After last night’s rude welcome to Coors Field, the Milwaukee Brewers boast a 52-62 record. If you’ve followed the Brewers’ batting performances in recent months, as well as their starting pitching, you might expect the Brewers to have a slightly better record. In fact, the Brewers’ current run differential of 518 RS / 528 RA suggests that the Crew should be able to claim a record closer to 56-58 than their current mark.
As J.P. Breen mentioned in his morning round-up, “The most glaring problem for the ’12 Brewers has clearly been the bullpen.” That’s about as clear as I can put it, too, but there’s an overall feeling about this ballclub that is nagging at me. My intuition is that, flat out, the 2012 Milwaukee Brewers are an unbalanced club. What I want to accomplish in this post is to explore that intuition.
On July 3, the 2012 Brewers distributed their runs scored in the following manner:
2012 Milwaukee Brewers
26-3 with 6+ RS
7-21 with 3-5 RS (4 wins between 3-4 RS)
4-18 with 0-2 RS
These divisions are rudimentary, but they hopefully capture the basic idea of how the Brewers perform (a) when they score significantly more runs than average, (b) when they score runs in an average range, and (c) when they score significantly fewer runs than average. Adjusting for Miller Park, the average runs scored per game for the National League (and Brewers) is approximately 4.45. Since the Brewers can’t score fractions of runs, these ranges should give the basic idea of how the Brewers perform in certain types of ballgames.
Compared to the 2011 Brewers, the 2012 club seemed to distribute their runs scored rather evenly:
2011 Milwaukee Brewers
49-9 in 6+ runs scored
40-21 in 3-5 runs scored (27 wins between 3-4 RS)
7-36 in 0-2 runs scored
At 79 games (the total going into July 3’s slugfest), these 2011 ratios suggest that the Brewers should have scored 6 runs approximately 28 times, 3-5 runs approximately 30 times, and 0-2 runs approximately, 19 times. At that point in the season, the differences were subtle, but the 2012 Brewers were a ballclub slightly more likely to score 0-2 runs in a particular ballgame, even while scoring 6+ runs at a rather strong rate.
Since July 3, the Brewers played a staggering 14 games in which they scored 6 runs. Those games happened to feature some of their most improbable and signature pitching blow-ups, too, and instead of helping to correct the season’s path, the Brewers only yielded 8 wins in those 14 games.
Furthermore, the Brewers also scored between 0-2 runs in nine games since July 3 (unfortunately, they went 0-9 in those games). In their “moderate” outings of 3-5 runs scored, the Brewers went 7-5 from July 3-to-present. This leaves us with the following ratios, including the expected performance using 2011 ratios in parentheses and italics:
2012 Milwaukee Brewers
34-9 with 6+ RS (approx. 35-6 with 6+ RS)
14-26 with 3-5 RS (9 wins between 3-4 RS) (approx. 28-15 with 3-5 RS, including 19 wins between 3-4 RS)
4-27 with 0-2 RS (5-25 with 0-2 RS)
On the surface, we might expect a ballclub that scores (at least) 6 runs more frequently to be a better ballclub than one that scores 6 runs less frequently. This might seem extremely reductive or an oversimplification, but even in a relatively high run environment (such as 2012 Miller Park), 6 runs scored provides the ballclub with a significantly above average offensive output. In contrast to this expectation of greatness, the 2012 Brewers (a) have lost a higher percentage of 6+ RS games than the 2011 Brewers, and (b) have played significantly fewer games with moderate run scoring output.
What does this mean?
Against their league, the 2012 Brewers bats boast an above average performance. Their total of 518 RS already places the club approximately 10 runs ahead of the league (and Miller Park) with just under 30% of the season remaining. The offense isn’t quite as good as the 2011 Brewers’ club, which boasted 721 runs scored against a league/park of approximately 694 runs.
Unfortunately, even though the Brewers’ offense is above average overall, the ballclub distributes their runs scored unevenly between strong (6+ RS) and weak (0-2 RS) games. This distribution diminishes their basic number of games in which they score a moderate number of runs; while one might suggest that simply scoring a lot of runs is a good thing, the strength of those runs scored is undermined if the club also frequently scores 0-2 runs.
Coupled with a bullpen that underperformed their expectations, the Brewers’ overall distribution of runs scored hurt their chances to win ballgames. If their 14-26 mark when the club scores 3-4 runs looks painful, think about this: 9 of those games are one-run losses with 3 runs scored; oddly enough, that set of games only includes 1 one-run loss with 5 runs scored, and 2 one-run losses with 4 runs scored.
(One might argue that this problem of run distribution is meaningless without a good pitching staff. The 2012 Brewers’ pitching staff is already approximately 60 runs off the pace of the excellent 2011 Brewers’ pitching staff. This results, of course, in more losses when the offense scores more than 6 runs. Furthermore, the club simply has to score more runs more frequently; in 2011 the Brewers allowed 2-or-fewer runs in 68 of their games; thus far, the 2012 Brewers allowed 2-or-fewer runs in 29 games — on pace for 41 such games over a full 162).)
If we look at the Brewers’ overall performance in one-run games in 2012, we might be further inclined to pin their overall record on the pitching staff. Frankly, this is a reasonable assumption given the fact that the pitching staff has performed much worse than the offense over the course of the season. Furthermore, the number of recent implosions in games where the Brewers managed to score 6 or more runs makes it even easier to pin the blame on the bullpen.
Batting Order Dyanmics
However, my feeling is that there’s more to the story in 2012. Why is it that the Brewers are seemingly doomed once relievers blow the game? Of course, I’m excusing games in which the Brewers already scored a bunch of runs and the ‘pen STILL gave up the lead; when the club comes back — as they did in Philadelphia recently — or continues to add runs — as they did against Washington recently — and the bullpen surrenders the game, it’s unreasonable to say, “hey, you already scored 10, why not score 15?” or something like that. BUT, what I’d like to know is, is it reasonable to ask an offense that frequently scores 6 runs to simply push across a few more runs when the Brewers are trailing, say, 4-to-3?
One of my favorite, quick and dirty ways to get a picture of how a player is performing at the plate is to look at his R and RBI together. I know that these stats are extremely dependent upon teammates, batting order, and context, but it’s a great way to simply see, “is this player performing up to his actual stats?” Which is really to say, “is the manager employing this player’s talents in a context that encourages runs production?”
Yuniesky Betancourt is my favorite example of this question, from the 2011 Brewers. Although Betancourt’s overall extreme contact approach and batting statistics suggested that he was a below average shortstop, his place in the Brewers’ batting order (and his ridiculous ratio of batting the ball in play) provided him great opportunity to produce runs for the Brewers. Although his .672 OPS failed to measure up against the NL SS OPS of .688 (and that’s before park adjustments!), Betancourt’s sacrifice flies and batting order position allowed him to produce runs at a rate consistent with the NL average for SS. This disjoint helped to fortify the Brewers’ batting order, and provided a productive dynamic throughout the line up in 2011.
My quick and dirty way to look at R and RBI is to use a harmonic mean ((2*R*RBI)/(R+RBI)) to combine a player’s run production into one extremely oversimplified metric. Combining R and RBI in this way helps to place a player’s run production into his batting order spot, and also expresses overall run production in a number that compares to a basic runs created calculation. (For instance, if a player scores 100 runs and bats in 75 runs, we wouldn’t say that player is worth 175 runs overall; but, saying this player is worth approximately 86 runs helps us to compare his actual runs produced on the field to an abstract estimate drawn from his batting stats. This basic metric helps to show just how much some lead-off hitters’ batting stats are skewed by their place in the order, and it immediately places a player in their team context).
Moving beyond simply placing R and RBI together, we can also place those statistics against PA to create a simple ratio of expected runs produced. This is especially valuable to produce at a league level, in order to compare league production to individual production. This is an especially valuable starting point for comparing players against their batting order and fielding position.
My basic calculation to estimate the expected R/PA for each fielding position is to:
(a) Calculate “RRBI” by taking the harmonic mean of R and RBI: ((2*R*RBI)/(R+RBI))
(b) Place RRBI against PA: (RRBI) / (PA)
This equation does not explain anything other than a player’s or fielding position’s or batting order’s performance within the context of a league, team, park, etc. It’s extremely dependent on other players and the overall interaction of a batting order. But, it can be easily compared to runs created by taking RC/PA, and in that case we can provide a foundation for whether a particular player is underperforming or not.
Although this type of analysis is certainly flawed, it provides a basic starting point for comparing the dynamic of different batting orders. For instance, even though the 2011 Brewers had a good offense overall, their three worst fielding positions (C, 3B, SS) put together batting stats that compared unfavorably to the league. However, because of the overall composition of the batting order, and the fact that the Ron Roenicke largely confined those bats to the low-middle and low batting order spots, the Brewers were able to minimize the production damage from these positions:
C (621): .1036 R/PA (.708 OPS) (Brewers: .1066 R/PA, .696 OPS) (approximately 64.3 NL R vs. 66.2 MIL R)
3B (678): .1108 R/PA (.705 OPS) (Brewers: .0864 R/PA, .598 OPS) (approximately 75.1 NL R vs. 58.6 MIL R)
SS (650): .1028 R/PA (.688 OPS) (Brewers: .1013 R/PA, .672 OPS) (approximately 66.8 NL R vs. 65.8 MIL R)
Overall, the 2011 Brewers’ three worst fielding positions probably cost the ballclub at least 15 runs over the course of the season (maybe more, depending on how we build from this analysis). However, the bulk of their trouble from these positions was thanks to Casey McGehee at 3B; despite a below average OPS, the Brewers yielded an average performance from their catchers, and their shortstops were not far from the overall league average.
By comparison, the 2012 Brewers’ three worst fielding positions yielded consistently below average run production for the ballclub. These performances arguably affect the 2012 Brewers in a worse way, given the batting order positions afforded these fielders throughout the season. Overall, these fielding positions are on pace to provide the Brewers approximately 100 more PA from their worst performers in 2012 (compared with their three worst in 2011).
2B (496): .1110 R/PA, .716 OPS (Brewers: .1045 R/PA, .710 OPS) (approximately 55.1 NL R vs. 51.8 MIL R)
CF (505): .1130 R/PA, .739 OPS (Brewers: .0968 R/PA, .694 OPS) (approximately 57.1 NL R vs. 48.9 MIL R)
SS (439): .1043 R/PA, .698 OPS (Brewers: .0948 R/PA, .651 OPS) (approximately 45.8 NL R vs. 41.6 MIL R)
While Rickie Weeks‘s improvements at 2B might help this lot, and bring the Brewers’ 2B performance to the NL average, the placement of Nyjer Morgan and Carlos Gomez in the batting order exposes their performances to more plate appearances (thereby influencing the batting situations for Ryan Braun and Aramis Ramirez to some extent).
Players from the Brewers’ three worst fielding positions have started in the 1st or 2nd batting order spot approximately 133 times in 2012; in 2011, the Brewers’ 3B, SS, and C did not start in the 1st or 2nd spot more than 10 times. As a result, the Brewers’ least productive bats have fewer opportunities to produce behind Braun and Ramirez, and instead, they simply must rely on their teammates to pick up their performance. I believe that this is the reason that none of the 2012 Brewers’ worst fielding positions approach the NL average; their batting order placement does not maximize their ability to produce runs.
Since the 2011 Brewers’ least productive bats regularly spent time behind the middle of the order, I gather that they were able to ride the coattails of Braun and Prince Fielder. By contrast, the 2012 Brewers’ least productive bats are regularly in the position to set the table for the club’s more productive bats, which should mean that (a) they have fewer opportunities to drive in runs, and (b) they influence the conditions for the more productive bats to produce.
My point here is kind of opaque, and it needs more analysis. The basic argument is that because of the batting positions of the Brewers’ least productive players in 2012, the Brewers were unable to consistently distribute their run production in a manner that helped them to win close games. This is a bit of “hocus pocus,” I suppose, but I’d wager that the overall plate appearances afforded to the Brewers’ least productive batters helped contribute to the season’s Murphy Law — or at least, the overall feeling that the ballclub is consistently out-of-sync.
At the very least, we can help to explain why the Brewers have been able to score 6+ runs more frequently while continuing to consistently score 0-to-2 runs. This basic distribution of runs scored helps to contribute to the overall composition of a ballclub that simply can’t get the best performance out of its parts. If we’re looking for those extra 4 wins, we can begin to understand the systematic culture of underperforming this season.
Ransom (Associated Press): http://blog.pennlive.com/patriotnewssports/2012/07/cody_ransoms_grand_slam_powers.html