Connect with us

Pittsburgh Pirates

How Much has luck affected Pittsburgh Pirates pitchers?

Nate Werner

Published

on

After a rough start to the season for the starting pitching, it has been the bullpen that has let the Pittsburgh Pirates down in recent weeks. The question is, have these performances been deserved, or are they just the result of good and bad luck?

The traditional way of determining if a pitcher is either over or underperforming, is to look at their ERA and compare it to their FIP; if the two statistics differ to a large degree we say their due for some “regression”. While there’s nothing wrong with that approach, in fact it can be quite useful, it does leave a lot of ambiguity as to how the pitcher will regress and why they were able to over/underperform in the first place.

What is useful is to use the idea of clustered runs, or that the sequence that a pitcher gives up hits matters. Take for instance two different pitchers who both pitch 6 innings, and give up a walk, a single, and a home run; one pitcher gives them up in that exact sequence, yielding 3 runs, the other pitchers spreads the 3 events out over his 6 innings of work, thus only yielding 1 run off the homer. The first pitcher has an ERA of 4.50 for that game, while the other has a 1.50 ERA. Over the course of an entire season, this clustering is largely random, meaning pitchers will neither cluster all batting events into one inning, like the first pitcher, nor will they perfectly spread them out like the second pitcher.

If a pitcher has innings or stretches with a lot of hits clustered together, while having other stretches where they have few or no base runners, this is more likely the result of bad luck. The difference between good and bad pitchers is that good pitchers just let fewer base runners on overall, but they don’t affect the clustering of those hits very much, as hits are largely random events.

To find clustered runs, or runs that result from too much clustering of hits, we just have to get the estimated number of runs a pitcher would have given up based on the kinds of hits they gave up. This is fairly easy to calculate if we have their wOBA against; we can translate that into wRC, which amounts to the number of runs they would have given up under an average clustering of those hits. Clustered runs are just the difference between a pitcher’s actual runs given up and their wRC against.

Since we’re often talking about pitchers who have different amounts of playing time, it’s much more useful to talk about these things in terms of rate statistics rather than counting statistics. Using a pitcher’s wRC for runs in the numerator of the RA9 formula we get their wRA9, or what we would expect their runs against per 9 innings pitched to be with an average amount of clustering; not too much clustering and not too much diffusion of hits either. Taking RA9-wRA9 gives us what we’ll call Clustered Run Average (CRA), or the average number of runs a pitcher gives up per 9 innings pitched, as a result of clustering hits together.

Starting Pitching

Here’s how the Pirates starters have looked in terms of clustering:

Pitcher RA9 wRA9 CRA CRuns
Nick Kingham 4.34 3.04 1.30 4.20
Jameson Taillon 3.97 3.30 0.67 4.91
Ivan Nova 5.69 5.37 0.32 2.22
Trevor Williams 3.84 3.80 0.04 0.29
Chad Kuhl 4.26 4.99 -0.73 -5.52
Joe Musgrove 2.37 3.39 -1.02 -2.16

The NL average wRA9 so far this season has been 4.19 and the NL starters average RA9 is 4.28.

Nick Kingham has had the worst luck of the starters this season, giving up nearly 1 and 1/3rd runs per 9 IP as a result of clustering, for a total of more than 4 runs in less than 30 IP. Jameson Taillon has also had some tough luck giving up about 2/3rds of a run per 9. What is good to see is that Taillon, who is well below the NL starter’s average RA9, has managed to do so even with a spell of bad luck.

Chad Kuhl, on the flip-side, has prevented more than 5.5 runs this season by diffusing hits more than average. I found this to be a bit surprising from Kuhl; if there was one pitcher on the team I thought was susceptible to “the big inning” where a lot of hits are clustered together, it would have been Kuhl. This suggests he’s actually been quite lucky in terms of limiting the big inning per opportunity to have one. It also suggests that he might be due for a bit of regression by giving up those big innings rather than sneaking out of them as he has done this year.

Musgrove’s numbers should be taken with a grain of salt, as of writing this, he’s only pitched in three games and 19 innings; without taking the time to look at this formally, I doubt any of the above metrics have had time to stabilize. As good as Musgrove may have looked during his first two starts in particular, I doubt he can sustain a 2.37 RA9 for the rest of the season. That being said, if something like his 3.39 wRA9 can hold up as his RA9 for the rest of the season, this starting rotation will really have solidified into a strength of the team, rather than a liability.

Relief Pitching

Now about the bullpen:

Pitcher RA9 wRA9 CRA CRuns
Felipe Vazquez 5.79 4.07 1.71 4.44
Edgar Santana 3.42 2.72 0.70 1.85
Tyler Glasnow 4.36 3.90 0.46 1.70
Michael Feliz 6.08 5.67 0.42 1.10
Kyle Crick 2.75 2.62 0.12 0.27
Steven Brault 2.63 3.14 -0.51 -0.77
Richard Rodriguez 2.45 3.34 -0.89 -2.17

The NL average RA9 for relievers this season is 4.19.

Perhaps the most striking name on this list is Felipe Vazquez, who has given up the second most clustered runs this season, coming in less than half of a run behind Jameson Taillion, who has pitched 43.1 more innings this season than Vazquez. In terms of CRA Vazquez ranks as the worst on the team, giving up close to 2 clustered runs per 9. While this certainly doesn’t explain all of the lack of performance from Vazquez this season; a reduction in speed and a change to his pitch selection strategy have likely impacted the change as well; it is still worth noting that clustering and a healthy dose of bad luck have also played a role.

There is also a seeming inversion in how the bullpen has performed, and how Clint Hurdle has utilized it. Vazquez and Feliz have been the high-leverage 8th-and-9th-inning guys most of the season, all the while performing the worst in terms of RA9 and wRA9. Santana, Crick, Brault, and Rodriguez are below average in both categories, with Glasnow not far behind were it not for some bad luck.

While Crick, the reliever who has pitched the best with the least amount of luck, is now being used for 8th inning duties, the fact still remains that this bullpen has quality relievers; they’re just not being used optimally because of preconceived notions about what their “roles” should be.

The net “luck” of the entire staff has been just slightly bad. While the Pirates rank as having given up the 10th most clustered runs in the majors this season at 0.12 CRA clip, they’re about 0.4 of a standard deviation from the from the league average in terms of both total Clustered Runs and CRA. What this means is that even though there are individual candidates due for some improvement or regression, the total output of this staff isn’t likely to improve all that much on the basis of a change in luck alone.

A word of caution when using CRA and wRA9; they are meant to be descriptive statistics to quantify a particular phenomenon, namely the clustering of hits against a pitcher. While there is predictive power of wRA9 in the 1st half of the season on a pitcher’s 2nd half RA9, it is not as great as the FIP to ERA relationship; the Spearman correlations are 0.40 and 0.46 for the halves of the 2017 season respectively. There is, however, some evidence of a “balancing out” in luck occurs on the basis of clustering. There was a -0.14 Spearman correlation between first half CRA and second half RA9 last season, meaning the greater a player’s first half CRA was, or the worse “luck” they had, the better they did overall in the second half. This is better explanatory power than the first half difference between ERA and FIP (0.02 Spearman correlation).

Nate Werner is a senior at Penn State, where he is studying for his B.S. in Economics. He is a lifelong Pirates fan that uses the tools of statistical analysis to dive deeper into the numbers of baseball. His goal is to take the style of analysis used in front offices across the Major Leagues and bring it to the computer screens of everyday fans. You can read some of Nate’s more general analyses of baseball on goldboxstats.wordpress.com and follow him on Twitter @GoldBoxStats.

Advertisement
scores by the Score
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Pittsburgh Pirates

Parsing GM speak, living with Gregory Polanco and more

Published

on

On the latest episode of Locked on Pirates, host Jason Rollison explains how to parse GM Speak, how to live with Gregory Polanco and wonders out loud if the Washington Nationals just opened the floodgates.

All that and more on the latest episode of Locked on Pirates, live for your listening pleasure:

Pittsburgh Pirates fans – Subscribe, Rate, Review!

Locked on Pirates is the semi-daily Pittsburgh Pirates podcast of choice for discerning Pirates fans.

Jason Rollison from Bucs Dugout and others will help you through the 2018 Pittsburgh Pirates season.

You can help chart the podcast’s success by subscribing to our podcast and writing a review!

We are on every podcast app that ties into iTunes, so search for us! Or, if you’re on an Android or iOS device, please just click the links below to subscribe, and while there, why not leave a review?

Using something other than these three? Just search for “Locked on Pirates” in your podcast app of choice.

Continue Reading

Pittsburgh Pirates

Off Day Mailbag with 93.7 The Fan’s Josh Taylor

Published

on

On this episode of Locked on Pirates, host Jason Rollison brings in Josh Taylor of 93.7 The Fan to help answer your burning Pittsburgh Pirates questions and more.

The latest episode of Locked on Pirates is available for your listening pleasure. Each off day, we open up our mailbag to your various gripes, salient points and even more gripes (understandable) to help us get through the off day.

Josh is always a great guest, and together we talk Jung Ho Kang, Gregory Polanco and so much more. It’s an engaging bout of Pirates talk without the BS. We hope you enjoy, and if you do please consider subscribing at the links below!

NOTE: Had some audio issues a the beginning which level out at around the 1:00 minute mark. 

Pittsburgh Pirates fans – Subscribe, Rate, Review!

Locked on Pirates is the semi-daily Pittsburgh Pirates podcast of choice for discerning Pirates fans.

Jason Rollison from Bucs Dugout and others will help you through the 2018 Pittsburgh Pirates season.

You can help chart the podcast’s success by subscribing to our podcast and writing a review!

We are on every podcast app that ties into iTunes, so search for us! Or, if you’re on an Android or iOS device, please just click the links below to subscribe, and while there, why not leave a review?

Using something other than these three? Just search for “Locked on Pirates” in your podcast app of choice.

Continue Reading

MiLB

The Locked on Interview – Adam McInturff from 2080 Baseball

Published

on

In the return of Locked on Pirates and the Locked on Pirates interview, Alex Stumpf talks the 2018 Pittsburgh Pirates draft with 2080 Baseball’s Adam McInturff. Together the duo go through the 2018 Pirates draft and talk about some of the lesser-known prospects for the West Virginia Power.

Pittsburgh Pirates fans – Subscribe, Rate, Review!

Locked on Pirates is the semi-daily Pittsburgh Pirates podcast of choice for discerning Pirates fans.

Beginning with Opening Day, Jason Rollison from Bucs Dugout and others will help you through the 2018 Pittsburgh Pirates season.

You can help chart the podcast’s success by subscribing to our podcast and writing a review!

We are on every podcast app that ties into iTunes, so search for us! Or, if you’re on an Android or iOS device, please just click the buttons below to subscribe, and while there, why not leave a review?

 

Continue Reading
Advertisement

Trending