Winning on the margins

Every team has their core players. They’re the ones who are supposed to carry the team. And all teams have their marginal players on the periphery. No one expects those guys to be good. They are just warm bodies stepping in to give a core player a day off or fill in for an injury or pitch some mop-up innings.

No one expects a peripheral player to be good, but that doesn’t mean that the fringe folk don’t have an impact. Once in a while, a team can get really nice production from its fungible players, and unexpected club depth can pad the season’s win total.

Then again, just because peripheral players aren’t expected to be good doesn’t mean they can’t underachieve. There are degrees of bad, and what happens if the end of the bench isn’t just bad by normal ballplayer stands, but also sucks by end-of-the-bench player standards? Then you’ve got a problem. Then a team can pick up extra losses.

Those peripheral players—they are so easy to overlook, but they can make quite a difference, maybe even the difference between a postseason appearance and an October off.

So let’s look at which teams have gotten the most from their men on the margins through the first half of 2013.

Defining the core and the periphery

So what exactly is meant by peripheral player or core player? For purposes here, anything not coming from the core comes from the periphery.

There are three places to look: offensive players, starting rotation and bullpen.

Offense: The core offensive players are the guys listed at Baseball-Reference.com as the top player at each position for the team. The periphery? Those are all the other guys. The offensive contributions of all non-pitchers who aren’t in the core—that’s the periphery.

Starting rotation: The five pitchers with the most starts are the core. All other starts belong to the periphery. (Note: in case of swingmen who have starts and relief appearance, only their performance as starters counts here).

Bullpen: The reliever with the most saves is automatically in the core. Joining him are the four other relievers with the most innings pitched out of the bullpen. (Again, with swingmen we’ll look only at innings in this role, not overall innings.) Usually a closer has among the most relief innings, but sometimes he doesn’t—and it doesn’t make any sense to declare the closer not a member of the core. At any rate, everyone out of those five are peripheral relievers.

Offense by the outsiders

So let’s find out how to do this. First, let’s get the info. That’s simple enough: It’s team stats minus core guys’ stats (and, for offense, minus what the pitchers did at the plate).

Now that we have the raw data, we pick some key metrics. Here’s the key point: Let’s use rate stats.

There are two things when looking at the periphery’s performance: their quality and quantity of production. For this study, we want to compare quality. You want to look at how a team’s peripheral players compare to what a normal group of peripheral players would do in the same amount of playing time. It’s really key to adjust back for playing time. Otherwise you’re learning more about how much a team uses its extras rather than how the extras are doing.

So let’s use Runs Created (RC)per 27 outs. Here are how the benches for all 30 teams do via that stat (adjusted for park):

Team	RC/27
BOX	5.30
ATL	4.82
CLE	4.43
CHC	4.27
STL	4.26
KCR	4.01
DET	3.95
LAA	3.88
NYM	3.85
TBR	3.84
SEA	3.74
TOR	3.72
PIT	3.69
SDP	3.64
SFG	3.63
LAD	3.62
ARI	3.60
MIA	3.58
PHI	3.57
MIN	3.24
TEX	3.22
BAL	3.21
HOU	3.19
OAK	3.18
COL	2.78
MIL	2.70
DCN	2.67
CIN	2.62
NYY	2.46
CWS	2.28

Bench players in the AL average 3.66 runs per 27 outs. In the NL, it’s a hair lower: 3.62.

Apparently, the Red Sox are having a mighty nice time of it with their bench. In fact, that’s actually better than the Boston big nine are doing. They’ve relied on three main bench players, all of whom have hit well: Daniel Nava, Jose Iglesias and Mike Carp. That’s carried the team’s bench score.

Strange as it might sounds, through the All-Star break, five teams had their benches outhitting their starters: Boston, Kansas City, Atlanta, Miami and the Cubs. At the other extreme, three clubs have a fall-off of more than two runs per 27 outs from their core to their periphery: Cincinnati, Milwaukee, and worst of all Colorado (whose core is doing 2.7 runs better per 27 outs).

You can reverse-engineer this to get a sense of what an “average” periphery would do for each team based on the actual bench’s playing time. Let’s look at that Red Sox team. Well, those guys made 660 outs prior to the All-Star break. Given that an average AL periphery would produce 3.66 runs in 27 outs, that means an average team would produce about 89 runs. Adjust for park, and it’s 93.9 runs created by an average bunch of bench players. In reality, those real life Red Sox have produced 129.5 runs (adjusted for park) in their playing time.

Boston’s bench bats have given the team 35.6 more runs then a normal bench would’ve done in their playing time. When you do it for all 30 squads, here are the results:

Team	Runs
BOX	 35.5
CLE	 26.0
ATL	 22.4
SEA	 12.7
NYM	 12.0
CHC	 11.2
SDP	 10.7
STL	  9.4
SFG	  8.4
PIT	  7.5
TBR	  7.1
LAD	  6.4
LAA	  6.3
KCR	  5.8
DET	  1.6
MIA	 -0.9
TOR	 -1.2
ARI	 -2.3
PHI	 -2.7
OAK	 -7.2
BAL	 -8.0
MIN	 -9.1
HOU	-10.7
TEX	-11.5
CIN	-22.4
MIL	-24.1
DCN	-24.6
NYY	-28.6
CWS	-30.9
COL	-44.2

That’s a six game difference between the Red Sox and the Yankees. That’s the difference in their seasons so far; the gap between first and fourth places—not the pitching, not the front line hitting, just in their bench patters.

Meanwhile, the bench batters have sunk the Rockies. The highest OPS+ by any of them is 83. That’s normally where replacement level is. So the entire bench is below replacement level. That’s bad.

Starting pitchers

Let’s try the same approach with starting pitchers. Instead of focusing on Runs Created, the metric of choice is something much more traditional: ERA.

AL peripheral pitchers have a total ERA of 5.51. In the NL, it’s a lot different, actually: 4.92. Not sure what to make of that, but since I compared hitters to the AL or NL total and not all major league baseball, we’ll do the same here.

Some teams have had to use extra starting pitchers a lot more often than others. At dueling extremes, the Pirates had 28 starts from outside their top five before the All-Star break, while the Braves had just one.

Still, the system adjusts for playing time—it’s how many runs normal peripheral starters would’ve given up in the innings they had versus how much the team’s actual auxiliary arms did. As always, park will be accounted for. Here are the results, with the best extra pitchers on top and the worst on bottom:

Team	R.Saved
MIA	 20.2
TBR	 17.1
NYY	 13.5
CWS	 13.0
TEX	 11.6
ARI	  9.6
CHC	  9.3
CIN	  6.7
TOR	  4.9
KCR	  3.0
NYM	  2.8
DET	  1.7
ATL	  1.1
DCN	  0.9
STL	  0.7
PIT	 -1.9
BOX	 -2.4
MIN	 -3.3
LAD	 -4.0
SEA	 -4.5
SFG	 -4.6
OAK	 -4.7
PHI	 -6.8
MIL	 -7.0
LAA	 -7.7
COL	-11.2
BAL	-12.6
CLE	-12.9
HOU	-13.2
SDP	-14.0

Miami’s extra pitchers have been a godsend. As bad as the Marlins are, they’d be historically bad if they had the back of the Padres’ rotation. The Marlins’ extra arms have a 3.70 ERA (in a slight hitters park) in nearly 140 innings, the most by any team. The big help here is Jacob Turner, who has a 2.44 ERA in nine starts at the midpoint, sixth most on the team. He actually has more innings than Alex Sanabia, but Sanabia has one more start, so is considered part of the core. Flip them, and Miami’s score drops.

The Padres’ extra starters have an ERA around 7.00, despite performing in one of the game’s best pitchers’ parks. In 12 starts, they’re averaging barely four runs per outing.

Six teams actually are getting a better ERA from their peripheral arms than their five core starting pitchers: Arizona, Atlanta, Miami, Kansas City, Tampa and the Yankees.

Bullpen

Lastly, let’s look at how the extra relievers are doing. Once upon a time, teams had just a five-man bullpen, but that’s changed. Still, by that sixth reliever, the bullpen typically looks bleak.

The usage split here isn’t nearly as stark as it was with starting pitchers. The fewest innings came from the White Sox’ extra relief pitchers, with fewer than 80 innings. The most is San Diego (of course! What with the Padres’ horrible marginal starting pitching). They had 138.1 innings pitched from auxiliary relievers before the All-Star break.

Again, we’ll use ERA. AL peripheral pitchers have a total ERA of 4.89. In the NL, it’s a lot different: 4.01. That’s an even bigger split than with starters. We’ll keep comparing teams to just their leagues, not all baseball. Here are the results:

Team	R.Saved
MIL	 26.4
TOR	 19.1
MIN	 18.7
KCR	 14.8
ARI	 14.1
TBR	 12.7
SFG	 11.2
SDP	 10.7
ATL	 10.2
TEX	  8.3
OAK	  7.0
CHC	  4.4
NYY	  4.3
CIN	 -0.5
BAL	 -1.0
COL	 -2.4
DET	 -3.8
PIT	 -4.7
DCN	 -5.1
LAA	 -5.1
PHI	 -7.9
NYM	 -9.4
CWS	-10.8
CLE	-13.0
MIA	-13.6
STL	-13.7
BOX	-14.6
HOU	-15.4
SEA	-15.7
LAD	-18.6

How ‘bout them Brewers? They have a terrible starting rotation, and their backup position players leave much to be desired, but they have tons of depth in their bullpen.

Actually, their high score is partially caused by how you divvy up core versus periphery. Francisco Rodriguez is still listed as periphery, even though by the time you read this he might lead the team in saves. He didn’t when I got the data, but he does have an ERA barely over one.

As a result, the Brewers are one of six teams with a better ERA from their peripheral relievers than their core ones, along with the Cubs, Diamondbacks, Padres, Giants and Rays. That’s five NL teams and one AL. No wonder the overall ERA for NL peripheral relievers is so much better—their clubs keep giving more innings to worse relievers.

Putting it together

Now that we have all three areas, let’s see what teams have gotten the most from the end of their benches and which have gotten the least:

Team	   Bat	SP	RP	ALL
TBR	  7.1	 17.1	 12.7	 36.9
ATL	 22.4	  1.1	 10.2	 33.7
CHC	 11.2	  9.3	  4.4	 24.9
KCR	  5.8	  3.0	 14.8	 23.6
TOR	 -1.2	  4.9	 19.1	 22.8
ARI	 -2.3	  9.6	 14.1	 21.4
BOX	 35.5	 -2.4	-14.6	 18.5
SFG	  8.4	 -4.6	 11.2	 15.0
TEX	-11.5	 11.6	  8.3	  8.4
SDP	 10.7	-14.0	 10.7	  7.4
MIN	 -9.1	 -3.3	 18.7	  6.3
MIA	 -0.9	 20.2	-13.6	  5.7
NYM	 12.0	  2.8	 -9.4	  5.4
PIT	  7.5	 -1.9	 -4.7	  0.9
CLE	 26.0	-12.9	-13.0	  0.1
DET	  1.6	  1.7	 -3.8	 -0.5
STL	  9.4	  0.7	-13.7	 -3.6
MIL	-24.1	 -7.0	 26.4	 -4.7
OAK	 -7.2	 -4.7	  7.0	 -4.9
LAA	  6.3	 -7.7	 -5.1	 -6.5
SEA	 12.7	 -4.5	-15.7	 -7.5
NYY	-28.6	13.5	  4.3	-10.8
CIN	-22.4	  6.7	 -0.5	-16.2
LAD	  6.4	 -4.0	-18.6	-16.2
PHI	 -2.7	 -6.8	 -7.9	-17.4
BAL	 -8.0	-12.6	 -1.0	-21.6
CWS	-30.9	 13.0	-10.8	-28.7
DCN	-24.6	  0.9	 -5.1	-28.8
HOU	-10.7	-13.2	-15.4	-39.3
COL	-44.2	-11.2 	-2.4	-57.8

Eleven teams have a periphery topping their core in at least one of these three areas. Six teams actually have their margin men topping their central players in two areas: Kansas City, Arizona, Atlanta, Miami, Tampa and the Cubs. Sometimes they’re getting great work from the end of the bench. In some cases their foundation has let them down. And in some cases it’s both.

So it isn’t too surprising to see those five of those teams in the top six overall. (Miami’s peripheral hitters might be worse than normal end-of-the-bench production, but it’s still a touch better than what the team’s starting position players are doing. Scary reality for south Florida).

Colorado had just been killed, absolutely killed, by the end of its bench. Sure no team expects to get much from its backups, but Colorado’s guys haven’t just been bad, but several steps below that.

What does that mean for their record? Let’s figure it out. Below is what happens to teams’ records if you adjust their midseason records by this info (using standard shorthand of 10 runs to one win):

Team	AdjW	AdjL	Adj.Pct
BOX	56.1	40.9	0.578
BAL	55.2	40.8	0.575
NYY	52.1	42.9	0.548
TBR	51.3	44.7	0.534
TOR	42.7	51.3	0.454

DET	52.1	41.9	0.554
CLE	51.0	44.0	0.537
KCR	40.6	51.4	0.441
CWS	39.9	52.1	0.434
MIN	38.4	53.6	0.417

OAK	56.5	38.5	0.595
TEX	53.2	41.8	0.560
LAA	44.7	48.3	0.481
SEA	43.8	51.2	0.461
HOU	36.9	57.1	0.393

DCN	50.9	44.1	0.536
ATL	50.6	44.4	0.533
PHI	49.7	46.3	0.518
NYM	40.5	50.5	0.445
MIA	34.4	58.6	0.370

STL	57.4	35.6	0.617
PIT	55.9	37.1	0.601
CIN	54.6	40.4	0.574
CHC	39.5	53.5	0.425
MIL	38.5	55.5	0.410

COL	51.8	44.2	0.540
LAD	48.6	45.4	0.517
ARI	47.9	47.1	0.504
SFG	41.5	52.5	0.441
SDP	41.3	54.7	0.430

Only nine teams actually have their places change from where it actually was as of the All-Star Game, but some of those changes do have playoff implications.

Tampa falls from second to fourth place in the AL East. So the Rays’ October hopes are based heavily on roster depth.

In the NL East, the Nationals and Braves flip positions. That’s surprising given that the Braves had a big six-game lead at the break, but that’s because their backups have been great while the Nationals’ performance has been all too marginal at the margins.

Finally, the Rockies would’ve ended the first half of the year in first place not by getting great production from their peripheral players, but by getting just the standard lousy play teams expect from those guys. Because it’s been so much worse, their postseason hopes look bleak. They’re a club that might miss October because their weakest links are indeed that weak.

References & Resources
Info comes from Baseball-Reference.com

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Fantasy Waiver Wire: Week 17, Vol. 1
Next: NL East division update: July edition »

Comments

  1. Brad Johnson said...

    I noticed the Nats’ poor performance over the weekend for my update article. They’ve had 5 players combine for over 700 PA of truly execrable performance. Roughly .175/.210/.225. One of those guys is Danny Espinosa.

    I assume that you have Rendon (one of the bright spots) as a periphery and Espinosa as a core 9. What happens to their scores if you flip them around?

  2. Bryan Cole said...

    Is the DH for NL teams counted as periphery or core?

    And how much of the Braves’ offensive score is Evan Gattis being incredible and B.J. Upton being a different type of incredible?

  3. Chris J. said...

    Bucdaddy – they’re saved by Garret Jones, who is listed as a backup at B-ref.  Some of the scores are determined by how you slice up core/periphery. There’s no great way to figure that one out.

  4. blw said...

    Yes, it could be true that Atlanta’s starters might be out-averaged by the peripheral starters – but it’s worth observing that the peripheral start category consists of exactly one start by a pitcher other than the five in the primary rotation.  That one start went pretty well, so it’s hardly surprising that the regular rotation was “beaten.”  There’s likely going to be a second start from outside the rotation later this week, but this is hardly an indication of anything wrong with the ATL starting rotation.

  5. pat rocket said...

    Pitcher hitting should be peripheral, especially if you count NL dh abs that way.  Any offense a pitcher gives you is gravy.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>