When Stars And Scrubs Doesn’t Work
The Milwaukee Brewers have two of the top ten players in WAR, and neither of them is named Ryan Braun. Carlos Gomez (+3.2 WAR, #1) and Jean Segura (+2.5 WAR, #9) have been revelations for the Brewers, while Braun has been his usual dominant self, putting up +1.9 WAR with his usual brand of excellence. Toss in a strong performance from the underrated Norichika Aoki and 78 terrific plate appearances from the occasionally healthy Aramis Ramirez, and the top end of the Brewers line-up has been as good as any in baseball. For context, here is the total combined line for Gomes, Segura, Braun, Aoki, and Ramirez:
| Name | PA | AVG | OBP | SLG | wOBA | wRC+ | Fld | BsR | WAR |
|---|---|---|---|---|---|---|---|---|---|
| Big Five | 875 | 0.323 | 0.384 | 0.532 | 0.393 | 153 | 6.2 | -2.3 | 8.7 |
That’s a stellar performance. For comparison, here’s how that stacks up against the “Big Five” from the Cincinnati Reds, a team that has scored 245 runs, second in the NL only to the Colorado Rockies.
| Name | PA | AVG | OBP | SLG | wOBA | wRC+ | Fld | BsR | WAR |
|---|---|---|---|---|---|---|---|---|---|
| Reds Big Five | 1110 | 0.287 | 0.381 | 0.485 | 0.371 | 136 | 2.2 | 1.2 | 8.3 |
Even with injuries limiting the amount of playing time that Ramirez (and Braun, to a lesser extent) have gotten, the Brewers top tier players have outperformed the Reds group overall, and have hit significantly better than Cincinatti’s quintet, even though both Shin-Soo Choo and Joey Votto have also been fantastic this year.
But, as we near the beginning of June, the Brewers have scored just 196 runs and their position players have just a 103 wRC+, 14th best in baseball. Because, here’s what the hitters on the Brewers beyond the Big Five have done.
| Name | PA | AVG | OBP | SLG | wOBA | wRC+ | Fld | BsR | WAR |
|---|---|---|---|---|---|---|---|---|---|
| Everyone Else | 855 | 0.200 | 0.256 | 0.304 | 0.249 | 52 | -10 | -0.6 | -3.3 |
There are 10 players in that group — including regulars Yuniesky Betancourt, Jonathan Lucroy, Rickie Weeks, and Alex Gonzalez, plus every reserve the team has used — and they’ve combined to hit almost exactly as often as the Brewers best hitters. The best offensive performance by anyone in that group belongs to Betancourt, who has a 78 wRC+. All 10 of the players in that group have either been exactly at replacement level or below.
A graph of the Brewers position players performance shows the distinctions.
Ramirez’s poor defense and lack of playing time have made him the team’s only position player with a WAR between +0 and +1.0. Every other player has either been excellent or terrible. There’s no middle ground here. This is the clearest example we’ve seen in some time of how a true stars and scrubs team would do. And the answer is not particularly well.
Granted, the Brewers scrubs have been worse than you could reasonably expect anyone’s scrubs to be, but their stars are also performing at an extraordinarily high level as well. Even if you regress the Brewers scrubs back towards replacement level, you have to do the same for guys on the high end, and then the adjustments mostly cancel out. No matter what we’ve been told, a few great players simply cannot carry a bunch of terrible ones to success. The great performance of a few just cannot overcome the brutal performance of the many.
I know it’s probably old hat for most FanGraphs readers by this point, but every roster spot matters. The down-roster role players count too. Major League Baseball is not a sport where star performance wins. It is a sport where even the best single player in the game doesn’t make that much of a difference. The Brewers, by 2013 performance, have had three of the very best hitters in baseball, and they still have an average offense and are headed for the top ten in next summer’s draft.
What separates a good team from a bad team is often not the quality at the top of the roster, but the quality at the bottom of the roster. The Brewers have some really good pieces, but they also have more dreck than just about anyone besides the Astros and Marlins. And that kind of roster construction just doesn’t really work in this sport.
Dave is the Managing Editor of FanGraphs.
I thought this was going to be an examination of the 2013 Los Angeles Angels of Anaheim.
That was my reaction to the head line: 80%p Angels, 20% Dodgers, 10%p Toronto. (I know).
Actually, the Angels are an example of a team constructed to not have any “soft spots” in the lineup, no true scrubs as starters. Their worst positions are still rather solid: Iannetta/Conger at C and Callaspo at 3B. Shuck has filled in pretty well for Bourjos (.361 OBP).
In other words, the team–at least going into the season–was meant to be a core group of three stars surrounded by a bunch of average to above average complementary players. Not bad plan, even if hasn’t worked out as hoped (until the last couple weeks, at least).
too many chiefs, not enough indians
Last year’s position players finished #1 in WAR. The lineup is constructed to be great all around and it really hasn’t been a problem this year.
This is not true with regards to the pitching staff, which is just about all scrubs all the time.
Even funnier, I thought it was going to be about the LA Dodgers.
Excellent use of the word “dreck.”
The Brewers pitching staff was constructed similarly, as well.
Except that Doug Melvin forgot the “stars” part of “stars and scrubs”.
Except without the stars?
As a Brewers fan, I had an inkling this article was going to be about the Brewers…
I do see the roster evening out a bit. As Gomez and Segura cool, Lucroy and Weeks should improve, and Hart will return from injury. When healthy, the Brewers have a very solid 1-8. The problem is that they have zero depth and injuries happen to every team. A lot of bumbling backups (Alex Gonzalez at 1B?) have been pushed into action for the crew, augmenting the weight of Weeks’s brutal beginning and Lucroy’s slow start.
Contrast the Brewers to the Pirates, who have only two position players below 0.0 WAR–Inge at -0.1 and John McDonald at -0.5. But Inge and McDonald have just 105 PA between the two, with 70 of those PA beloning to Inge, who still provides value with his glove.
The Pirates’ 12 0.0+ WAR position players break down like this:
2.0
1.9
1.7
0.7
0.6
0.5
0.4
0.3
0.2
0.2
0.2
0.0
Pretty good mixture it seems.
Has a team with reasonable playoff aspirations ever started off a season using a backup shortstop as their everyday 1st base man? I mean, you can always dig up a Brandon Moss, Corey Kotchmann or Daric Barton and hope for the best, no matter when in spring training the injuries happen. I don’t know that I’ve ever seen so inexplicable a self-inflicted MLB wound.
They tried to sign Lyle Overbay before spring training, he turned them down to sign with the Red Sox. And Mat Gamel got hurt.
Understood on Gamel, but you can always very cheaply scare up a Quad-A guy like Brandon Moss was, like Daric Barton is. When was the last time a reasonable contender settled for a month+ of their backup shortstop starting at 1st base? Which has now turned into 2 months, incidentally.
I watched them play a four game series against the Pirates and they had a different starter at first each game. I think it was Yuni, Gonzalez, Maldonado, and someone else who I can’t recall at the time.
Probably Lalli. He’s the AAA catcher, so it makes sense they would start him at 1B.
1B is the weakest position in MLB right now, so if it’s ever going to happen it’s now. Doesn’t excuse playing the Betancourt/Gonzalez platoon, of course.
It probably doesn’t help when your rotation is at least 80% scrubs. I feel like his article is like moneyball, the movie. Entertaining, yet completely focuses on offense as reason for success/failure.
This article is solely focused on runs scored not team wins. Nothing like Moneyball.
A couple of these players seem unlikely to perform as poorly as they have done thus far, notably Weeks and Lucroy. Of course that’s apt to be offset by some regression of the high performers.
The real issue is a porous infield defense behind pitchers who don’t get many strikeouts.
And this, I think, is the mystery of the Giants.
Correct, 9.7 WAR for the Giants starting lineup.
Dave had a speculative article at the end of last season which argues the Giants win with depth. Basically, get a couple stars (Posey and Sandoval), and fill the rest out with solid players.
http://www.fangraphs.com/blogs/the-giants-win-with-depth/
If they had a #1 pitching staff it would almost cut it, but yeah. The A’s are a team that usually benefits in the other direction from this, they have a deep roster when it comes too replacement level players. One of the biggest reasons they succeeded last season, was having replacements turn into upgrades. Beane saw what happened with his stars heavier teams the last few years, and realized depth, and roster versatility was the key to success. I do like how most players on the team play multiple positions, giving overlap and redundancy. Being able to move things around a lot with your roster lets you always take the best replacement option at the time, instead of the best positionally specific.
The Diamondbacks also seem like a case of a team with a balance of quality players despite whatever the hell it is that Kevin Towers is doing.
They would be heading for a top ten pick in the draft – if they hadn’t surrendered their first round pick by signing Kyle Lohse. Winter is coming Brewers fans.
They lost their 2013 first round pick, which was #17. The 2014 top 10 pick is still safe (for now).
Top 10 picks are protected. If they do get a top 10 pick, they will surrender their next highest draft pick for Lohse.
right, except, that Bob M is also correct, it is the wrong year. Top 10 picks are protected, but it doesn’t matter unless they sign a free agent who received a qualifying offer from their former team during this offseason.
I am not incorrect on the year. In return for signing Lohse, the Brewers lose their draft pick in next month’s draft. Which occurs in 2013, and is the #17 pick.
The Brewers are currently playing like a team that will get a top 10 pick. That will occur after the season results are finalized – which means it occurs in 2014.
You guys are not disagreeing! You both think it’s the wrong year. Reconcile, and go forth as brothers.
More than the Brewers offense (#22 in runs scored), their starting pitching (#27 in runs allowed, #27 innings pitched) has been the reason they are a truly bad baseball team. They give up a lot of runs, but make up for it with short outings.
The bullpen is actually pretty solid (#13 in runs allowed, #10 in innings pitched), which is almost a miracle after how bad it was last year (#29 in runs allowed, #9 in innings pitched).
If the rotation is only 27th in innings pitched, then the bullpen by definition will be doing quite well in that same metric (if you can call it that).
A much better formula is the Orioles: Stars, and scrubs playing like stars. (Chris Dickerson 0.5 WAR in 19 games? lol.)
Dickerson isn’t a scrub. He’s a decent platoon/4th OF-type who has gotten few opportunities to play, partly due to injuries. Seriously, his career numbers are fine.
Why use WAR and not wRC+ or wRAA when you’re mentioning their difficulty in scoring runs? Gomez doesn’t look quite as elite when you use those instead of WAR. I think the argument still stands, but you should probably graph something instead of WAR to show the Brewer lineup’s extremes (e.g. in wRAA Gomez is 9th among MLB qualifiers, compared to his 1st place standing in WAR).
Another thing to think about with the Brewers that this doesn’t show is where in the lineup these players are batting. Gomez has batted further down in the lineup (6th or 7th) for a good portion of the games so far this season (RR is finally moving him up further in the order), with Weeks batting 5th and 6th (even 4th for a couple of games) in a fair amount of games this season. Putting a .160 hitter who strikes out a TON in the 5th or 6th spot in a lineup can really kill chances at multiple run innings.
While there is no excuse for their 1B situation ,if the Brewers even had a mediocre starting staff they could be in contention in their division .
The problem is also their division. The only two teams that are farther out of 1st in their divisions are Houston and Miami, which is not to say that Milwaukee is anywhere near as bad as they are. In fact, they have a winning percentage 100 pts higher.
So, Dave, in dissecting why the Brewers are no good, you didn’t see fit to mention the fact their rotation has a 5.29 ERA, worst in their league? You actually think their WAR through less than 50 games does a better job explaining this?
Good God, Dave, change the record already.
Never mind for a second the fact you’ve ignored the Brewers’ run prevention in its entirety (they’ve allowed the most runs in the NL, 4th in the majors, in case you give a shit) but Just so that you’re clear Dave– you think less than 50 games worth of defensive metrics should be taken completely at face value, right? Because by referencing these guys’ WAR totals, that’s precisely what you’re doing.
I get that Ramirez isn’t a good defender, especially at his age. I get the Gomez is. But Dave, you either have to accept the obvious shortcomings of metrics like UZR in meaningless samples– which is what less than 50 games worth of data is– or just go ahead and admit you don’t care about that. You can’t have it both ways though.
You’re happy to pay lip service to this idea, but then when you evaluate why a team has a poor 50 game stretch, you don’t hesitate to lend this data complete credence. So which is it, Dave?
Meaningless sample sizes have to do with predictability and repeatability.
Small samples as used in the article truly capture what actually happened.
Remember the first day of the season and every makes their “Player X is on pace to hit 162 HR” joke. That sample size is meaningless to actually predict a season’s home run total…but I’ll be gosh-darned if Player X didn’t actually hit a homerun in that first game.
If you think that UZR and similar measures are decriptive statistics that measure “what actually happened” than you have a fundamental misunderstanding of them. Because that basically could not be further from the truth.
Likening a measured home run–or RBI, or what have you– to a metric using a subjectively constructed algorithm to measure +/- ‘x’ runs a player saved/cost his team in a tiny span of games, is absurd. That should be plainly obvious.
The Brewers this year just make me sad. I bet my brother $100 the Crew would be last place in the division this year during Spring Training and he called me out on it and said i was crazy for thinking that. The totality of the problems with the team is ridiculous and easy to predict.
Their starting rotation after the Loshe deal looked horrible outside of Gallardo (who has been bad too). A bunch of minor league pitchers, an aging had a career best year Loshe (horrible signing) and Gallardo will not get it done against the Cardinals and Reds.
Then Hart gets hurt, then Gamel, and we sign…….Betencourt and Gonzalez, two light hitting SS. Who would have guessed that this would end up badly?
I love the Brewers. I’ve been with them for my entire life. But the organization needs an overhaul. They cannot seem to recognize when they need to tread water instead of spending money on overrated players (Suppan, Loshe). They also havent drafted well since Hart/Gallardo/Braun/Fielder. The longer they give up higher draft picks to not compete like the Loshe deal, the longer their non-contention is going to last because as of right now they have a bottom 5 major league team and a bottom 5 minor league system. Where are they going to be able to go from there in the next 5-7 years but down?
I recall having exactly this debate with you six and seven years ago, Dave—and you were the one arguing _for_ stars and scrubs then (I even used that exact phrase), whereas I was pulling for a 25-man-team approach. Well, I don’t know when you came round, but I’m glad to hear it. As you now concur, stars can’t carry a squad, no. There’s no excuse for carrying a bunch of negative WAR guys down the roster, the minors are full of replacement guys who can kick in the necessary smidgen of positive value. Carrying zeros is like giving away outs in a game, with much the same outcome.
It would be interesting to see a study comparing teams with the broadest distribution of positive value—25-man teams—relative to those with the most skewed distribution of value—stars and scrubs—relative to their projected win totals. Regardless of what those projections were, I’d take the hypothesis that teams with the broadest positive distribution are more likely to outperform projections. Notable organizations of that kind once were the Orioles and the Oakland As of the 1970s, although the fact that they both had historicaly strong starting rotations makes the observation harder to generalize. As others mention in this thread, the current San Francisco Giants fit that description better than any other present organization which comes to mind—despite the fact that when they had Buff Barry Bonds it was *them* who were used as the example of a stars and scrubs outfit. Now, they seem more the other type of design.
If a comprehensive study has been done, it would be interesting to have that linked too.