First, how do we check. My first thought is a simple measure of ERA for bullpens. To keep from doing too much data mining lets check ML bullpens every 5 years...2013, 2008, 2003, 1998, 1993, etc. back to 1948 (you'll see why).
| Appearances | ERA | OPS Against | Pen vs Starter Runs Saved Per Year |
|||||||
| Year | Teams | Relief | Per Team | Relief | Start | Spread | Relief | Start | Spread | |
|---|---|---|---|---|---|---|---|---|---|---|
| 2013 | 30 | 14336 | 478 | 3.59 | 4.01 | 0.42 | 690 | 726 | 36 | 68 |
| 2008 | 30 | 14157 | 472 | 4.10 | 4.44 | 0.34 | 731 | 759 | 28 | 55 |
| 2003 | 30 | 12958 | 432 | 4.16 | 4.52 | 0.36 | 736 | 764 | 28 | 58 |
| 1998 | 30 | 11962 | 399 | 4.19 | 4.55 | 0.36 | 734 | 765 | 31 | 58 |
| 1993 | 28 | 10301 | 368 | 4.04 | 4.26 | 0.22 | 727 | 739 | 12 | 36 |
| 1988 | 26 | 7331 | 282 | 3.54 | 3.81 | 0.27 | 693 | 697 | 4 | 44 |
| 1983 | 26 | 6760 | 260 | 3.59 | 3.99 | 0.40 | 699 | 721 | 22 | 65 |
| 1978 | 26 | 5891 | 227 | 3.61 | 3.72 | 0.11 | 706 | 701 | -5 | 18 |
| 1973 | 24 | 5323 | 222 | 3.71 | 3.76 | 0.05 | 707 | 702 | -5 | 8 |
| 1968 | 20 | 4790 | 240 | 3.00 | 2.98 | -0.02 | 645 | 636 | -9 | -3 |
| 1963 | 20 | 4808 | 240 | 3.44 | 3.47 | 0.03 | 683 | 680 | -3 | 5 |
| 1958 | 16 | 3555 | 222 | 3.75 | 3.90 | 0.15 | 715 | 720 | 5 | 24 |
| 1953 | 16 | 3113 | 195 | 4.27 | 4.10 | -0.17 | 741 | 724 | -17 | -28 |
| 1948 | 16 | 3049 | 191 | 4.25 | 4.09 | -0.16 | 715 | 708 | -7 | -26 |
Interesting...
- The biggest spread in ERA/OPS is the most recent year, the one with the most appearances by relievers.
- The year with the biggest negative difference was the 2nd last year listed, which saw just 4 more appearances per team than the 2nd worst in most categories.
- The games per team grew every time up to 1968 when it stayed stable, then dropped in '73, started to recover in 78 and back to the normal pattern in '83 although the ERA/OPS against had a heck of a peak in '83 which wasn't beat until this year.
- Massive improvement from '73 to 83 but nothing revolutionary since, although if you take out '83 you get reasonable improvement up to 1998.
I suspect key elements are...
- 50's to 60's: Save leaders used to be sub-20 with rare exceptions (save wasn't an official stat but it does show guys trusted with leads late). 1959 was the last year of that (3 guys tied for ML lead with 15) then guys like Roy Face started showing there was value in a closer. 1965 saw the first 30 save man (Ted Abernathy) the year after the first 29 save man appeared and the record was broke again the next year.
- Mid-70's: Goose & Fingers were the class of the multi-inning firemen who'd come in the 7th or 8th and last to the end. Then Bruce Sutter came on the scene and averaged around 100 innings a year from 79-84 and led in saves all but one of those years before blowing out his arm, showing the value of more specialization. We also saw the Yankees sign Goose even though their closer won the Cy Young the year before and they had a good setup man too, then they won the WS again after doing that which would've opened a lot of eyes.
- Early 90's: Oakland and the LOOGY/ROOGY/pure 1 inning closer - Tony LaRussa introduced us to the 6 and 7 man pen in this era with super specialization. Note how appearances per team jumped by over 80 from '88 to '93.
- Mid-2000's: Acceptance - games per team jumped by 40 from 2003 to 2008, partially due to teams in general accepting the 7 man pen due to the wear and tear of the steroid era I suspect. Clubs started needing more and more specialists to get through the Barry Bond types of that era and learned that more relievers means more ways to beat the other team. You can see overall ERA's have dropped since 2008, back to pre-1993 levels but the massive use of relievers is just growing and the effectiveness is finally over that freakish 1983 season.
The next revolution will be teams going to tandem starters I suspect - limiting starters and all relievers to one shot through the batting order ultimately in order to maximize their usefulness. I'm sure many in different teams have noticed that the pen, despite adding 2 guys since the 80's and adding 4 more teams (thus another 4 full pitching staffs) are seeing lower ERAs and lower OPS against. Focusing pitchers on using all their skills for fewer innings per game leads to better results. There is no question on that one. We see guys who in the 50's wouldn't be in AAA due to lack of multiple pitches who are getting ERA's in the 2s and even the 1s at times thanks to more focused use. I'd bet strongly that Gibbons would love to do that - go to a 3 inning max for all pitchers - due to his ability to juggle a pen effectively. Of course, that would probably lead towards a 15 man staff (5 guys available per game, 3 day rotation, 3 guys per start with 2 relievers available each game to bridge any gaps).
Now, what about offense? I'll save that for another day (trying to figure out a good method to measure regulars vs non-regulars... first attempt showed regulars, 500+ PA, reducing in quantity while relief use increased which seems counter to what one would expect so I'm going to have to think about it some more).
