Rubbing Mud: Addison Russell Can’t Get No Relief From Relievers

Joe Maddon has been known, on more than one or two occasions, to buck convention just for the hell of it. Sometimes, in Maddon’s view, it hardly matters which way of going about some small, esoteric part of the game is right and which way is wrong. The trick is to make his team believe that their way of doing things is not only right, but unique, and therefore, a powerful secret weapon they wield over every opponent they face. The psychological edge you get from that sense of secret superiority and invincibility is worth more than the purely mathematical margin between two strategies.

In light of that attitude, it wasn’t hard to see Maddon’s approach to building an NL lineup coming. He implemented the second-leadoff-hitter theory right away, and he hasn’t gone away from it for a single NL-rules game all season. Maddon’s pitchers bat eighth, because the book (literally, The Book) says it can confer an extra run or two over the course of a season; and more than that, because it’s different.

Of course, if the pitchers bat eighth, someone still has to bat ninth. For a huge majority of this season, that someone has been Addison Russell. The result has been a bizarre rookie season for Russell, the youngest regular in the National League. He has run hot and cold, sometimes looking like a budding superstar, other times seeming to have trouble controlling the strike zone or driving the ball. In 381 plate appearances, Russell is hitting .241/.303/.371, which comes to a .249 True Average. That actually makes him a really good player, given the sterling defense he’s played on both sides of second base so far, but it’s an underwhelming offensive line, especially in contrast to the ones being put up by teammate Kris Bryant and fellow wunderkind shortstop Carlos Correa.

If you took the Cubs at their word when Russell was first shuffled into this role, you’re probably surprised to find him having such a seemingly rough season. Here’s Maddon from a piece by Patrick Mooney of CSN Chicago in mid-May:

“Here’s what you have to understand about him hitting ninth,” Maddon said. “He’s not going to see any better pitches than he’s going to see hitting ninth. I don’t want to put him in the one-hole or the two-hole right now. I don’t want to lay that on him.

“If you put him seventh — or put him eighth and put the pitcher ninth — he’s going to see (fewer quality pitches). The whole game plan, in my mind’s eye, by hitting him ninth, is twofold: To be the second leadoff hitter, in a sense, with a lot less pressure on you (and) the potential to see better pitches, because 1, 2 and 3 are hitting behind you.

“That’s where people are failing to think this all the way through. They just see the pitcher hitting eighth: ‘Oh my God, hit the pitcher ninth, because it’s been done like that for the last 150 years.’ (But) by him hitting ninth, he’s going to see better stuff to hit.”

If Russell is batting ninth to foster his development and ensure he’s in a position to succeed, why isn’t he thriving? The best way to answer that question, as it turns out, is to reject the premise. It isn’t true. Whether the Cubs know it or not, Russell is in a really tough spot. Maybe Maddon’s quotes were a misdirect, an easy way to explain away a strategy that is making his youngest, least experienced hitter’s rookie season a brutal grind, because someone has to do that job and it would be harder to convince a veteran that he would be better off hitting there. Maybe Maddon genuinely thought Russell would be in a position to thrive and is simply being proved wrong. In either case, the fact of the matter is that Russell is probably playing much, much better than you realize this season, a testament not to the role he’s playing but to his ability to adapt and succeed in the face of adversity.

The Full-Time Ninth Hitter
Of those 381 PA, 355 have come from the no. 9 slot. That’s 93 percent of his total. It’s extremely rare, these days, for a player to be so consistently slotted into the bottom spot. Obviously, it’s simply never happened in the NL, but this isn’t even done in the AL anymore. The second-highest percentage of total PA taken from the ninth spot this season (among batters with at least 350 total PA) belongs to the Tigers’ Jose Iglesias, at 54 percent. Over the two years prior to 2015, only two players qualified for the batting title with at least half of their total PA coming in the nine hole: Leonys Martin and (Joe Maddon’s Rays shortstop) Yunel Escobar, at 58 and 52 percent, respectively, both in 2013.

Unless a change is made, Russell will probably be the first player to qualify for the batting title with at least 90 percent of his PA coming as the ninth hitter since Gary DiSarcina in 1998. Yet, all the way up until DiSarcina did it for the last time, it wasn’t terribly uncommon for AL teams to field a light-hitting, good-glove player all season, batting him ninth just about every day. Since the DH made this sort of thing possible in 1973, there have been 50 player-seasons consisting of at least 350 total PA in which 90 percent or more came as the ninth batter, but only Russell, Jason Bartlett (2006), Adam Kennedy (2005), and Felix Martinez (2000) have done it this century.

There could be a half-dozen competing explanations for that, nearly all of them holding some grain of truth, none of them capturing the phenomenon perfectly. Part of it is simply the introduction of interleague play, which might have forced some weak hitters on AL teams to bat eighth for a week. Part is the shifts we’ve seen in scoring, and in teams’ evaluative priorities, even at defense-first positions. Part is that benches, while thinner and generally worse than they were 15 or 20 years ago, are used differently. More of a team’s fringe contributors cycle in and out of the starting lineup from one day to another, instead of a player starting a lot and being pinch-hit for later in games. Part is that lineups are much less consistent and rigid than they once were, with managers more often shuffling players just to see what sticks. Hitters don’t get slotted into concrete roles as often anymore, anywhere in the order. Part might even be an acknowledgment that batting ninth every day minimizes reps and leads to some really tough splits, so most players are negatively impacted by it.

What’s important is to establish that Russell is batting ninth more often than anyone has in nearly 20 years. That allows us to dig into the implications of that fact.

Seeing the Starter a Third Time (or Not)
This shouldn’t surprise you, but to start at the beginning: starting pitchers are facing 24.8 batters per game in 2015. That’s the lowest number ever.

That means that fewer starters than ever are getting as far as Russell’s spot at the bottom of the order for a third time. In fact, since the Cubs push opposing starters out sooner than all but two other teams, starters are even less likely to get to Russell than they are to get to, say, Omar Infante.

The end result is that Russell and the opposing starter have faced off for a third time in only 26 games, a number that constitutes just 7 percent of his total plate appearances. That’s easily the lowest figure in the league, well below Stephen Drew‘s 10 percent. It’s the lowest percentage, actually, for any batter with at least 350 plate appearances in any season for which we have this data. Chris Iannetta took 9 percent of his PA against the starter on the third go-round in 2009, but he did it in exactly 350 PA. Raise the bar to 500 plate appearances, and the lowest share of PA coming in that split belongs to Stephen Drew, who had 11 percent of his in that situation in 2013. This season, Alcides Escobar and Adam Eaton have each seen starters for a fourth time in a higher percentage of their PA than the percentage in which Russell has seen them a third time. Russell is one of a fistful of regulars who have not yet seen a starter for a fourth time this season.

Obviously, those plate appearances Russell isn’t getting against tired starters are instead coming against relief pitchers. As one might expect, he leads the league with 42 percent of his PA coming against them. (Drew is second, again, at 41.5 percent.) Perhaps because of the sheer volume of relief arms he’s seeing, or perhaps because he’s seeing them all for the first time, Russell is struggling mightily once a reliever comes on:

Addison Russell, Splits by Time Facing Opponent in Game, 2015

Those are radical struggles, to be sure, but it’s not unusual for hitters to struggle against relievers these days. The league is hitting a collective .243/.312/.379 against relief pitchers this season, but .270/.328/.438 the third time hitters are seeing a starter. The gap between expected production in those scenarios has steadily grown over the last 40-plus years.

So Russell, with his unusual breakdown of opportunities, is often seeing one of his plate appearances in a game turned from a good chance to do damage to a tough task. Starling Marte of the Pirates is the median regular both in percentage of PA coming against relievers and in percentage of them coming against starters the third time through the order. A quick comparison:

Addison Russell and Starling Marte, 2015

The stats are mostly there for reference; focus on the right-most column. The difference between Russell’s distribution of those mid-game PA and a typical one is jarring. While Marte’s performance doesn’t follow the league trend, if you figure that Russell’s generally would in a bigger sample (a very safe bet; some of the deviation from what you’d expect is simply BABIP right now, and Russell has mashed opposing starters the first two times through the order), you’re talking about seeing significantly more production from him if he were batting higher in the order and seeing more starters on their last legs.

Is There an Argument Here?
Obviously, augmenting Russell’s offensive numbers isn’t Joe Maddon’s job. The second-leadoff-hitter theory is a team-level strategy, and the guy whose job it is to bat ninth inevitably is getting the short end of the stick, whether the team wants to admit as much or not. There are only two questions we really need to tackle here:

1. Is Russell, a rookie who has had to make huge adjustments throughout the season and do so without the comfort of getting to see starters for a third time and avoid seeing many different relievers, suffering developmentally, and is he thus better off swapped with a hitter like Chris Coghlan or Miguel Montero?

2. Does the second-leadoff-hitter theory work in the Age of Relievers?

To answer the first question: I do tend to think that the Cubs should switch a veteran into that spot if they’re going to continue this lineup structure. Coghlan and Montero both have much smoother stats in terms of their splits against starters and relievers, and as the game progresses, than Russell has. More importantly, perhaps, Maddon is willing and able to pinch-hit for either of them if the opponent makes a pitching change and the matchup turns sour. The same can’t be said for Russell. Maybe the idea of keeping those vets a bit higher in the lineup is to try to force the opponent’s hand and get them to make a pitching change sooner, but that could actually have more value in the ninth slot. Most of the time, the Cubs aren’t going to let a pitcher get as far as the 27th batter under any circumstances, so by the time Montero or Coghlan would come up batting ninth, a reliever would already be in the game. If the opposing manager was forced into a bad decision at that stage—be it a wasted change Maddon could answer with a right-handed bat off the bench, or a change not made when it should have been because of the threat of that substitution—it could pay off big for Chicago.

To answer the second question: This probably belongs in the hands of someone smarter and better at mathematical modeling than I. All I’ll say is that the interaction of the times-through-the-order effect and the second-leadoff-hitter theory probably should be explored in greater detail. It’s possible that losing an at-bat for a good hitter against a tired starter cancels out the positive impact of being able to bat the team’s best hitters a slot higher. It’s also possible that the second-leadoff thing really only works if you happen to have a player with some OBP skills but a poor overall offensive profile. The Cubs don’t have that guy, and they might be better off simply stringing together their best hitters and minimizing the plate appearances they give to their pitchers.