.

Now that we've had our PSA (public service announcement) on Netiquette and the SSI way of doing bidness ... let's return to the (reasonable) question at hand.  Why refer to W-L record, if stats #5, #12, and #18 capture the same things?  Do W-L records have any unique information?

It's an interesting question, taken in a vacuum.  Possibly some SSI readers would like to noodle around in the specifics of the answer...

Sandy and Bat571 weighed in.

.......

For one thing, San-Man said,

It provides the one piece of information that implies "perfection". The ERA, xFIP, WAR, WPA, etc., would not (by their nature) alert a viewer to any level of "perfection" - unless the ERA or xFIP were to actually reach 0.00. (And the irony here is that even with a 0.00 ERA, one can actually still lose. - See more at: http://seattlesportsinsider.com/comment/93200#comment-93200

This relates to our original observation that Tanaka never had a bad game.  It's one thing for a pitcher to have a 2.99 ERA or xFIP; it's another thing for him to have 0% starts in which he allowed more than 4 runs.  ERA doesn't measure consistency.

........

For another thing, Bat571 said,

look at the team's record without that pitcher - call it the Lefty Carlton Gambit. Rakuten was 82-59 for the season and won the pennant for the Pacific League. Without Tanaka, they were 58-59, not figuring Tanaka's 3 ND (which I believe the Eagles still won). Like Lefty, he was a major part of the team's success. W-L has value. W-L pct. has value, particularly if the pitcher made all his scheduled starts during a season. And W-L pct of a pitcher compared with the W-L pct of a team tells you if he's cruising or dominating. - See more at: http://seattlesportsinsider.com/comment/93200#comment-93200

Why keep a +/- record in basketball or in hockey?

..........

For a third thing:

Consider Jeff Sullivan's original article, in which he examined a single Paxton start with 10K and 0 BB.  Sully pointed out that all ML pitchers who had achieved even one such start --- > had a group ERA+ of 107.

Somebody might object, "What is the unique information that one game with 10K, 0BB contains ... that isn't contained in his 7.2 K/9 rate?!"

Well, by that standard, not much.   A game with 10K, 0BB doesn't really contain information that isn't "captured" by his season K rate.

But!  It is one more camera angle by which we establish that a pitcher is excellent.  The group DID have that 107 ERA+.  So what do we say about a 24-0 won-loss record?  Do such pitchers correlate with good ERA's?  

..........

For a fourth thing:

Supposing, for a moment, that W-L did not contain "unique" information.  It would remain a concise way to express excellence. Bill James likes to take large groups of pitcher seasons and compare them across time.  How readable is this line?

• 19-5, 1.27
• 10-4, 1.87
• 24-0, 1.27

So I reject the proposal to censor W-L and ERA out of existence because skill can be captured in other ways.  Legibility is its own reward.

.............

For a fifth thing:

In the original Tanaka article, I outlined six (6) things, expressed by a 24-0 won-loss record, that would not be expressed by xFIP, IP, and WAR.

• Tanaka's consistency.  He went unbeaten because he had no* bad games.
• His adaptability.  His "game" was effective in every possible condition thrown at him.  
• His durability, especially as manifested in the next game after a long start.
• etc.

Of course a distinct stat has unique information in it.  The question is whether we know what it is...

.............

For a sixth thing:

The real live element, in this movement to suppress W-L, ERA, RBI, etc., is snobbery.  Present company excepted!  :- )  The attitude involved, in the censorship of "common" stats, is an elitist one.

.

There are saber blogs with an exaggerated respect for saber erudition, and which dislike the paradigms of those they regard as lower class.

Some sabermetricians, present company excepted!, want to wrest the game away from the riff-raff, and make it their own exclusive property.  We'll thank you to use xFIP in the future, sir, and shun all use of the gauche ERA stat.

Hey, man.  All stats contain information.  At SSI, express the information in any manner that best expresses your own view of the world.

Power to the people,

Dr D