As teams prepare for the season opener, the final cuts, some of which are very difficult, are being made. It has gotten even more difficult in the last few years as teams now regularly break camp with a twelve-man staff. When I was a kid in the Seventies, ten-man staffs were de rigueur and even some of those guys seemed never to get used (the Phils carried Ron Schueler for years but seemed to use him less than a September callup). 

The other day in Philly, Karim Garcia, a non-roster invitee, become ostensibly the final cut of spring when the Phils declined to sign him to a major-league contract. Garcial, who hadn't played in the majors since 2004 (when he batted .229) was a dubious pickup in the offseason to be sure, but he performed as well as almost anyone in camp (.305 batting average, 7 runs, 1 HR, and 7 RBI in 59 at-bats).

Compare that to the man he was basically competing with for the fourth outfielder/pinch-hitter job, Jayson Werth, who batted .275 with 7 runs, 1 HR, and 6 RBI in 40 at-bats. However, Werth, who can play all three outfield spots more capably and has even learned how to catch in a Luis Aguayo/emergency type role. So Werth was more versatile—and had been out of the majors since just the middle of 2005—and won the job. The fifth outfield spot went to Michael Bourn, who is needed for defensive insurance for Pat Burrell in late innings and for his speed (6 stolen bases and 15 runs to lead the club this spring). So bye bye, Karim. Oh well, he won't be missed much.

Back in the Seventies, I remember the Phils carrying multiple corner outfield/pinch-hitter types. In 1980, they had three: Del Unser, Greg Gross, and George Vukovich. But teams carried six outfielders, three catchers, two utility infielders, and a partridge in a pear tree back then.

It makes me wonder if we are ready to expand the rosters to 26 men. It gets worse in the American League where teams must carry a designated hitter as well. Remember when the NL voluntarily limited their rosters to 24 men supposedly out of altruism for their brothers in the AL, though I'm sure cutting one salary didn't hurt either.

Anyway, it also makes me wonder what kind of men are filling those final rosters spots as bench players are expected to do more and more. First, I ran the numbers comparing bench players against starting position players. For each, I calculated the percentage of players, Win Shares, games, at-bats, and plate appearances for starters as compared to all position players:

So even though bench players numbers have thinned somewhat in the past decade, their production (measured by Win Shares) has fallen off more. However, that's due more to their lack of use (games, at-bats, and plate appearances) than to their poor performance. Actually, their disuse starting in the Nineties, at Nineteenth Century levels, and is now off the charts. There has never been an era in baseball in which the bench was used less. 

So who are these bench players that we've heard so much about? Well, one thing about them is that they are expected to play more positions than ever before (did I mention this already?). Here are the totals per decade for all position players who played four and five positions in a given season (min. of five games per position):

 So the last two decades have seen more and more multi-position players being used. Aside from a brief spike in the 1890s (mostly from 1890 to 1892 when players were being stretched thin—figuratively—because of demands from multiple leagues) there has never been anything like the versatility we see from today's bench players. By the way, there has yet to be a six-position player who meets the criterion (min. 6 games per position), but give them a couple of years, and they'll produce one.

So what has all this skimping and Rafael Belliard-ing around with the bench allowed the team to do with the pitching staff? Well, take a look at the average number of pitchers used in at least ten, twenty, and thirty games per year below:

Decade	10+ GP	20+ GP	30+ GP
1870s	1.53	1.09	0.78
1880s	3.16	2.32	1.74
1890s	4.82	3.44	2.35
1900s	6.07	4.51	2.99
1910s	7.72	5.58	3.94
1920s	8.70	6.93	4.96
1930s	9.31	7.44	5.03
1940s	10.32	7.97	5.02
1950s	11.22	8.71	6.02
1960s	11.87	9.58	7.32
1970s	11.67	9.20	6.81
1980s	12.58	9.62	6.82
1990s	14.89	11.13	7.85
2000s	16.50	12.27	8.97
Overall	10.27	7.86	5.58

Note that the slow historical growth of the staff went through the proverbial roof in the Nineties and has continued unabated ever since then. I wasn't kidding when I said Seventies staffs could hide a pitcher or two: under seven pitchers were used for at least 30 games per year back then. Now, almost nine pitchers are used that much, and I suggest that the 30-game numbers haven't shot up as much as the other two because teams cannot find as many pitchers that have enough talent for teams to stick with them a full season. Teams are using sixteen or seventeen men to fill out their twelve men staffs over a full season.