clock menu more-arrow no yes

Filed under:

2013 Closers: Pitch Location vs Velocity

New, 2 comments

Building on the strike zone pitch location quality metric introduced last week, we look at how 2013 closers fare in locating their pitches in combination with pitch velocity.

How does all-time great Mariano Rivera rank in this analysis of pitch location and velocity?
How does all-time great Mariano Rivera rank in this analysis of pitch location and velocity?
Mike Stobe

Last week, I introduced a metric that in its current form attempts to quantify how well pitchers locate their pitches within the strike zone. At this point, the measure is completely context neutral, in that it does not consider league, ball park, defense, base state, out state, count or even the eventual result of the pitch in terms of what the batter was able to do with it. It uses league average wOBA against per pitch type per strike zone sub-zone to determine the typical success of each type of pitch in each section of the strike zone.

While I plan to add other components on top of just location to build it into a potential performance analysis tool, as it is it offers some useful information about a pitcher that I do not think we currently have in one number.

The next pitch characteristic that I will endeavor to work into the equation will be velocity. There was a relatively small, yet statistically significant positive correlation (0.22) between the strike zone pitch location number (essentially estimated wOBA against) and velocity.

After looking only at qualified starters last week, before I move on to consider velocity I thought it would be interesting to examine another subgroup of pitchers where velocity tends to be high - closers. Closers tend to be relievers that have experienced some amount of success and in general throw the ball very, very hard. If nothing else, I wanted to see which 2013 closers were the best at locating their pitches within the strike zone in 2012. I also wondered what sort of correlation I would see with velocity in this small group of specialized pitchers.

Here is the table of 2013 closers, with their estimated strike zone wOBA against based on location alone, along with their average fastball velocity as reported by PITCHf/x. The list of closers is taken from this closers depth chart. I replaced "interim" closers Mitchell Boggs (.3177, 95.7) and Bobby Parnell (.3133, 95.8) with expected-closers-when-healthy Jason Motte for the Cardinals and Frank Francisco for the Mets. For a few pitchers who did not throw pitches classified under FF/FA by PITCHf/x, I used the velocity of their fastest pitch type (sinker).

Team Closer xszwOBA vFA
sfn Sergio Romo .2954 87.9
oak Grant Balfour .2986 92.7
nya Mariano Rivera .2991 90.5
col Rafael Betancourt .3017 91.4
was Rafael Soriano .3025 92.3
sdn Huston Street .3043 88.9
min Glen Perkins .3051 95.2
nym Frank Francisco .3055 93.7
tex Joe Nathan .3058 94.0
mil John Axford .3058 96.1
phi Jonathan Papelbon .3074 93.8
tor Casey Janssen .3080 91.7
ari J.J. Putz .3083 92.8
kca Greg Holland .3084 96.1
chn Carlos Marmol .3088 94.0
det Joaquin Benoit .3097 93.6
bos Joel Hanrahan .3115 95.8
hou Jose Veras .3121 93.7
bal Jim Johnson .3133 94.4
ana Ernesto Frieri .3139 94.0
cle Chris Perez .3143 94.0
mia Steve Cishek .3146 92.2
cha Addison Reed .3155 94.5
pit Jason Grilli .3168 93.6
sea Tom Wilhelmsen .3170 96.2
lan Brandon League .3178 95.0
atl Craig Kimbrel .3185 96.8
cin Aroldis Chapman .3186 98.0
sln Jason Motte .3215 97.1
tba Fernando Rodney .3249 96.2

To see graphically how the strike zone pitch location quality metric correlates with velocity, we can run a simple linear regression. While the sample size here is small, the relationship is actually pretty stunning for closers.


We see a much stronger positive correlation here between strike zone location quality and fastball velocity than we did with the set of qualified starters, all the way up to 0.71. With such a small group, just changing out a handful of names with others could like change the correlation quite a bit, but there does certainly appear to be a trend here. I should say that correlation does not necessarily imply causation. So we cannot look at this and say that throwing harder causes a pitcher to have less control over the location of his pitches within the strike zone, for example.

While not all closers are equal, we know that for a pitcher to be tagged the closer on a particular team, the organization must feel that he can perform at a high level relative to the other options that they have. Given that these pitchers tend to all be quite successful, the relationship above should not be surprising, but it does serve as evidence of something we inherently believe to be true - that pitchers who throw gas do not need to be as refined in their pitch location as those who can't just blow it by hitters time after time. Closers who can pitch at an elite level without elite velocity tend to have much better strike zone command than the fireballers.

This helps to give us a quantitative explanation of how guys like Mariano Rivera, Sergio Romo, Huston Street and Rafael Betancourt can cut it as closers without standout velocity. These pitchers know how to locate strikes in areas where hitters do not as much damage as others.

As far as interpreting the graph, pitchers below the line would seem to possess a combination of strike zone pitch location quality and velocity that is more desirable, while those above the line are worse in this regard. We can think of those below the line as having additional velocity for a given amount of strike zone command, or else additional command within the strike zone for a given fastball velocity.

John Axford, Grant Balfour, Glen Perkins, Greg Holland and Joe Nathan are the five 2013 closers who rate best in the combination of these two pitch characteristics. Fernando Rodney, Steve Cishek, Jason Grilli, Brandon League and Huston Street rank the most poorly of all closers.

I should point out that both Fernando Rodney and by a slim margin Craig Kimbrel appear on the wrong side of the line here for 2012, so this combination is certainly not proof of actual performance. As a reminder, the location quality metric is also not adjusted for home ball park or team defense, two aspects that certainly would help Rodney's performance as a closer. In fact, as a group those pitchers below the line averaged a more damaging basic home park factor of 101.7, while those above the line averaged a more favorable 98.8 home park factor.

The measure of location quality is also specifically focusing within the strike zone, meaning pitchers that operate most effectively outside the zone (i.e. high O-Swing%) may not be entirely appreciated using this approach. Location and velocity also make up just two thirds of individual pitch characteristics available through PITCHf/x, with movement also being ignored in this analysis.

On the whole, I believe even at this point that this is an interesting way to look at the closers around the league today. If I had to choose closers without knowing anything else about them I would take guys who possess a good combination of strike zone pitch location quality and high velocity, although would lean toward more velocity if I had to choose. Long term successful closers Mariano Rivera, Joe Nathan and Jonathan Papelbon all ranked as better than average using this methodology, to cherry-pick some familiar names.

Next week I will build velocity into the strike zone pitch quality metric and see where that takes me.

What do you think about this way of assessing the relative merits of closers?

You can follow me on Twitter at @MLBPlayerAnalys. <a href="" class="twitter-follow-button" data-show-count="false">Follow @MLBPlayerAnalys</a>
<script>!function(d,s,id){var js,fjs=d.getElementsByTagName(s)[0];if(!d.getElementById(id)){js=d.createElement(s);;js.src="//";fjs.parentNode.insertBefore(js,fjs);}}(document,"script","twitter-wjs");</script>

Credit and thanks to Baseball Heat Maps for PITCHf/x data upon which this analysis was based and to Fangraphs for other pitching statistics and park factors.