The 2013 NL East, and Strong and Weak Divisions

Coming into the 2013 season, The National League East was supposed to be a competitive division. The Nationals won 98 games last year, the Braves won 94 and added two Upton brothers to their outfield, and while the Phillies had disappointed in 2012, it was possible to hope that a bounceback year from Roy Halladay would anchor a rotation that could hang with anybody. Instead, the Braves are 57-44 (a 91-win pace) and every other team in the division is under .500. As a matter of fact, it is extremely historically rare for so many teams in a division to finish the season below .500.

For the purposes of this article, I am defining a “Weak” division as one in which 75% of teams finish below .500. I am defining a “Semi-Weak” division as one in which 60% of teams finish below .500. By the same token, a “Strong” division is one in which 75% of teams finish at or above .500, and Semi-Strong is one in which 60% of teams finish at or above .500.

(For a Weak Division, since 1969, every team but one would have to finish below .500. In previous years, when divisions had eight or ten teams, that meant at least six out of eight, or eight out of ten, would finish below .500. A Semi-Weak division is one in which at least three of five, four of six, five of seven, five of eight, or seven of ten finish below .500. The same proportions hold for Strong and Semi-Strong.)

There have been a lot more weak divisions lately than there used to be, as you might imagine. There are more divisions, and unbalanced schedules widen the win-loss gap between the haves and the have-nots. In the first 19 years of the three-division era — 1994-2012 — there were 114 division-years (19 * 3 divisions * 2 leagues), yielding 11 Weak Divisions, and 46 Semi-Weak Divisions. That includes the shortened seasons in 1994 and 1995. In the 17 162-game seasons from 1996-2012, there were just seven Weak Divisions and 38 Semi-Weak divisions.

Interestingly, in absolute terms, there have been a lot more Strong divisions than Weak divisions.

  Strong Semi-Strong Semi-Weak Weak
1994-2013 16 56 46 11
1969-1993 5 35 20 2
1901-1968 4 54 22 0

Why is that? Maybe it’s because of competition: almost all teams try to win, so strong divisions occur when those teams succeed in their goals, while weak divisions only occur when nearly every team gets snakebit. I did say “almost”; one of the reasons that the NL East is a Weak division this year is that the Marlins play in it.

Maybe the weakest division in recent memory is the 2005 NL West, which the San Diego Padres won with an 82-80 record, finishing five games ahead of the second-place 77-85 Diamondbacks. (The following year, the Cardinals won their division with an 83-78 record, but it was a much closer pennant race, as the Astros finished 82-80, so it was only a Semi-Weak division.)

Conventionally, the AL East is often thought of as the strongest division in baseball, and it was strong in 2008, 2010, and 2011, which is remarkable. But the AL West has actually been more competitively difficult by this analysis: it was a strong division in 1995, 2000, 2002, 2004, 2009, and 2012.

Here’s the complete list:

Strong Weak
2012 AL West 2011 AL Central
2011 AL East 2008 AL West
2010 AL East 2005 NL West
2009 AL West 1999 AL Central
2008 AL East 1998 AL Central
2008 NL Central 1997 NL Central
2007 NL West 1997 AL Central
2005 NL East 1995 NL East
2004 AL West 1995 AL Central
2003 NL East 1994 NL West
2002 AL West 1994 AL West
2000 AL West 1984 NL West
2000 NL West 1983 AL West
1997 NL West  
1996 NL West  
1995 AL West  
1991 AL West  
1988 NL West  
1981 AL East  
1979 AL East  
1969 NL West  
1928 NL  
1926 AL  
1916 AL  
1911 AL  

(Data obtained from

The relatively high number of Strong divisions in recent years provides a good argument in favor of expanding the Wild Card. Strong divisions make for compelling pennant races, but they also highlight the unfair nature of the game: it’s hard for the Rays to compete with the Red Sox, Yankees, and Blue Jays, all of whom have spent hundreds of millions of dollars in the last few offseasons, as they all compete for the same flag.

On the other hand, Weak divisions really are embarrassing for the game. There were no weak divisions before the 1980s, so they are a relatively recent phenomenon. One reason for weak divisions could be non-competitive owners. But another could be poverty. The American League Central is one of the weakest divisions in baseball this year; it often is. One reason is simple economics, as wealth in America is concentrated on the coasts, not in the Midwest.

Another is that it doesn’t take much to win the division, so there is little upward pressure forcing all of the teams to innovate or die. Just to illustrate that point, the Cleveland Indians won three straight division titles from 1997 to 1999, and over that entire timespan, they were the only team from their division to finish over .500.

Next year, the NL East will probably be reasonably competitive again. By average age, the Braves and Nationals are two of the youngest teams in the league, so they’re likely to continue scrapping no matter what happens to the Phillies, Mets, and Marlins. So this analysis has little predictive use. But it’s interesting to think about.

Print This Post

Alex is a writer for FanGraphs and The Hardball Times, and is a product manager for The Washington Post. Follow him on Twitter @alexremington.

35 Responses to “The 2013 NL East, and Strong and Weak Divisions”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. 2nd Edition says:

    I don’t like your cutoff points – I think it belies your thesis. For example, this year’s NLC is only semi-strong because it has only 3 teams, 60%, over .500. But all 3 winning teams are more than 10 games over. To me this has to be a strong division if the word is to have any meaning.

    +14 Vote -1 Vote +1

    • Table says:


      Vote -1 Vote +1

    • I get what you’re saying, and I think that would be a useful measure, too. Basically, I’m measuring strength or weakness as a negative trait: a “strong” division is one that has very few doormats, and a “weak” division is one that has very few juggernauts.

      I think you’re asking for a measure of relative team strength — how good the individual teams are in relation to one another. That could also be interesting. But I first just wanted to measure the divisions as a whole.

      Vote -1 Vote +1

      • Baltar says:

        Sorry to disagree with you, Alex, but to me, the strength/weakness of a division is best measured simply by it’s win/loss percentage vs. other divisions. (A second-order analysis might bring some additional precision.)
        I don’t think having more or fewer strong or weak teams, especially when the criteria include games against the teams in the same division, means anything.

        Vote -1 Vote +1

        • That’s a completely different measure of strength. I’m not interested in whether the AL East teams en masse are better than the AL Central teams en masse. What I’m interested in is the experience of a single team playing in the AL Central, looking at all of its divisionmates. Are there any doormats? Are there any juggernauts? How many teams would that team have to leapfrog to win the pennant?

          Vote -1 Vote +1

        • Bip says:

          It doesn’t really matter how many teams you have to leapfrog, you just have to be better than the best team. The NL West is probably better than the NL East on the whole. Leapfrogging the Mets and Marlins is probably easier than leapfrogging the Padres and Rockies, but the NL East has the Braves, who are probably better than any NL West team.

          So the Dodgers are in first in a semi-weak division when the probably would be in second in the weak NL East.

          Vote -1 Vote +1

        • Doug says:

          Not to keep pressing Alex, but your measure doesn’t capture whether there are doormats or juggernauts. It measures whether the division has teams that had records above or below the mean for the league. If anything I’d imagine your “strong divisions” are more likely to correlate to divisions with “doormats” than “juggernauts” as to find a lot of winning teams they need to be getting those wins off of someone.

          Vote -1 Vote +1

      • Jason B says:

        I was essentially going to make the same point as the 2nd Edition but knew someone would beat me to it. I would judge the relative strength of the division based on aggregate winning percentage, not % of teams above or below .500. Which of these divisions is really ‘weak’?

        Division the first
        Springfield 71-29
        Shelbyville 69-31
        Ogdenville 48-52
        Brockway 47-53
        N. Haverbrook 45-55

        Division the second
        Croatia 55-45
        Slovenia 54-46
        Macedonia 52-48
        Montenegro 52-48
        Serbia 22-78

        Two .500+ teams in the first but an overall win% of .560. Four .500+ teams in the second but an overall win% of .470.

        Not surprisingly Jeff Loria owns the Serbian team.

        +21 Vote -1 Vote +1

        • Bryan says:

          Three more wins would have put Ogdenville on the map! If only Lyle Langley hadn’t quit on the team when he skipped town in June.

          Vote -1 Vote +1

  2. Doug says:

    Did you take into account the different sizes of divisions? It’s a lot easier for a 4 team division to have 75% of it’s teams with winning (or losing) records than for a 5 team division to accomplish the same. This certainly contributes to the AL West accomplishing this feat more regularly than the more celebrated AL East.

    +15 Vote -1 Vote +1

    • I applied the 75% threshold to all divisions regardless of team size, as I wrote:

      For a Weak Division, since 1969, every team but one would have to finish below .500. In previous years, when divisions had eight or ten teams, that meant at least six out of eight, or eight out of ten, would finish below .500. A Semi-Weak division is one in which at least three of five, four of six, five of seven, five of eight, or seven of ten finish below .500. The same proportions hold for Strong and Semi-Strong.

      Vote -1 Vote +1

      • Baltar says:

        I think you missed Doug’s good point, as your reply wasn’t responsive.

        Vote -1 Vote +1

        • I didn’t miss it, exactly — Antonio Bananas made a similar point below. All I could do is clarify my methodology. I applied the 75% threshold equally across all divisions regardless of the number of teams. Obviously, that meant that it was more likely that the AL West would be “strong” or “weak,” and less likely that the NL Central would be. That’s a flaw in my methodology, if you like. But that’s what I did.

          Vote -1 Vote +1

  3. Scoops Häagen-Dasz says:

    Wasn’t Houston leaving the NLC supposed to expose the division for how terrible it really was? And all those teams that feasted off the terrible Stros would take a big step back cuz they didn’t get to beat up on the weakest team in a 6-team division?

    Vote -1 Vote +1

    • cass says:

      Yep. Except the Cardinals are always good, the Reds have had a good team for a few years, and the Pirates have been smartly rebuilding for a number of years and it finally paid off. Basically, the fact that a monster team built on prospects and smart acquisitions developed in the division made up for the fact that the Astros left.

      Vote -1 Vote +1

  4. cass says:

    How often have all teams in a division been at or over .500? I remember back in 2005, the last-place team in the NL East, the brand-new Nationals, had a .500 record.

    I also don’t see the point in looking at the leagues before divisions were implemented. Back in those days, there was no interleague play, so the wins and losses would have to even out to 0 within each league. That’s never been the case in the divisions because teams have always played teams outside their own division. The NL East is weak this year because NL East teams keep getting beat by NL Central and NL West teams.

    Would a better measurement of division strength be the total combined winning percentage of all teams in the division? Seems like that’d be more accurate.

    Vote -1 Vote +1

    • It has happened exactly twice in history: the 2005 NL East, when the last-place Nats finished 81-81, and the 1991 AL West, when the last-place California Angels finished 81-81 (and the worst-to-first Twins won 95 games and eventually the World Series).

      Vote -1 Vote +1

  5. Table says:

    I don’t think it makes sense to simply use whether a team is over .500 as the defining factor. How far over or under a team is would provide better data

    Vote -1 Vote +1

    • I have all of the data. What specific query did you have in mind?

      Vote -1 Vote +1

      • Pirates Hurdles says:

        Why not just base it on win percentages and define the cutoffs by examining the data distribution? That eliminates the obvious caveat of differing N when using your percentage cutoffs.

        Vote -1 Vote +1

        • Because, especially in the era of unbalanced schedules, it’s easier to have really good teams when there are really bad teams. It’s a lot harder to be a 100-win team when there are no 90-loss teams in your division.

          Vote -1 Vote +1

        • Table says:

          It just doesn’t make sense to use a .500 cutoff. is a 49-51 team really different than a 51-49 team? Especially once you consider that the 49-51 team might play in a stronger division?

          Vote -1 Vote +1

        • Eric Feczko says:


          Using aggregate win-loss percentages controls for the unbalanced schedule. The unbalanced schedule reflects the fact that there are more intra-division than inter-division games. However, the aggregate (or mean) percentage for intra-division games will always be .500. As a result, the only difference in aggregate or mean W-L percentage results from inter-divisional games, thereby measuring the relative strengths of teams (if you buy into W-L percentage as a measure of “strength”, which is debatable for the small sample of a season’s worth of games).

          In fact, your arbitrary cutoffs introduce the confound of unbalanced schedule, because they do not control for intra-divisional games; a bad team and four average teams may become a strong division, while a five average teams could become a “weak” division.

          Vote -1 Vote +1

      • dougiejays says:

        Why not do it by aggregate games over/under .500 or aggregate win %? The AL East this year has 4 teams over, but the Yanks and Jays are basically both .500 teams if you disregard their head-to-head record.

        Vote -1 Vote +1

  6. Antonio Bananas says:

    Maybe instead stack rank a league, then use a points system like in team races. Lower scores are better.

    I think division competitiveness would be interesting too. Maybe use a Gini coefficient with wins as what you measure and plot.

    Vote -1 Vote +1

  7. Dave says:

    Actually, by your own analysis the AL West is weaker than the AL East because while the AL West has three seasons of being a weak division(1983, 1994, and 2008) the AL East has zero. Also many times divisions are remembered for World Series Championships not in-season achievements, which makes people more likely to think of the AL East as stronger than the AL West, due to the East having more championships.

    Vote -1 Vote +1

    • Antonio Bananas says:

      Because the West had 4 teams. So they had to have less teams qualify for the arbitrary 75%. If you bump that to 80%, no year from e AL West is special.

      Vote -1 Vote +1

  8. Antonio Bananas says:

    Also, maybe the AL west was disproportionate because it only had 4 teams. So less teams had to be good to reach your 75% threshold. For a 5 team division, really, 80% would have to be over .500 and 4 teams overly, which is 33% more teams than the 3 teams and 75% required for the AL West.

    Vote -1 Vote +1

  9. Sylvan says:

    The 1973 NL East is technically not a “weak” division” according to your methodology, but, man…

    NYM 82-79
    STL 81-81
    PIT 80-82
    MON 79-83
    CHI 77-84
    PHI 71-91

    Vote -1 Vote +1

    • You’re right. Technically, it’s semi-weak.

      As I said above, I agree that this measure could be usefully augmented by a measure of the strength of the individual teams in the division. But I’m using the word “strength” somewhat differently, which is why I capitalized it and put it in quotation marks. I mean “strong” and “weak” to refer to the number of teams in that division who are competitive.

      Vote -1 Vote +1

  10. R says:

    I’m going to say this article is semi-weak

    Vote -1 Vote +1

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>