with.

Matt Perez said...

If you throw Gausman in the sixth, O'Day in the seventh, Miller in the eighth and Britton in the ninth then starters only need to go five innings.

How much more valuable is it to have Gausman pitch one inning in the sixth vs having a starter go 6 instead of 5?

Matt Kremnitzer said...

Interesting question. Perhaps this requires further exploration...

Phil said...

Nate's comment shows that Gonzo in the rotation going up to 5 innings, facing the lineup twice, and Gausman in the bullpen to only see the lineup once is of greater value to the team. Gonzo would be, based on the sample data used by Nate, less effective than Gausman for a one or two inning relief appearance.

Gausman's continued development and confidence in his off-speed pitch will be the key to his future success. Is there data for how 2-pitch guys vs. 3+ pitch guys fare in the 2nd and 3rd times through the lineup? (i.e., pitchers using just two pitches more than 95% of the time vs. guys who mix it up more?)

Unknown said...

Matt, that's an interesting point, but things are rarely that rigid, and I don't think they should be, especially in the playoffs when it's all hands on deck.

Anonymous said...

I'd also be concerned about that big difference in FIP given that the Orioles defense will not be at full strength.

Pat Holden said...

How about keeping Chen out of the rotation if we face Detroit? They crush lefties. And Buck has clearly kept Chen away from them during his career for this reason

Unknown said...

I'm interested to see what they do with Chen vs. Detroit. Everyone seems to be assuming that he will start game 2. I'd rather them push back Chen, or put him in the pen and start Gausman. Detroit has been absolutely destroying LHP this year.

Anonymous said...

To be fair, Detroit has crushed both Gonzalez and Gausman this year:

Gonzalez: 7 ER in 3 1/3 innings.

Gausman: 5 ER in 4 innings

As well as Bud:

Norris: 9 ER in 12 2/3 innings.

I understand this was earlier in the year, but if we do end up with the Tigers, I don't feel great about anyone other than Tillman. Gonzalez did pitch well in the postseason in 2012. And just to make this whole post seem irrelevant:

Ubaldo vs Detroit:

0 ER in 7 innings