On Tunnels and CSAA

At the end of January, BaseballProspectus introduced several new metrics to dive deeper into pitching analysis. Called Strikes Above Average is one potentially useful metric for judging a pitcher’s command since it assesses the called strikes that a pitcher creates (controlled for the umpire, catcher, and other factors). Pitch tunnels is another potentially useful metric, for it assesses the movement of a pitcher’s arsenal as the batter might see it within their brief decision window (ex., “swing?” or “take?”). Like any statistic, these items cannot necessarily be used on their own to drive sweeping conclusions, but used in concert they present a fantastic suite for expanding discussion of Brooks Baseball PITCHf/x classifications and Deserved Runs Average (DRA).

Background Reading
Pitch Tunnels
Two ways to Tunnel
Modeling Questions

Zach Davies might present the best motivation for using these statistics. The pure 50 OFP, back end starting pitching prospect never had a scouting report that jumped off the map, and anyone looking to hang a relief judgment on the young righty could use Davies’s size or stuff to plead their case. Yet, Davies immediately proved effective for the 2015 Brewers, making adjustments in his stuff and approach during a short call-up, which foreshadowed the righty’s exceptional first full season (3.5 WARP, 3.47 DRA, and a 3.55 strikeout to walk ratio). Looking at Davies, it is clear that the righty will need to succeed by using his own approach, never deviating, and gaining every advantage possible. It so happens that Davies ranked as the very best CSAA percentage pitcher in the 2016 MLB: no one in the majors gained a higher percentage of called strikes for their team than Davies.

Tunnel statistics paint a different picture of Davies. Here, the righty does not have a Tunnel Differential that jumps off the map (that’s “how far apart two pitches are at the Tunnel Point—the point during their flight when the hitter must make a decision about whether to swing or not”), nor does he have a wicked Post Tunnel Break, which “tells us how much each spin-induced movement is generated on each pitch between the tunnel point and home plate.” Yet, despite numbers that don’t jump off the map here, Davies ranks among the top third of the league in Post Tunnel Break, and is better than sixty percent of MLB pitchers in terms of Tunnel Differential. Following the assessment of Kyle Hendricks that helped motivate discussion of these pitches, it is quite clear how Davies excels despite lacking overpowering stuff: elite command, pitches that are relatively difficult to discern from the batter’s perspective, and pitches that have relatively strong break after the batter makes their swing decision. One might use these statistics to suggest that Davies can get ahead in the count, deceive batters, and induce weak contact with strong break.

Entering Spring Training, Zach Davies is the best pitcher on the 40-man roster, or among spring training invites. In order to assess the full staff, I looked at CSAA, Tunnel Differential, and Post Tunnel Break for any pitchers that had more than 5.0 MLB innings pitched on Milwaukee’s roster (these statistics are not yet available for minor leaguers).

Brewers Pitchers CSAA (Percentile) TunnelDiff (Percentile) PostTunnelBreak (Percentile)
RHP Z. Davies (2016) 3.51% (99th) 0.822 (61st) 0.2275 (66th)
RHP W. Peralta (2016) 1.97% (97th) 0.8427 (45th) 0.1450 (18th)
RHP C. Anderson (2016) 1.21% (91st) 0.8038 (74th) 0.2505 (71st)
RHP C. Knebel (2016) 0.75% (82nd) 0.9065 (15th) 0.3437 (95th)
RHP M. Blazek (2016) 0.58% (76th) 0.9103 (13th) 0.2745 (80th)
LHP B. Suter (2016) 0.39% (72nd) 0.8244 (59th) 0.2052 (50th)
LHP T. Milone (2016) 0.38% (72nd) 0.7816 (84th) 0.2537 (72nd)
RHP T. Cravy (2016) 0.32% (69th) 0.9645 (4th) 0.2324 (64th)
RHP T. Jungmann (2016) 0.23% (66th) 1.0216 (2nd) 0.3814 (97th)
RHP R. Webb (2015) 0.06% (62nd) 0.7773 (83rd) 0.1807 (37th)
[RHP R. Scahill (2016)] 0.05% (59th) 0.8366 (50th) 0.1499 (20th)
RHP J. Marinez (2016) 0.00% (53rd) 0.8369 (49th) 0.1382 (15th)
RHP M. Garza (2016) -0.00% (53rd) 0.8604 (34th) 0.2493 (71st)
RHP J. Lopez (2015) -0.23% (27th) 0.8279 (50th) 0.2068 (54th)
[RHP D. Goforth (2016)] -0.80% (21st) 0.8265 (57th) 0.1236 (10th)
RHP J. Barnes (2016) -0.82% (19th) 0.7636 (90th) 0.1663 (28th)
RHP N. Feliz (2016) -1.17% (11th) 0.8018 (75th) 0.1724 (31st)
RHP C. Torres (2016) -1.29% (6th) 0.8812 (24th) 0.1845 (38th)
RHP J. Guerra (2016) -1.55% (3rd) 0.8450 (44th) 0.2065 (51st)
RHP J. Nelson (2016) -2.79% (0th) 0.8264 (57th) 0.1834 (38th)

This table features some fascinating, surprising, and supportive results. If one thought that Jimmy Nelson had command issues (which could previously have been assessed through mechanical tests and Brooks Baseball analysis, among other sources), CSAA amplifies that finding, as Nelson is the opposite of Davies for the Brewers staff. On the other hand, Taylor Jungmann has that big herky jerky delivery, and sure enough, that righty does not fare well in Tunneling his pitches, but he compensates with exceptional Post Tunnel Break. That splitter truly works for Junior Guerra, who also does not have strong command, but Guerra compensates for that lack of command with solid Tunnel Differential and Post Tunnel Break; basically, one might suggest that Guerra’s pitches are slightly difficult to discern, and they have solid late break. Command is also an issue for Jorge Lopez, but the righty showed solid Tunnel Differential and Post Tunnel Break during his brief 2015 stint, which forms a foundation for judging a successful 2017 should the mechanics remain in tact.

Perhaps the best surprises on the list were Wily Peralta and Chase Anderson. Peralta’s late season surge arguably saved his rotation spot, and arguably earned him a place in the 2017 rotation. The righty succeeded because of fantastic command; Peralta was only slightly behind Davies in terms of adding called strikes to the Brewers. Interestingly enough, the righty also does a moderately good job of Tunneling his stuff, which leads one to question whether he can succeed by playing up his fastball and slider combo. Anderson’s results are simply stunning: the metrics love the righty for command, difficult to discern pitches, and late movement. This leads one to gather that other aspects of Anderson’s pitching strategy are hindering the righty. A late season shift in his arsenal helped Anderson reclaim his season, and the suite of excellent Tunnel and Command statistics leads one to be bullish on the righty entering 2017.

Among relievers, Corey Knebel boasts the best command and movement combination, perhaps foreshadowing the youngster’s move into a higher leverage role this season. Tommy Milone and Brent Suter should be considered rotation darkhorses and potential relief surprises in 2017, as their combination of Command and Tunneling makes them southpaw variations of Davies. With maximal pitch arsenal strategy, these two southpaws could surprise observers as foundations of the pitching staff. Of all the non-roster invitees, perhaps Ryan Webb boasts the best profile, as the righty uses a strong combination of command and Tunnel deception out of the pen (only Tommy Milone is better).

These statistics cannot tell the whole story for the Brewers staff, but they can offer hints about where a pitcher has strengths that can drive other areas of their game, or where a pitcher absolutely needs to improve to keep a job. If the Brewers blend considerations of Command, Deception, and Late Break with their performance metrics, they can maximize a relatively low-cost and unassuming pitching staff that could be ready to succeed in the right circumstances.

Related Articles

2 comments on “On Tunnels and CSAA”


Your new stats are fascinating and illuminating and are making my head spin. Like Maddux and Zito, Davies and Guerra achieve success with radically different approaches. Nelson, Peralta, Anderson and Garza are also on opposite ends of the spectrum, but all had very effective stretches and stretches of total ineptitude. Is it possible to separate the 3 categories of stats by good and bad outings four the latter 4?


Curious if these measurements exist: pitches outside the zone called strikes and inside the zone called balls. That would require access on every pitch to the Fox Tracks or whatever; and would allow a further measurement of what their true CSAA would have been.

I want a real strike zone that is fair for all hitters and pitchers. I don’t need the human element behind the plate.

Leave a comment