Do St. Louis Cardinals closer Trevor Rosenthal's 2014 stats support him throwing exclusively out of the stretch?

St. Louis Cardinals closer Trevor Rosenthal had a rocky 2014 even though he notched 45 saves in 51 opportunities for a converted-save rate of 88%.The primary issue was walks. Rosenthal more than doubled his BB% from 2013 to 2014. Back in October we took a gander at what was wrong with Rosenthal, and the numbers showed that he was throwing fewer first-pitch strikes, fewer strikes overall, and opposing batters were swinging less often against the righty. Given my belief in the importance of the first-pitch strike for a pitcher, my recommendation of getting on top more often probably wasn't very surprising.

What this analysis left out was the fact that opposing batsmen fared well against Rosenthal's first-pitch offerings when they swung at them. In 29 PAs, a batter swung at and made contact with Rosenthal's first pitch. They posted a .519 BA, .500 OBP, .778 SLG, and 1.278 OPS. If you were giving up contact that resulted in that stat line, you might become a little gun shy about cutting the plate with your first offering.

Derrick Goold interviewed Rosenthal down in Jupiter, which served as the foundation for an interesting article over at stltoday.com. I was pleased to read Goold's reporting that Rosenthal wants to return his focus on throwing first-pitch strikes in 2015 to the same level it was while he was coming up through the farm system:

To help that first-pitch focus, he’s going retro. Rosenthal also said in the minors he was conditioned by the Cardinals’ development work to count his first-pitch strikes.

"Charted it every game," he said. "I got in the habit of mentally keeping track of it. I think it would be valuable to get back to that as far as being mentally sharp from the first pitch. It’s not as easy as (throwing more strikes) because if I’m going to start throwing a fastball over the plate every time that first batter is going to hit a double and he’ll be on second, not just first. It’s more of the quality of the first pitch. It’s about preparing for it."

A pitcher keeping track of his first-pitch strike total during starts? That warms my heart. And I suspect Joe was pleased to read Rosenthal discussing the need to diversify his pitch selection. Back in the dead of winter, Joe researched Rosenthal's pitch usage and concluded:

While Rosenthal's fastball is one of the most electric pitches in baseball, it might be beneficial for him to increase his changeup use in the future, especially on the first pitch. Also, his changeup is so good that he really shouldn't shy away from throwing two of them in a row, either. The mere threat of a ~100 MPH fastball does a pretty good job at keeping hitters from sitting on back-to-back changeups.

During episode 16 of the VEB podcast, RB and I discussed Rosenthal's struggles a bit, due to an email one of my cousins sent, expressing his concern about the St. Louis closer. As we talked about the flamethrower's struggles last season, RB made the excellent point that Rosenthal had particular trouble against the first batter he faced when entering a game. I made a mental note at the time to look into that further. RB's observation struck me as true, but I hadn't sought out the numbers reflecting that particular split. Then I forgot to look up Rosie's splits. Well, Goold's article rekindled the desire—in particular, because of Randy Choate's idea to help Rosenthal be more focused for the leadoff man of his outing.

The following chart shows how opposing batters performed in 2014 against Rosenthal when they were the first batter he faced in an appearance or the leadoff batter of an inning, compared to the overall batting line against the righty. It also includes tOPS+ which compares an individual split's OPS to that against Rosenthal overall. The sOPS+ column shows how the performance against Rosenthal compared to MLB overall in that specific split.

Not very good in the BA department, but the walk rate was actually below his overall BB% for the season.

Goold reports further:

Rosenthal intends to experiment this spring with a few different ways to focus on the first batter and controlling counts.

One change will be obvious with the first batter he faces.

During a session against teammates Monday, Rosenthal for the first time threw every pitch out of the stretch. He, some teammates and the coaches have discussed abandoning the windup this year. He can maintain his power and control, and out of the stretch was when he pitched most effectively a year ago.

The Cardinals tried this experiment last year with Justin Masterson. You'll recall that St. Louis traded for the sinkerballer last year while he was on the disabled list, convinced that mechanical adjustments would return him to the form he showed the year before. It didn't work out that way. Masterson was such a mess that, per the reporting of Goold, the coaching staff wound up having him attempt to pitch solely out of the stretch in order to simplify his mechanics and make them more repeatable. Goold quoted Matheny as follows regarding the impetus for the Masterson stretch-only experiment:

"We know it’s there, it’s just repeating, repeating, repeating," Matheny said. "What is getting in the way of not repeating? It’s the same thing when I work with catchers. What are the little things that I can chip away that are extra movement. Just chipping. It’s chipping away the little things that are causing the inconsistency."

Rosenthal has had problems repeating his delivery, so simplifying his mechanics might not be the worst idea in the world. Nonetheless, this proposed experiment made me wonder if the numbers showed that Rosenthal was in fact a better pitcher out of the stretch (i.e., with men on base) than out of the windup (i.e., with the bases empty). So, as I am wont to do, I made a chart for comparative purposes with the same stats as the one above.

For starters, this is putting an awful lot of stock in 300 PAs, which is something Rosenthal recognizes. Goold quotes him in the article using the phrase "small sample size." Using that phrase and a focus on first-pitch strikes? Rosenthal is vaulting up my list of favorite Cardinals (which includes every Cardinal).

What one makes of this chart really depends on how one defines a pitcher's effectiveness. Do you believe in assessing a pitcher independent of his fielders, using only strikeouts, walks, hit batsmen, and homers? Or do you use stats that include batter results such as BA, OBP, SLG, ISO, and OPS? By those metrics, Rosenthal pitched better over 165 PAs with ducks on the pond than the 143 PAs he threw with the bases empty.

However, the primary issue with Rosenthal in 2014 was walks and he issued more free passes pitching out of the stretch with men on base than he did throwing out of the windup with the bases empty. But five of those walks with men on were of the intentional variety, which means that Rosenthal's unintentional-walk rate with the bases empty of 12.6% was a bit higher than his unintentional-walk rate of 11.5% with men on. That percentage-point difference is the equivalent of about one walk.

All of this makes me wonder whether throwing exclusively out of the stretch is really the cure for what ailed Rosenthal in 2014 or if it's just unnecessary tinkering based on a not-so-meaningful small sample size.