(The data which this post discusses can be found here. Feel free to open it now, or wait until you get further down!)
As anyone who knows me in real life knows, I am an enormous Chicago Cubs fan. As that criteria eliminates (probably) every single one of you, I will continue without referencing this fact again. Save once. Right now.
Today, I was pondering who my would be on my All-Time Cubs team. I wanted folks who'd contributed while wearing a Cubs uniform - while Andre Dawson may have won the '87 MVP with Chicago, there were outfielders who contributed more to the Cubs during their time in pinstripes (Hack Wilson, Sammy Sosa, and Billy Williams, as it turned out, though Dawson was on the Montreal/Washington list). Similarly, Greg Maddux is clearly one of the greatest pitchers who ever played for the Cubs, but his contributions in Chicago were actually somewhat less than a handful of other pitchers who called Wrigley home (Jenkins, Reuschel, Rush, Brown, and Vaughn).
With these thoughts in mind, I went to Fangraph's handy-dandy leaderboard and started the process of creating my team.
I enjoyed the results I got, so I went ahead did the same thing for every franchise currently in the majors. You can find the results here. (Spoiler: It's good news for Yankees, Tigers, Giants and Braves fans.)
A few notes on process, before we get to the more interesting stuff:
(1) Leaders are Fangraphs WAR leaders, at their positions, during their time with any given franchise. Players were allowed to be on more than one franchise (Alex Rodriguez for the Yankees and Mariners, for example, or Nolan Ryan for the Angels, Rangers, and Astros), but not twice for one franchise. As for that ...
(2) If a player played more than one position for a franchise, as was the case with a number of folks, they were placed at the position where they had the greatest WAR differential with the second-best player at that position. For example, Stan Musial played both 1B and LF for the Cardinals, producing 126.8 WAR in the process. The second-best Cardinals LF, Lou Brock, produced 41.5 WAR. The second-best Cardinals 1B, Albert Pujols, produced 83.0 WAR. So for the purposes of this exercise, Musial was considered a Left-Fielder, allowing Pujols (who was less-worse than Brock) to play first;
(3) Outfielders were separated by position (LF/CF/RF), and had the principles described above applied;
(5) In order to avoid penalizing franchises who had multiple superb players at the same position, I added (in addition to five pitchers and a position player at each position) a "bench" player, who was simply the player with the highest WAR for that franchise who would not have otherwise been included. For example, Carl Yastrzemski was "crowded out" by Ted Williams in Boston's LF. He was included here;
(6) Only data from 1901 - 2013 was considered. Sorry, early National League players You take the hit on this one.
The teams themselves were quite interesting, and I encourage you to play around with the chart to find teams you like. But wait, there's more! Once I'd found the results (which again, can be found here) I played around with it a bit, adding STDEVs and Z-SCOREs for each "category" of WAR (pitching, infield, outfield, and bench). This allowed me to get a sense of which teams have historically been heavily reliant on their pitching (Diamondbacks, Mets, Blue Jays, and Twins), their infields (Brewers, Orioles, Rangers, and Marlins), their outfields (Giants, Rays, Tigers, and Padres), and that made up category, "bench" (Red Sox, Reds, Braves, and Padres). Play around with the numbers a bit; you'll see what I'm talking about. It was a lot of fun.
There are of course many things I could also have done, and I'm looking forward to hearing from all of you about what some of those things are. I'm planning on running some R-squareds between various categories, to see if there's any correlation between having a relatively "balanced" attack (evenly distributed between offense and pitching) and overall WAR success. I'd also like to find a way to normalize the results so teams which haven't been around for very long (Marlins, Rays, etc.) aren't penalized for not having had many franchise players.
All in all, though, it was an interesting exercise. Take a look, and let me know what you think, and interesting things you notice!
tWAR: Total WAR produced by all players considered for each franchise.
pWAR: Total WAR produced by the five pitchers considered for each franchise.
iWAR: Total WAR produced by the starters at the infield positions for each franchise.
oWAR: Total WAR produced by the starters at the outfield positions for each franchise.
bWAR: Total WAR produced by the player that produced the most WAR for each franchise, who wasn't included as the representative for any of the other positions.
(z)-tWAR: The z-score for each franchises' value in tWAR.
pWAR%: The percentage of each franchises' tWAR that came from pWAR (pWAR% = pWAR / tWAR).
(z)-pWAR%: The z-score for each franchises' value in pWAR%.
iWAR%: The percentage of each franchises' tWAR that came from iWAR (iWAR% = iWAR / tWAR).
(z)-iWAR%: The z-score for each franchises' value in iWAR%.
oWAR%: The percentage of each franchises' tWAR that came from oWAR (oWAR% = oWAR / tWAR).
(z)-oWAR%: The z-score for each franchises' value in oWAR%.
bWAR%: The percentage of each franchises' tWAR that came from bWAR (bWAR% = bWAR / tWAR).
(z)-bWAR%: The z-score for each franchises' value in bWAR%.