Finding the Toughest and Weakest Divisions

I hope you don’t think this is going to be an extensive study. I hope you don’t want for this to be an extensive study. Because, look, here’s the deal:

divisionalwar2014

And that’s it. Those are the results. You can leave now, enlightened, if you so choose. All that’s coming is explanation, commentary, and history, and some people want that and some people don’t. The AL East projects to be strong! The NL East projects to be strong, too, relative to something other than the rest of this year’s major-league baseball. Relative to that, it’s pretty weak.

I wouldn’t say it’s critical to have an understanding of divisional strength, but at the very least, it’s neat, and more than that, it can help reconcile differences between team strength and playoff odds. After all, if you play in a weak division, your odds will look better than they would in a strong division. I feel a little weird writing when I already gave away the conclusions, but, always move forward, that’s what I say.

The graph above was simple to generate. As you know, we’ve got author-maintained depth charts for every team in the league. As you know, we’ve got 2014 projections that blend both Steamer and ZiPS, and as you might know, this page exists, showing off a detailed team-by-team WAR summary. Making the graph above was almost as easy as just adding numbers together. I did have to make one adjustment, because the linked page projects a few too many WAR. But that did nothing to change the order — it just brought all the numbers down a little bit.

I wound up with the AL East projected for 188 WAR. In second place is the NL West, at 176, so that makes the AL East the toughest projected division in baseball. At the other end, the NL East is projected for 142 WAR. Next-lowest is the NL Central, at 154, making the NL East the weakest projected division in baseball. Both the Nationals and Braves, of course, are pretty good, but the division’s mighty thin behind them, and it stings the Mets extra bad to have to play this season without Matt Harvey. Not that Harvey being healthy would be enough to turn the Mets into certain contenders, but a healthy Harvey is the very image of a difference-maker, and, well I didn’t expect to spend this paragraph daydreaming about the Mets.

Over the past decade, the best division in baseball has averaged 195 WAR, and the worst division in baseball has averaged 142. The projected numbers, naturally, regress toward the mean, so there’s going to be a smaller spread, but there’s a clear gap between the upper tier and the lower. It certainly shouldn’t be a surprise to see the AL East on top, but it’s worth taking a moment to discuss some of the implications.

For example, the Orioles and the Mariners are projected for just about identical WAR totals. Yet we give the Orioles a 12% shot at the playoffs, while the Mariners come in at 41%. The Mariners also have higher playoff odds than the Blue Jays, despite projecting for a lower WAR. The Orioles and Phillies are tied in playoff odds, even though the Phillies project for a significantly lower WAR. While the Marlins project for one more WAR than the Astros, they also project for six more wins. Everybody take a breath, and the next paragraph will contain more examples.

By total projected team WAR, the Braves currently rank 16th, between the Indians and the Pirates. By playoff odds, the Braves rank sixth, between the Red Sox and the Angels. There’s a definite upper class of five teams, by playoff odds, but by WAR, there isn’t that same separation, as the Rangers and Rays are right there with the Nationals. Even though those three teams are equivalent on the WAR page, the Rangers and Rays combine for playoff odds of 86%, while the Nationals are at 77% on their own.

I’m just going to let you eyeball the rest. Obviously, the projections aren’t perfect, but they’re also the best we’ve got right now, and they should convey a pretty good sense of things. It isn’t just about how good a team is. It’s also about the context in which that team exists and competes, and while over the long run these things should average out, in any given year the divisions are going to be uneven. Right now, the NL East looks particularly weak. The AL East looks strong, but then, that’s usually the case.

For funsies, here’s an overview of the last ten years. For the nine years that the AL West had four teams, I multiplied the divisional WAR by 5/4. For the nine years that the NL Central had six teams, I multiplied the divisional WAR by 5/6. For so long, we accepted that one division had four teams while another had six. The more distant that gets, the more hilarious it seems. That’s absurdly unfair! All right.

divisionalwar20042013

The weakest division was the 2005 NL West. That’s the NL West the Padres won with an 82-80 record. The five teams combined for a WAR of 126. The strongest division was the 2008 AL East. That’s the AL East the Rays won with a 97-65 record. The five teams combined for a WAR of 210. The four strongest divisions were all the AL East. For that matter, the AL East was the strongest division in six of the ten years. In three of them, the AL West was the strongest division. In one, it was the AL Central. That means that, over at least ten years, the National League hasn’t had the strongest division in baseball. Nor does it project to have the strongest division in 2014. The gap between the leagues seems to be shrinking, but it has undeniably existed for a while, and that’s evident in the WAR. Especially since WAR includes a league adjustment, on account of the NL having been inferior.

On average, over the decade, the AL East came in at 189 WAR. The NL Central came in at 154. It’s been helpful to have the Red Sox and the Yankees in the same division, so that each could compete with the other. And this makes the accomplishments of the Rays all the more incredible. In time, the Rays will regress, as long as they don’t spend a lot more money. Already, the farm system is drying up, as they haven’t been able to lean on high draft picks. But plenty is already in the books, and lately the Rays have been a little bit miraculous. Which, yeah, sorry, Toronto and Baltimore. A lot of it’s cyclical, but that’s no consolation today.




Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


71 Responses to “Finding the Toughest and Weakest Divisions”

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

    Nice post. I’m bookmarking this for the next time somebody complains the AL Central is too easy.

    Vote -1 Vote +1

  2. PG says:

    How does the DH rule in The AL get factored into this? AL teams generally get positive WAR out of their DH’s while NL teams usually get negative WAR from their pitchers hitting.

    Is it that the difference is immaterial or is it accounted for in some other manner?

    +8 Vote -1 Vote +1

    • Vin B says:

      I’d imagine the pitchers would have a higher FIP projection in AL than NL and thus a lower WAR.

      Vote -1 Vote +1

    • Pirates Hurdles says:

      This was my complaint when Jeff posted this in his chat last week (or the week before). Basically subtract 5-10WAR per AL division due to DH WAR.

      I don’t think NL pitchers have higher WAR than AL because it is league corrected no?

      Vote -1 Vote +1

      • Dan says:

        Does that mean we should also subtract the extra WAR that NL pitchers would receive from facing fellow pitchers rather than a DH?

        Vote -1 Vote +1

        • matt w says:

          But do NL pitchers receive extra WAR for facing pitchers? As Hurdles said, I think WAR is league/park corrected.

          Vote -1 Vote +1

      • pft says:

        AL teams collectively got a total of 23 WAR from the DH, but almost half of that was from AL East teams. So 5 WAR for Central and West, and 10 for AL East seems reasonable.

        However, AL teams won 53% of their games against the NL, which is about 20 wins per division better than NL divisions(over 162 games), so the relative superiority of the AL should be taken into account, and not sure the league adjustments do that beyond adjusting RC and FIP. So maybe the AL divisions should be getting more of a boost even with the DH factor (10-15 wins per division).

        Vote -1 Vote +1

        • piratesbreak500 says:

          But in interleague play, AL teams have the advantage, as the difference between a DH and Pinch Hitter is greater than the difference between an AL pitcher batching and an NL pitcher batting.
          -This is due to the allocation of resources. Both leagues sign pitchers because they’re good pitchers- how well they hit isn’t a factor in the signing decision. However, NL teams have to sign bench players with an emphasis on ability to play positions, whereas AL teams need to sign at least 1 bench player primarily for their bat, leading to the best DH suited players being signed by the AL, and a competitive advantage in interleague games in AL ballparks (and possibly a slight edge in NL ballparks, as the DH is likely a better pinch hitter, but the advantage is tiny).

          Vote -1 Vote +1

    • JZ says:

      NL pitchers don’t have to face DH’s so does that raise their WAR?

      Vote -1 Vote +1

  3. Chickensoup says:

    What goes into the league adjustment? I mean I don’t doubt the AL East is tops because they have 2 teams who routinely buy wins in FA, but how much is also due to the DH spot? If i were to guess, the pitcher gains in the NL don’t overtake the positive from DHing and the difference is the league adjustment. I could be wrong on this though

    Vote -1 Vote +1

    • nv says:

      Huh? Because the Dodgers and Tigers and Angels and Nationals and Giants, etc. don’t sign FA?

      NY’s occasional FA benders are an outlier, sure, but Boston is in the next group. I’m pretty sure the Angels have spent much more on FA over the past few years than Boston — of course, I’m not sure how many wins they’ve bought.

      Vote -1 Vote +1

      • Chickensoup says:

        I was talking more about the decade trend than just the last few years. Not every NY expenditure works out, but many are at least somewhat effective. Boston and NY buy a lot of players on the FA market and are big traders who take on a lot of old debt. Those two teams plus the dodgers do more than most teams do, and they are in the same division. The Dodgers are right there with them now, the giants are different in how they do it (more extensions than anything else). The Angels have given out a lot of money for 2 or 3 years etc.

        I never said other divisions/teams don’t pay for FAs. I said that those 2 teams combined probably do it more than any other division rivalry in baseball, especially over the long haul. For every Pujols contract that a team like the Angels give out there’s an equal contract the Yankees or Sox have given, plus another one. This is in no way saying anything other than those 2 teams spend a lot of money on FAs and extensions, probably more than any other 2 teams in a division does

        Vote -1 Vote +1

  4. Nate says:

    Jeff,

    Would you be able to provide this breakdown with hitters and pitchers only? Then it may be possible to see which teams are going to face the toughest competition throughout the year, and which would see the weakest.

    Vote -1 Vote +1

  5. Scruge says:

    I’d love to see something like this but broken down on a team-by-team basis. After all, the AL East is a lot tougher for the Blue Jays than it is for the Red Sox since teams never play themselves.

    Vote -1 Vote +1

  6. Sean C says:

    The word of the day is ‘funsies’.

    Vote -1 Vote +1

  7. The Stranger says:

    How about rankings of divisional competition for each team? Remove each team from its own division, then figure the WAR of its competition. The Red Sox, for instance, account for 45 of the 188 AL East WAR, so their divisional opponents must have 143. Whereas the Padres account for 32 of 176, so their divisional opponents have 144 – does this mean that the Padres effectively play in a tougher division than the Red Sox? I imagine this goes into the playoff odds already, but it would be a fun leaderboard to see on its own.

    +20 Vote -1 Vote +1

  8. triple_r says:

    An alternative title: “What We Complain About When We Complain About The Orioles Getting Screwed By Baseball”.

    Vote -1 Vote +1

  9. Biesterfield says:

    This might be opening a can of worms, but do you see in the future having division adjusted WAR too? Or even taking it a step further, player adjusted WAR? After all it is harder to get a hit off of David Price than it is Joe Saunders but WAR currently treats them as the same.

    Vote -1 Vote +1

    • tz says:

      Once upon a time, when there were balanced schedules and all, the argument could be made that everyday players faced the same level of competition, so adjustments wouldn’t be worth the effort.

      However, it would make sense to adjust the raw components (hitting and pitching runs above or below) for the level of competition because:

      – Specialists (relievers, platoon players, PHs, etc.) face a different subset than the overall league

      – Differnent team schedules have an impact upon the average pitcher/hitter faced. From BBREF, there is about a 0.5 run per game difference between the toughest (NYY) and weakest (Atl) average opponent played:

      http://www.baseball-reference.com/leagues/MLB/2013-standings.shtml

      This has to make a difference.

      +5 Vote -1 Vote +1

      • The Stranger says:

        Why just use schedules? We have PITCHf/x, and someday soon we might have this new thing MLB is implementing. If we wanted to, we could probably correct for opponents’ talent using the actual opponent faced in each plate appearance instead of using schedules to approximate that.

        I did a little bit of looking into this for a since-abandoned project, and I found that the differences in quality of opposition over a full season weren’t insignificant. Admittedly, that was before adjusting for league and park effects, but there might be something there.

        Vote -1 Vote +1

  10. stan says:

    Its kind of crazy to me to think that anyone could write an article about “best divisions” without any discussion of actual W-L records of the divisions. This article treats WAR as if its is the actual goal of baseball teams and fans rather than wins. In this era where the teams play 20 interleague games each and the leagues (finally) have the same number of teams, there’s no real reason not to use the actual records to determine which was the “best division”. By the way, if the WAR scores are very different from the actual results on the field, that’s a pretty good indication that there’s something wrong with the calculation of WAR.

    For the “record”, the AL East was 28 games over .500 last year. The AL Central was -5. The AL West was -18. The NL East was -14. The NL Central was +16. The NL West was -7.

    -12 Vote -1 Vote +1

    • larry says:

      unfortunetly what you are asking for are the end of year rankings, and those wont be available until the season is over. At this point projected WAR and projected win-loss racords should line up.

      The problem with using last years win-loss records is that 1) teams have changed, some of them significantly. 2) the win-loss records might not actually be indictative of true talent as there is sequencing and luck that has a huge effect

      +11 Vote -1 Vote +1

    • Aaa says:

      Please go ahead and use last season’s W-L records to project 2014 W-L records. I’m sure no one has thought of this before.

      Vote -1 Vote +1

      • stan says:

        So, you guys actually needed me to spell out for you that I was referencing the end of the article and the divisional WAR by year table?

        Vote -1 Vote +1

        • KDL says:

          How do we draw meaningful comparisons from year to year if we use stat A (W-L) for one year, and stat B (WAR) for another year?

          I think Votto is going to hit a tone HR this year…look at this graph of his BABIP over the last 5 seasons!

          My facetious Votto example will be retracted if you admit that WAR and W-L tell us similar info…but you’re fighting against that very idea by complaining about using W-L not being used. So, you can’t have it both ways.

          Either W-L and WAR tell us vastly different things and should not be compared to each other. Or they’re similar enough for you to want to compare WAR and W-L (as in your last comment) – which makes your initial complaint about using WAR pretty silly.

          Vote -1 Vote +1

        • stan says:

          @KDL. How can you determine who is the “best” without looking at actual wins? Shouldn’t that be the primary if not the only determinant of who is the “best”? I can see why you’d use WAR or another predictive tool to guess at who is best in the coming year, but I don’t see any logic to using WAR from past years to determine which divisions were the best when we have the actual win loss records to go by.
          You fail to recognize the distinction between predictive analysis and retroactive analysis, which is every bit as different as WAR and Wins and Losses. As I’ve said twice now, I was talking about the end of the article.

          Doesn’t it bother anyone that the actual W-L records of the divisions are so different than the designations of total WAR? To me that screams that WAR is not a very accurate tool of past worth.

          Vote -1 Vote +1

        • Xao says:

          Except that wins won’t tell you which team is better either. If one team wins 89 games and another wins 90, how confident are you that the 90-win team is actually better?

          Vote -1 Vote +1

        • Win-MIN-Twins says:

          Here you are, the divisions by wins and average winning percentage:

          ————————
          AL East
          Wins – 433
          Average Winning % – .534

          AL Central
          Wins – 400
          Average Winning % – .494

          AL West
          Wins – 387
          Average Winning % – .477

          NL East
          Wins – 391
          Average Winning % – .483

          NL Central
          Wins – 421
          Average Winning % – .520

          NL West
          Wins – 399
          Average Winning % – .493
          ————————

          (Note that the AL East and AL West played one more game than the other divisions because of the Rays-Rangers play-in to the play-in.)

          The AL West has the lowest win total and winning percentage. The obvious thing to attribute this to is the presence of the Astros. The problem is, the top two teams (Oakland and Texas) were obviously both very, very good.

          The NL West is, by this approximation, the third-worst division when its worst club is Colorado (who won 74 games). With the exception of the AL East, every other division’s fifth-place team (and some fourth-place teams) are below this. Then there’s the Dodgers, who basically everyone was high on going into the postseason. They won 92 games, four less than the Braves and five less than the Cardinals and Red Sox. It feels like wins and winning percentage are more indicative of parity within the division than anything, at least in the case of the NL West.

          That’s what I’ve got. (I personally prefer WAR.)

          Vote -1 Vote +1

  11. Roger says:

    Are the number of WAR generated in intra-division contests consistent and we are mostly seeing a difference in WAR based on strength in interdivision and interleague play, or does the WAR difference indicate that the events counted by WAR happen more frequently in intradivision play in the stronger divisions?

    What happens to a player’s WAR as he changes divisions and leagues? Can a hitter moving from the AL East to the NL East be expected to face a WAR drop greater than would be expected by aging alone, or would you expect a bump up in WAR?

    Vote -1 Vote +1

  12. Aaa says:

    Might be helpful if you could see each division broken down by team. That way you could see that non-Tigers AL Central teams make up only ~72% of the projected WAR of that division, non-Braves teams ~75.5%, etc.

    Vote -1 Vote +1

  13. Jay29 says:

    I was taught by professors at a very good engineering school that it’s improper to put a non-zero minimum on the y-axis of bar graphs. Yet I’ve heard from smart people since I was in college that that’s a frumpy old rule that shouldn’t be held onto.

    What do you think? Is it deceptive to imply that the AL East is almost twice as good as the NL East? Or is it too boring to show the whole bar and diminish the difference between the bars?

    Vote -1 Vote +1

    • Biesterfield says:

      Its fine. When you introduce a new graph and want to compare them that is when you want to make sure you have the same axes.

      Vote -1 Vote +1

  14. WormFace says:

    “For so long, we accepted that one division had four teams while another had six. The more distant that gets, the more hilarious it seems. That’s absurdly unfair! All right.”

    Please explain this to me. I don’t understand. I tried to google it but couldn’t find an answer.

    Vote -1 Vote +1

  15. here goes nothing says:

    So, a comment that I think is true: this isn’t the whole story (OK, duh). What I’m thinking is that WAR doesn’t capture level of play; it could be that in 2014, the level of play is literally higher than in 2005. If true talent level rises across the board, you won’t notice it in WAR, because the value of a replacement level player would rise correspondingly.

    Not that I think this is necessarily true, but it’s a valid methodological point.

    Vote -1 Vote +1

  16. gtb says:

    The weakness of WAR?

    Using a slightly more prosaic method (and arguably significantly more reliable)of determining divisional strength; i.e. the number of ACTUAL wins, yields the following rankings for 2013.

    Division Record WAR
    AL East 433 1
    NL Central 421 3
    AL Central 400 2
    NL West 399 4
    NL East 391 6
    AL West 387 5

    By this method the strongest division is still the AL East albeit by a much narrower margin that WAR would suggest. Coupling this with the inherent differences between projected WARs and actual WARs leads me to trust these projections very little.

    Vote -1 Vote +1

    • Sean says:

      Congrats, you’ve successfully analyzed the past using a new method that no one has ever heard of!

      Vote -1 Vote +1

    • Jovins says:

      That’s not a valid argument. Teams play far more games against teams in their divisions than out (and also against teams in their league) which will tend to benefit teams in bad divisions and punish teams in good divisions.

      It’s going to artificially bring the divisions closer together, and it’s also going to make teams in the weaker league seem strong.

      Vote -1 Vote +1

      • paperlions says:

        I feel the need here to point out that every division has a .500 record against itself on account of that whole, every game has a winner and loser thing….so, when comparing the number of wins in entire divisions, it works just fine.

        Regardless of the fact that that was last year…it is odd to me that a division that has made little to no improvements of the rosters since last year, the NL West, is projected to be so good when last year the division had one winning team.

        Vote -1 Vote +1

        • RC2 says:

          No, it doesn’t work fine.

          Let me give you a silly example: The NL West gets hit with a meteor, and everyone on all the teams and their minor leagues die. They hire the landscaping crew to play the games.

          The NL West, of course goes .500 against itself, and loses the rest of its games. Each team goes 38-38 in division, and loses the rest of its 86 games, to go a sparkling 38-124.

          That 38-124 record DRASTICALLY overrates the talent of these teams.

          Vote -1 Vote +1

  17. Vlad the Impaler says:

    Fangraphs is running out of ideas for articles. They really need the regular season to start.

    Vote -1 Vote +1

  18. Llewdor says:

    Toronto – 2008. Third best team in baseball. Finished fourth in their division.

    Ouch.

    Vote -1 Vote +1

  19. JunkyMonkey says:

    I made this to show what it looks like with each team split out.

    In case I didn’t do that correctly, here is a direct link
    http://i.imgur.com/RAhq2se.png

    Vote -1 Vote +1

  20. Ruki Motomiya says:

    “After all, if you play in a weak division, your odds will look better than they would in a strong division.”

    Isn’t what matters the record differential and not overall record?

    For example, a theoritical division with all teams projected for the same WAR, but an average WAR, is harder to win than a division with two really good teams and two more average teams, because Division 1 has 4 people in close compeition while Division 2 basically only has 2 competitors? So it is harder for any one team to win division 1.

    Vote -1 Vote +1

  21. Aron says:

    Is there any way to switch a team or two from the NL East to the AL East, you know to even things out, not because I’m tired of the Jays getting beat on or anything, just in the name of fairness…

    Vote -1 Vote +1

  22. bg says:

    If it weren’t for Jeff Sullivan, I think I would constantly be in a state of not knowing whether the worst MLB team was actually STILL A BETTER BASEBALL TEAM THAN THE BEST LITTLE LEAGUE TEAM.

    THANK YOU JEFF SULLIVAN

    Vote -1 Vote +1

  23. Antonio Bananas says:

    If there are 10 people, one is a millionaire, the others make 0, they have a high average, but that’s not a rich group of people. The AL central isn’t tougher than the NL East. Maybe have a playoff odds thing and list each league and score it like a cross country race (1 point for first, 2 for second, etc) and see which division has the most amount of teams likely to make the playoffs.

    Vote -1 Vote +1

  24. Or at least teams above .500. Maybe I’m bias though.

    Vote -1 Vote +1

    • Baseball Splits (Twitter) says:

      The Indians were pretty good last year. Other than the Braves, the NL East was not very good last year.

      Vote -1 Vote +1

  25. rpb says:

    It is always interesting to look at the details behind the projections. The 2014 projections show the Yankees with the 3rd best starting pitching staff in baseball, 2.0 WAR better than the Nationals. Tanaka sports a 4.7 WAR for his first season. He is apparently going to be almost 1 win better than Strasburg. CC improves 1.1 WAR over last year even though he is another year older. Those seem like rather optimistic assumptions.

    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>