Collette Calls: Analyzing Second-Half Numbers

I would like to preface this entire article by stating that second half splits are not in any way a rock-solid predictive method of forecasting a breakout for a player the following season. Anything can happen over a 10- to 12-week spread of a season. For example, entering the week, Francisco Lindor has a .406 weighted on base average. That makes him one of the 10 most productive offensive players in baseball since the break and makes him more productive than the likes of Jose Bautista, Carlos Gonzalez, Yoenis Cespedes, Miguel Sano and even Mike Trout. Now, if you want to draft Lindor highly next season because of what he's doing in the second half of 2015, be my guest.

What second half performances should lead you to do is to look at the underlying skills behind the numbers to see what has led to the change in production.

• Has the player made a mechanical adjustment?
• Has the player made a mental adjustment?
• Has the batter changed his approach?
• Has the pitcher added a new pitch?
• Is the player playing through an injury?
• Has the batter's role in the lineup changed?
• Has the pitcher's role on the pitching staff changed?

If you view second-half splits under the investigative lens of the questions above, they can become useful. Let's start with a role player who is on fire lately – J.P. Arencibia from Tampa Bay. Entering 2015, Arencibia had a career .207/.255/.403 line with a ton of strikeouts and few walks in more than 1,600 plate appearances. He was called up to Tampa Bay in August when the surprising power source Curt Casali was injured and in 50 plate appearances, Arencibia has hit .354/.360/.729 with five home runs and 15 runs driven in. Arencibia has made a few mechanical tweaks at the plate this season as he has flattened his bat and opened up his stance a bit more, but it is still the same old Arencibia beneath the recent hot streak, if not worse:

I would say the same thing about Brandon Phillips. Last season at age 33, Phillips stole all of two bases in 121 games played. In 2015, with the same team, and the same manager, he has swiped 21 bases and is one of 17 players to steal at least 10 bases in the second half. Players do not get incrementally faster in their waning years, so improvements on the bases mostly come from technique or role. In the case of Phillips, he was moved from the middle of the lineup guy to high in the lineup. He has had a .350 OBP in the second half of the season, giving him a lot of opportunities to run. He's also been much more aggressive in stealing third base this season, attempting one fewer steal of third base in 2015 than he did from 2011 through 2014:

Carpenter told people in spring training he felt he could hit 20 homers in 2015. Teammates scoffed because Carpenter had combined to hit 19 homers from 2013 through 2014, so Adam Wainwright said he'd give Carpenter a golf cart if he hit 20. Carpenter hit his 20th a few weeks back and has a nice TCU-decorated golf cart to show for it. It looked like a safe bet for Wainwright as Carpenter was mostly a slash hitter that focused on contact and didn't get much carry on his flyballs. Carpenter got his golfcart, but he had to rob his contact skills to pay up his end of the bet.

We also have guys who go from stellar in the first half to awful in the second half, such as Brett Gardner. I made a few predictions before the season, one of which was that I strongly felt that Gardner would out-produce Ellsbury in 2015. That has certainly happened as Ellsbury has once again had injury issues, but Gardner's second half has really cut into the awesome numbers he was on pace for. He hit .302/.377/.484 in the first half but has hit .201/.304/.299 since the break. This is the second consecutive season in which Gardner's production has taken a second-half dive, but since baseball is not linear we can't just call that a trend without looking into what Gardner may be doing differently at the plate. Strangely, there's not much different happening for Gardner:

A great example of a player changing career trends is Logan Forsythe from the Rays. Historically, Forsythe's overall numbers have been limited by his inability to do much against right-handed pitching. In nearly 700 plate appearances against righties before 2015, Forsythe had a measly triple slash of .215/.283/.296 with seven home runs. In 2015, over 421 plate appearances, he has already hit seven homers against righties and has bumped up that slash line to .274/.354/.376.

However, the improvement at the plate isn't founded in an improvement in plate discipline:

On the pitching side of the ledger, one of the more surprising numbers is seeing Ian Kennedy in the top 10 for K/9 for starting pitchers at 10.0, ahead of guys such as Noah Syndergaard, Jake Arrieta, Corey Kluber and Gerrit Cole. If you wrote off Kennedy at the break, I can't say I would blame you. He had a 4.91 ERA, a 1.33 WHIP and was giving out home runs like the Pope was handing out blessings this week as he made stops in the Megalopolis. In the second half, Kennedy has been stingier. He has lowered his ERA more than a full run down to 3.81 and he has been tougher to hit because of the increased strikeout rate. While his 21.2 percent strikeout rate in the first half was good, the 26.2 percent in the second half has really been impressive.

Kennedy has taken a more aggressive approach in the second half as he's cut back on the usage of his offpseed stuff in favor of using more cutters. His usage of the cutter increased late in the 2014 season as well. Additionally, the hamstring injury that cost him early and nagged him during his recovery is behind him and the pitcher is doing well as he heads into free agency this offseason. Pitching at Petco did not help Kennedy as much as people had hoped, but the pieces of Kennedy's game are starting to align in the right direction.

A pitcher out of the same mold as Kennedy is the equally surprising Marco Estrada. When Estrada was first traded to Toronto, many of us laughed at the thought of just how many homers the pitcher would allow in Rogers Centre as he has historically struggled to keep the ball in the park. Estrada has had the last laugh on all of us as he was serviceable in the first half and downright impressive since the break. Prior to the break, Estrada had a 3.52 ERA, a 1.18 WHIP and held batters to a .223 average while striking out 19 percent of the batters he faced. Since the break, he has a 2.70 ERA, a 0.98 WHIP and has held batters to a .191 batting average and has won seven of his 13 starts. In 80 second-half innings, Estrada has permitted 53 hits, and it's clear to see what he's doing differently now.

Estrada's fastball is very pedestrian, and he makes it look better by throwing it high in the zone to change eye levels for his changeup. He has historically thrown a curveball as his third pitch, but his usage of that pitch has declined in recent years. This season, he threw the curve even less than he ever has and has broke out a cutter and it has helped him reduce the damage inflicted on him by right-handed batters.

Pitchers with the fastball/changeup approach that Estrada has typically fare better against opposite-handed batters because the changeup fades away from lefties and many pitchers aren't fans of using changeups against same-handed batters because the pitch fades toward the batter and has a better chance of getting a good part of the bat if the pitch misses location. Estrada isn't one of those pitchers as he has thrown his changeup quite a bit to righties, but the extra pitch in the cutter against righties has helped him improve his numbers against righties.

My plan for this offseason is to do some more in-depth breakdowns of players whose 2015 seasons excited you or frustrated you beyond belief. I'm taking requests in the comments section.