It is fairly intuitive that a pitcher will suffer worse results each successive time through the batting order, but nonetheless, TTO (Through The Order) statistics can affirm that very assumption while providing a fascinating look at how a pitcher performs when facing the same batters multiple times in a game.
As such, I have compiled a series of tables displaying the results of the Rays’ starting rotation according to each time through the batting order. These TTO statistics won’t necessarily be able to describe all the various facets (e.g. fatigue) contributing to the decline as a starter works deeper into a game, but they do underline our assumptions that the more a batter see a particular pitcher in the game, the more likely he is to produce. They should also help pinpoint things like whether a pitcher has been more suited to be a reliever than a starter or if there is consistent point during a game when a starter should have been removed before falling apart.
Before we dive into the numbers, I’d encourage you to check out the statistical analysis on TTOP by Michael Lichtman in his article entitled “Everything You Always Wanted to Know About the Times Through the Order Penalty” that was featured on Baseball Prospectus a couple of years ago, and his follow up piece examining the correlation between the TTOP and a players pitching repertoire. Additionally, I should note that the primary rationale for leaving off the 4th TTO was the extremely small sample an individual pitcher would have.
While Chris Archer’s dismal first half of the season last year is no secret to baseball fans, it is intriguing to discover that his difficulties were largely the result of an uncharacteristic struggle the first time he faced a lineup. His .322 wOBA is a far cry from the lower .240’s he’s typically allowed the 1st TTO and also runs against the grain of the historical pitchers’ advantage early in the game.
One culprit for the poor performance is his observed spike in BABIP during that time, which jumped to nearly .323 after sitting around .263 from 2014 to 2015; unfortunately, there was more “bad luck” contributing than just that.
Archer’s strand rate sat about 6% lower than his norms, and despite the lowest “Hard” batted ball percent of his career, he just couldn’t seem to get guys out in their first trip to the plate. If we strip out some of the bad luck, it results in a 3.74 FIP that is less disheartening than the 4.33 ERA he actually put up.
Archer’s home run problem was an issue both the first and third time as the first stroll through the order witnessed his HR/FB increase 5% higher than the previous year, but that pales in comparison to the HR/FB soaring over 20% the third he faced the lineup. On a more positive note, both his K rate and his BB rate the 3 TTO were the epitome of mature pitching late into the game.
In short, there’s nothing in his TTO splits that are particularly alarming; it was a fluky season from which the Rays’ ace is a strong bet to rebound after a dominant second half where he delivered a 3.25 ERA that was on pace with his 3.23 ERA in 2015.
Jake Odorizzi’s initial run through the lineup was quite disappointing given the success he’s had the past couple of seasons, not much unlike Archer above; yet, it’s encouraging that his wOBA numbers were stellar the two subsequent trips through the lineup.
It shouldn’t be particularly surprising to discover that an elevated BABIP was a factor in the early inning decline, but what is disconcerting is that his hard contact percent jumped to 33% the 1st TTO because players were able to barrel the ball more often as they produced a .453 slugging percent. It could simply be an issue of predictability and sequence or trouble with location that’s mitigated as the game went on.
As we turn our attention toward the other TTO’s, a central affirmations of the second aforementioned Lichtman article was a correlation between the number of pitchers in a players arsenal and his success against a lineup the third time he faces the them. This isn’t too say that it’s always advantageous for a pitcher to add to his repertoire, but there’s a reason two pitch pitchers are typically relievers rather than starters.
In the case of Odorizzi, he utilized a whopping 7 different pitches (albeit the changeup was only thrown 13 times, according to Brooks Baseball) that contributed to his ability to post an ERA in the top third of all starters for the 3 TTO. A pitch or two of his were consistently crushed, which is another story, and it’ll be interesting to see if the 26 year old can refine his arsenal in hopes of replicating this past season’s success to either give the Rays another potent option in the front half of the rotation or a more valuable trade chip.
It’s always worth seeing how rookies handle a lineup a second and third time, and although initial struggles aren’t necessarily telling (as players are adjusting to the best crop of hitters they’ve ever faced), you do hope to see some improvement the following year as they learn how to acclimate to facing major league hitters a second and third time in the game.
In some ways, Snell was the stereotypical rookie, and in others, he was a step ahead of the game. His ERA and FIP were both elite the first run through the lineup; actually, to be more precise, the FIP ranked in the top 5 of all starting pitchers, and if not for a .368 BABIP and a low strand-rate, he could have even had an even more dominant ERA to accompany it. Perhaps the most incredible of aspects was a 2.5% HR/FB rate that trailed only Washington’s Joe Ross. The number is likely unsustainable on that level, but considering it only jumped to 6 and 12 the two subsequent trips through the order, there’s optimism it won’t get out of hand this year.
A few more encouraging signs were that his hard contact percent decreased by 3-4 percent each time through the order and his K rate increased the second TTO without a precipitous drop in the following turn.
Snell certainly needs to bring that walk rate down and improve as he pitches later in the game as the 5.74 FIP the 3rd TTO suggests, but there’s nothing so egregious in these statistics to do anything but affirm a desire to lock him an affordable extension now rather than later. It may seem risky without seeing how he adjusts; yet, the killer numbers the first time through bode well for a floor as a strong reliever (even if there’s no reason to think he can’t realize some of the vast potential he presents as a starter).
A quick glance at Cobb’s 2013-2014 statistics is rather depressing because they put into perspective how terrific an arm he was and could be if not derailed by injuries. He absolutely thrived the second trip through the order as his strikeout and walk rates both improved along with slight overall improvement in terms of wOBA and FIP compared to the first rather than being effected by the TTOP. While he’s not alone in that feat, it’s a trend that not many starters are capable of replicating.
As far as 2016 is concerned, 5 starts is too small of a sample size to give us much useful information, so I’d take any of those results with a grain of salt. He was very hittable the 1st TTO with a .452 batting average and high exit velocities, but the continued heavy groundball percentages and average pitch velocity are more reassuring that Cobb could be an effective starter if only his body can hold up to the rigors of a major league season. It might be more telling to examine his TTO statistics as the season progresses to see how his ability and endurance compare to his past results.
After the departure of Drew Smyly, the anticipated battle for the fifth spot in the rotation will probably come down to Matt Andriese or Erasmo Ramirez (unless Jose De Leon is promoted out of Spring), so these last two might be the most significant thing we look at today.
Unlike his competition, Andriese has only started games in the past two seasons, 8 in 2015 and 19 last season, which means he’s only started about one season’s worth of games. Nonetheless, exposure in two different seasons render a few TTO issues alarming.
First of all, Andriese has typically fallen apart by the second or third time he’s faced off against a lineup. His FIP takes nearly a 2 run jump to put him squarely in the company of the league’s worst versus the batting order a second time. To heap even more discouragement on that, his ERA going into the 3rd TTO jumped to an ungodly 7.50 despite better peripherals. Andriese’s batted ball profile reveals that hitters pound him much harder as the game progresses with a massive spike in both FB% and HR/FB %, which is certainly not a recipe for success.
He’s still a useful piece even if it’s increasingly less likely he’ll stick as a starter; the strong results the 1st TTO are consistent with arms that should be moved to bullpen because they’re much more effective in a one time through the rotation role. If you’d like further affirmation, check out his starter/reliever splits:
While Andriese could very well still become a quality starter at the back end of a rotation, he certainly needs to improve his performance on subsequent trips through a lineup if he wants to be more deserving of a starting gig than Erasmo Ramirez.
If you’ve followed Marc Topkin in recent weeks, he keeps suggesting Ramirez as a trade candidate. It could be the product of information that the Rays’ beat writer is privy to or he could just happen to think he’s the logical choice, but either way, Ramirez’ TTO statistics do nothing to dispel that notion.
Let me first note that the number of games he started from 2012-2016 are the following: 8, 13, 14, 27, and 1. Prior to 2015, Ramirez hadn’t exactly been the paragon of excellence when it came to starting games. It wasn’t so much that TTOP was bringing him down as it was he started off poorly and continued to pitch poorly before having some modicum of mediocrity towards the end of the game. The last part of that doesn’t sound so terrible, but the catch is that he only pitched well enough to face a lineup a third time in a limited number of instances, so it’s not as if that 3.79 FIP in 2014 was particularly useful then or reliable for these purposes.
To most people’s surprise, Ramirez put together a quality season in 2015 with a 3.75 ERA and 2.1 fWAR; yet, the team elected to not utilize him in that role this past season with the bevy of other options the team had. He indeed posted impressive TTO statistics in 2015, but their unsustainability on that probably played a role in the team’s decision to bring him out of the bullpen. For one, his BABIP the second time through the order was a good 70 to 80 points below the previous couple of years suggesting regression was soon to follow rather being 43 points better in wOBA than the first time through the order. Ramirez also had a remarkably stark Slugging Percentage the 1st TTO driven by a flyball percent well below his typical numbers.
Ramirez is also well suited to relieving, having showcased an ability to pitch any day of the week, but that’s not what makes his numbers great. He does boast a 6 pitch repertoire so it’s possible that 2015 was more than a mirage when it came to TTOP, but the 1 TTO are not as encouraging as Andriese’s when it comes time to shift to the bullpen as De Leon’s is ready to take the reins of a starting job.
In addition, Ramirez’s $3 million salary and three more years of control make it a logical time to move a 26 year old Ramirez while his 2015 starting numbers and TTO statistics have some shine on them. I wouldn’t expect much, but he could be a useful piece at the back of the rotation for a pitcher needy team like the Angels.
You might wonder why I excluded Jose De Leon from the list, but the reality is he only pitched 17 innings in the majors last season, and two of his four starts saw him exit before reaching the fourth inning. And having already taken a look, let me assure you that his TTO stats are every bit as dreadful as a cup of coffee in the big leagues would suggest.
Anyways, this look hasn’t really been intended to try and predict how much of TTOP (Through The Order Penalty) a player will face moving forward (because that’s a complicated and debatable issue) but rather to provide an examination of a pitcher’s ability to be successful the more he faces the same batters in a game. It’s mostly been an encouraging opportunity to pinpoint some unfortunate bad luck for the front end of the rotation at the beginning of games and to see just how auspicious Snell’s debut season in the majors was.
Additionally, while I’m not sure it has really displayed strong preferences for the fifth starter debate, it’s obvious neither Ramirez or Andriese should hold the position down for long without some kind of development to their approach.
At a minimum, Andriese seems to have fine a reliever career ahead of him.