This discussion will cover statistics of the amount of times our starting pitchers face the opponent’s batting order during a game.  But first, a background of how starting pitchers have pitched fewer innings in recent history:


As most are aware, the style of pitching that we see today is not how it has been for most pitchers in the history of the game.  The beginnings of the shift to the radar gun era of judging pitchers started about 2003. QuesTec was a company hired by MLB for pitch tracking technology for the purpose of judging home plate umpire’s performance.  Before this happened, most home plate umpires had very large strike zones — particularly on the corners.


Greg Maddux and especially Tom Glavine were famous for using pinpoint control to find just how far off the outside corner they could go and still get a called strike.  Leo Mazzone taught his pitchers that the best pitch in the game of baseball was a fastball low and away. Many pitchers tried to keep their pitches low and hit the inside and outside corner without using the maximum effort to light up the radar gun.  The key was command and movement — not velocity — for most guys. Livan Hernandez infamously exploited Eric Gregg’s strike zone in Game 5 of the 1997 NLCS to get pitches way off the plate called strikes.


The results of QuesTec led to umpires being more aware of calling pitches off the corner balls.  In 2003, both Maddux and Glavine declined tremendously. This was partially because of age, but also it was because the pitches off the edge were no longer being called strikes.  Glavine had to reinvent the way he pitched for the Mets, and instead of living off the outside corner, he used a curveball much more frequently and pitched inside more to become unpredictable.  He was able to become a solid middle of the rotation starter but was never again the ace caliber pitcher he had been in the past. By definition the strike zone was smaller. Umpires were enforcing the rulebook strike zone more frequently.  With pitches off the plate no longer being called strikes, soft tossers who lived off the edge of the plate had to throw the ball over the plate.


The widespread use of analytics started about the same time.  Moneyball: The Art of Winning an Unfair Game by Michael Lewis was published in 2003.  It was about how Oakland A’s general manager Billy Beane and his staff used metrics to sign and acquire players who had skills that were undervalued by the competition.  The A’s had one of the lowest payrolls in the league but yet won over 100 games in 2002.


Other teams who did not previously use metrics other than batting average, ERA, and counting stats saw the success the A’s had and began building their own analytics departments.  They hired many graduates from Ivy League schools and other prestigious colleges.


Pitchers became more aware of pitching “up and down” to change the batter’s eye level.  And if the strike zone was going to be smaller, then pitchers were going to have to throw harder in order get more swings and misses.  Analytics departments realized that strikeouts were advantageous for pitchers because it took fielding out of the equation. At the same time, it was learned that a hitter striking out was not necessarily that much worse than any other out.  There were guys who struck out well over 100 times who were still producing big numbers. Batters stopped cutting down on their swings with 2 strikes without the fear of being stagmatized for striking out a lot. Pitchers who could throw really hard became more emphasized in scouting. 


Charlie Morton has had a late career emergence to stardom because of this change in philosophy.  For most of his career, Morton was an ordinary sinkerball pitcher who threw in the low 90’s and wasn’t much more than a back of the rotation starter.  He realized that when he maxed out he could consistently sit in the upper 90s and had a hammer curveball. In 2016, his average fastball velocity jumped by 2.3 MPH to 94.3 from 92 MPH in 4 late season starts for the Phillies while coming off an injury.  The Astros noticed this and signed him to a 2 year deal. His velocity continued to climb and he helped Houston win the World Series. This Fangraphs article from 2017 tells the story of his rise: https://blogs.fangraphs.com/charlie-mortons-electric-stuff-has-never-been-more-electric/. Another change made to pitching was because of the discovery of the “times through the order” statistic.  It was found that most starting pitchers decrease in effectiveness the more times seen by a hitter during a game.  Usually by the 3rd time through the order, analytics departments discovered that a starting pitcher wasn’t any more effective than a replacement level middle reliever.


In addition, salaries in the game continued to climb.  Teams didn’t want their star pitchers getting hurt and being stuck on the shelf while being paid millions of dollars.  So why force them to pitch through fatigue when they are more likely to get hurt? Guys started being taken out of the game earlier.  Also, since pitchers were now using maximum effort to throw hard on every pitch, they were tending to get tired earlier in the game.


This brings us to our current starting rotation.  With the exception of Mike Soroka, all of our starting pitchers have struggled after the second time through the batting order.  I pulled the following splits from Baseball Reference:

Performance Against: Times Through The Order

Mike Soroka: BA OBP SLG OPS
1st PA in G as SP .216 .251 .257 .509
2nd PA in G as SP .272 .320 .361 .680
3rd PA in G as SP .211 .302 .325 .627
4th+ PA in G as SP .000 .250 .000 .250
Julio Teheran: BA OBP SLG OPS
1st PA in G as SP .197 .291 .319 .610
2nd PA in G as SP .215 .338 .333 .671
3rd PA in G as SP .268 .346 .455 .802
4th+ PA in G as SP 1.000 1.000 2.500 3.500
Dallas Keuchel: BA OBP SLG OPS
1st PA in G as SP .198 .270 .321 .591
2nd PA in G as SP .238 .315 .388 .702
3rd PA in G as SP .358 .411 .761 1.172
4th+ PA in G as SP .500 .500 .500 1.000
Max Fried: BA OBP SLG OPS
1st PA in G as SP .251 .301 .350 .651
2nd PA in G as SP .286 .330 .451 .781
3rd PA in G as SP .299 .353 .486 .839
4th+ PA in G as SP 1.000 1.000 2.500 3.500
Mike Foltynewicz: BA OBP SLG OPS
1st PA in G as SP .222 .280 .515 .796
2nd PA in G as SP .263 .333 .537 .870
3rd PA in G as SP .367 .418 .633 1.051

These stats clearly show that our starters in general struggle after the second time through the batting order.  Soroka has been near unhittable the first time through the batting order and very good afterwards, but is a 21 year old whose workload has to be managed.  Julio Teheran has been very good the first two times through the order and struggles afterwards — which matches the perception of him hitting a wall in the middle innings.  Keuchel has been excellent the first time through the batting order, mediocre the second time through, and the third time through he has been hit like Barry Bonds on steroids.  Fried has only been very good one time through the order — which is concerning for his long term future as a starter. Folty of course has struggled this year especially with the long ball, but his splits also get progressively worse through the game.


So how should Brian Snitker and Alex Anthopoulos use these numbers to their advantage? They can’t go to the bullpen after the 4th or 5th inning every game, can they?  I believe the answer is to either use an opener or to start piggybacking with other starters. Now, I don’t think that these strategies will be used in 2019 given that all of the relievers now have defined roles.  And these are human beings, one can’t just say “Oh, we’re going to make a radical change to how we use our pitchers in the middle of a pennant race” and expect everyone to be okay with it.


However, with the number of starting pitchers the organization has, I think there should be a serious look given in 2020 to piggybacking the starters.    Soroka, Teheran (who increasingly looks like a guy who may be brought back with his team option), Fried, and Folty are all under team control for next season.  And Bryse Wilson, Kyle Wright, and others are knocking on the door and some have had cups of coffee in the major leagues. And since these rookies all have options they could be shuttled to Gwinnett or swapped out if needed. 


With the surplus of starting pitchers, the team could feasibly use a starter for the first 2 times through the order, then go to a long reliever the next 2 times through the order before giving way to the late inning set-up man and closer.  Perhaps use a 4 man starting rotation with a long reliever paired to piggyback with each starter. Then the last 3 or 4 men in the bullpen would be the late inning set-up men and closer. This strategy would keep the amount of poor middle relievers out of the equation, while at the same time taking away the 3rd time through the order penalty.  If one of the long relievers is struggling and a starter is dynamite, then adjustments could be made to leave the starter in longer while limiting the struggling long reliever — or vice versa.


Do I think the Braves would try something like this?  I have my doubts, as they seem slower to make changes than other organizations such as the Rays, and they have a traditional type manager in Brian Snitker.  But I think it would make the pitching staff better. Go Braves!