clock menu more-arrow no yes mobile

Filed under:

A Sarcastic Reminder Why the Elias Free Agent Rankings Are Dumb

It's tough to detect sarcasm on the inter-tubes sometimes, so I'll make it clear when I'm being sarcastic below.  Look for the "Sarcasm!" warnings (not to be confused with the "Ichiro!" warnings.)  What is clear is that the Elias player rankings, which determine Type A and Type B free agents, are more than a little out of whack.  Here's what I mean...

Position Groups

Type A/B free agents are the top 20%/40% of free agents in each of these five positional categories:

  • Group 1: first basemen, outfielders, and designated hitters
  • Group 2: second basemen, third basemen, and shortstops
  • Group 3: catchers
  • Group 4: starters
  • Group 5: relievers

Looking at Group 1, you've got center fielders lumped in with designated hitters.  Yes, that makes sense, and those two positions have pretty similar defensive value.  (Sarcasm!  Center fielders are actually similar in value to second and third basemen, twenty runs more valuable than DHs.  First basemen are fifteen runs worse and corner outfielders are ten runs worse.  I don't see any team signing Jim Thome to play center field this off-season, do you?)

Looking at Group 5, we see that the top relievers are deemed just as valuable as top starters and top position players.  Naturally, that makes sense.  (Sarcasm!  Even a bullpen ace like Mariano Rivera is "only" a 4 WAR guy.  Eleven startering pitchers have hit that number already this season, with four above 5 WAR and two above 6 WAR.

Rating Metrics

Each group of players is rated on a similar set of numbers:


Perfect.  I agree 100%.  (Sarcasm!  I mean, seriously?  At least we know a Triple Crown winner will be a Type A free agent.  Whew!  (Oops, that's more Sarcasm!))

  • 2B/3B/SS: PA, AVG, OBP, HR, RBI, Fielding percentage, Total chances at designated position

What?  No stolen bases!?  I thought the middle infield was a place for speed, not power?  (Sarcasm!  At least an attempt is made to measure defense, even if the stats are nearly worthless.)

  • CA: PA, AVG, OBP, HR, RBI, Fielding percentage, Assists

Finally, some good defensive metrics.  (Sarcasm!  Really, this is just silly.  Having figure skating judges rate the aesthetics of each catcher's helmet wouldn't be a less valuable category.  Note that there's no sarcasm there.  Would including stolen base percentage or passed balls really be that overwhelming?)

  • SP: Total games (total starts + 0.5 * total relief appearances), IP, Wins, W-L Percentage, ERA, Strikeouts

Good, in order to balance out the rationality of ERA and strikeouts we needed to double up on the playing time metrics and include both Winz and W-L record.  (Sarcasm! That's actually how you get Gavin Floyd on the A list and John Danks on the B list.)

  • RP: Total games (total relief appearances + 2 * total starts), IP (weighted slightly less than other categories), Wins + Saves, IP/H ratio, K/BB, ERA

No sarcasm this time, I'm done pretending to be witty.  Why is IP weighted slightly less than other categories?  I mean, if you're suddenly going to be smart and weight the categories, why just one and why this one?

The Silver Lining

While the algorithm used in these rankings is ridiculous, the good news is that the rankings end up being moderately decent.  Over the course of two seasons, the better pitchers do tend to win more games and the better hitters do tend to have more RBI.  It's actually pretty hard to create a metric that's a hodge podge of random stats and that completely misses the boat.  But we can definitely do a lot better at creating a list of the best free agents -- instead of getting the tiers 70% right, 90% right seems like a better goal.  And it wouldn't be too hard to get there if, you know, mainstream baseball coverage wasn't in love with RBIs and fielding percentage.

All information from this MLB Trade Rumors post which got its information from Keith Law.