THE BOOK--Playing The Percentages In Baseball

I can’t submit a comment over at the BBRef site so I’ll do it here:

May I suggest, to find the beginning of the LOOGY trend, looking at Tommy Lasorda’s use of John Candelaria in 1991? In that season, 24 of Candelaria’s 59 appearances (41 percent) were for one batter. IIRC, this was Lasorda’s first use of a LOOGY, and the timing seems to fit nicely with Andy’s findings.

http://www.baseball-reference.com/players/gl.cgi?id=candejo01&t=p&year=1991

Andy is here now, so feel free to chat him up:

http://highheatstats.blogspot.com/

Good catch by GSfRB. The same suggestion was made contemporaneously. The STATS 1993 Baseball Scoreboard had an article entitled WHERE WILL THE TREND TOWARD “SPECIALIST RELIEVERS” END? and they also identified Lasorda’s use of Candelaria as the catalyst for what happened in 1992.

Their definition of specialist was a reliever averaging less than one inning per game (50g minimum), not limited to one batter appearances specifically. There was once such player-season in 1967, an average of one a year from 1980-89, then 3 in 1990 and 4 in 1991, jumping to 22 in 1992. But Candelaria’s 1991 was a real outlier, as he averaged 0.57 innings per game. No one else through that point had ever averaged less than 0.84 innings/game over a season.

One thing which might be fun to investigate is whether lineup construction trends had anything to do with this. My first thought was that more lefty specialists were driven by teams batting more lefties, but in that case sequences of back to back lefties, or two out of 3, would be more common, and presumably make appearances with one batter faced less common. So specialization, in particular 1 batter specialization, might be a response to less bunching of lefties in lineups.

The increase in left specialists in MLB during those years might well have been driven by sabermetrics, since platoon splits were quantified and publicly available starting in the mid 80s, and a good deal of attention was paid to whether strong or weak platoon splits were actually persistent.

FWIW, there’s also been an increase in switch hitters.  I posted the data about 3-4 years ago on this blog.  So, there might also have been a counter to the specialist role, or perhaps in response to all the platooning, that switch hitters started to become more prominent. 

It’s fascinating how there’s this shift historically.

Increase relative to when? It’s higher than the 1920s, sure, but I thought the 1990s had a slight decrease due to the end of the slapping swift switch hitter that filled the leagues (especially the NL, which never made sense to me) in the 1980s.

Charlie: I can’t find my blog post, which is really strange.

I know I posted a Google Docs of my data, so I’ll have to wait until I get home to check that.

I seem to remember switch-hitting plate appearances reaching 15% of all PA.  I don’t remember when it plateaued to that point.

I thought the graph really told the story well.  I’d guess you posted this due to the conversation on bjonline.

-- Detroit Michael (MWeddell)

When I saw the title of the post, my first name was “Rick Honeycutt”.

I just looked him up to see if the stats matched my memory, and starting in 1990, he had more appearances than IP ... meaning he came into games to pitch less than an inning ... hence “The First LOOGY”. He must be proud.

I actually recall some broadcaster (or maybe even LaRussa) saying the he deserved some HoF consideration for being the best LOOGY in history.