It’s the Bullpen, Stupid | The Process Report

It’s the Bullpen, Stupid

Despite their record the Rays have mostly looked like a .500 team this year. That’s not homerism, well maybe a little, but take out the stretch where their outfield consisted of two the dogs from the Cowboy Monkey Rodeo and a wad of gum filled with sunflower seeds and they’ve mostly played to that record for the majority of the year:

win

A brutal stretch in the middle of the season and another at the end dug a hole so deep that the team never had a chance of clawing out. The thing is, this tells us more about what happened than the why. It also looks like the bad stretches do outweigh the good throughout the year even if you remove the worst runs so it would be disingenuous to say that this was an average team that just had a couple of tough stretches due to injury. Let’s see if WAR (via Fangraphs) can help us shed light on the situation:

team-war

I have graphed the WAR contributions from each phase of the game for each team. The positional players (blue) averaged 18.6 WAR per team this year, and you can see that the Rays were basically right on that, if not a touch shy. However, they made up for it with their starting pitching (red) contributions, that were a full win better than the league average of 10.7. Ranking these two segments places the Rays 18th for their positional players and 15th for their starters. That’s pretty dang close to average.

The problem comes in when you search for, and can’t find, the bullpen contributions. The Rays had the 29th-worst corps of relievers in baseball this year with only the historically bad Cincinatti Reds being worse. If the Rays had gotten merely an average pen they might have looked more like the Orioles, Mariners, Marlins or Yankees. Bubble teams that have gone down to the wire, but mostly aren’t going to make the post-season. In the new Wild Card era that’s as good of a definition of average as you’ll find.  They didn’t have an average pen, though. They had a bad one that aggregated a WAR total of -0.2. They were sub-replacement as a group. A big reason for that is all the sub-replacement level pitchers they needed to rely upon.

I want to come back to that, but first let’s look at a reason for hope. Maybe some of you have been fans long enough to remember 2007. It was the year before the YEAR. That team could hit, boy, and the starter pitcher group was finally starting to flesh out some real arms as Scott Kazmir and James Shields began to take steps towards becoming top-of-the-rotation stalwarts. The thing is, they were miserable defensively and even worse in the bullpen. You can maybe get by with one or the other being bad if you’re great everywhere else, but you’ve got no chance if they’re both horrific. The deplorable defense deserves it’s fair share of blame, but let’s focus on the bullpen.

screen-shot-2016-09-30-at-9-21-16-am

Here I’m using Win Probability Added (WPA). It’s like WAR, but unlike the WAR it does not strip out context so that players are punished or rewarded for failing or coming through, respectively, in bigger moments. The kind of moments where bullpens show their mettle. I have ranked each of the Rays teams in franchise history from best to worst by this metric, and it is no surprise that the 2007 Devil Rays were the worst. In fact, there were only 570 teams in this sample so they were the second worst bullpen in the entire league since they became a team!

Here’s the hopey changey part. Look all the way to the very top. Yeah, that’s right. During the YEAR the Rays had the best bullpen in team history. Just one year after boasting one of the alltime terrible collection of relievers the team the team was able to fix the issue to not only have the best performance in team history, but the 17th best of all 570 team-seasons during this timeframe.

A big reason for this is that bullpens are volatile. Great guys get hurt or become unreliable or are mostly fine, but spend most of their season working off the lingering effects of a bad blowup. Sometimes they even stay very good. Let’s look at the key culprits of those wildly divergent seasons:

screen-shot-2016-09-30-at-10-11-57-am

I’m focusing on some of the more more useful statistics for relievers,  and isolating those that saw at least 10 appearances. Additionally, I have highlighted the guys that were with the team for both years. It should be noted that the team acquired both Wheeler and Balfour mid-season in 2007. The first thing that jumps out is how few guys they needed in 2008 compared to the year prior. This is a signal of both healthy, and effectiveness. In 2007 the team tried whatever they could to try to stop the leaking, and ultimately settled on a group including Reyes and Wheeler in the highest of leverage, and trying many options in the medium leverage situations that so often dictate the outcome of the game during those middle innings where the coin is still spinning in mid-air.

Nothing worked. Free agent signings, trade acquisitions, homegrown arms. None of it. Al Reyes was probably the lone bright spot, though Dohbermann should get the nod, as well. Every single other guy failed to perform in a positive manner. All of 15 of them. So what did the team do in 2008. You would be wrong if you said they got rid of everybody and started over.

You can see that six of the ten guys that saw more than ten appearances in 2008 also saw at least decent run in the abomination the season before. The team signed free agents Troy Percival, to be the de facto closer, and Trever Miller, for the less glamorous LOOGY role. A move to the pen for J.P. Howell (the original Erasmo Ramirez) with subsequent increases in trust ultimately leading to heavy usage in medium-to-high leverage situations gave them a true fireman that could go multiple innings. Lastly, the team gave up not much at the trade deadline to bring in Thunder From Down Under Chad Bradford who was mostly excellent down the stretch. Those were the only real additions.

Those guys were mostly ok with Howell being a true force in chilling games until the later innings. The real change was just that the rest of the guys got better. Wheeler went from being ok in high leverage to being very, very good in high leverage with the ability to go out there nearly every other day. Balfour went from being a minus to the second best reliever on the team. Al Reyes went from being the most trusted reliever to getting tased on his birthday in a bar brawl to being released from the team in August for poor performance. Results-wise, Dohmann was pretty similar without all the other off-field stuff.

The team identified the worst performers in the system and made sure to remove them, but also ear-marked some guys that they thought could improve with the right tweaks or usage. They were able to then use these guys in a way that promoted strengths, and hid weaknesses. Improved health and performance were big keys, but so was supplementing the mix with some character guys that could be relied upon. So that brings us back to the current team:

screen-shot-2016-09-30-at-10-13-35-am

There are just so many maybes that need to have decisions made on them, but if we isolate the guys that absolutely deserve to be in next years pen we are left with a list that is rather wanting. Alex Colome is a leadpipe lock. Xavier Cedeño should be, provided his arm functions properly. Matt Andriese will be used in some capacity, but could potentially be in the rotation. Erasmo Ramirez will probably continue to be the plow horse, and I’d use Chase Whitley in a similar capacity. Ideally, you can use all three of Andriese, Erasmo and Whitley for 5-6 outs on separate days so that you almost always have at least two of them available whether you need them or not.

If you can cover three to four innings between two of those guys then you can use Cedeño in more of a LOOGY role. Colome has also shown the ability to go get four to six outs when needed. Giving these guys proper rest will be the key, but I think the team can cover a lot of innings with those five guys. So that leaves room for another lefty arm and ideally a guy that can just flat out get righties even if lefties give him the fits. Perhaps Boxberger functions as that additional lefty, but I think the team would be right to have a short leash on him this year with a real lefty ready and waiting. As for the righty-killer I don’t think we have that internally. Ideally, the team can sign a few guys out of Sergio Romo, Koji Uehara, Jason Grilli, Drew Storen, Joe Blanton, Neftali Feliz, Junichi Tazawa, David Hernandez, Joaquin Benoit, Daniel Hudson, Greg Holland, Pat Neshek and hope to catch lightning in a bottle. Nearly all of those guys have some sort of health or performance-related issue or they wouldn’t be in the Rays scope of vision, but identify the right one and this pen might be closer than people think to putting this team over the top.



One Comment

  1. rb3 wrote:

    Is there a true-southpaw Boxy replacement internally? Seems like even slimmer pickings for that than for the righty-killer role. Marks? Genesis Cabrera? (Just kidding/too young still.)

Leave a Reply

#layout { padding-left:20px; }