This article is a theory on how to construct (not build) a major league pitching staff. It does not follow the sabermetrics of wins over replacement value, but reverts to the basics of the game itself.
The only true measure of each game is the final score. The whole purpose of baseball is to score more runs than your opponent. The simple fact is that during a course of a season, you need to be in a position to hold your opponent to less runs scored than your offense produces on average. If you score more runs than allowed on average, your club will be competitive (above .500).
You have to look at the basics of a season. Each team plays 162 games. Each normal game is set to last 9 innings. Therefore, on average, a team must have its pitching staff throw 1458 innings/season.
Modern team construction has each team with a five man starting rotation. And given the nature of pitch counts, injuries and a liberal definition of a quality start, most teams have a total pitching staff of 11 or 12 pitchers. The Cubs have been more apt to carry 12 pitchers and 5 position/bench players. So in a 162 game schedule, the five starting pitchers are expected to make 32.4 starts (lets round down to 32).
The current definition of a quality start is set too high. With the 2011 MLB staff average of 3.94 ERA, a "real" quality start should be defined as follows:
1 ER/5 IP = game ERA of 1.80, which is excellent.
2 ER/6 IP = game ERA of 3.00, which is very good.
3 ER/7 IP = game ERA of 3.85, which is just below the MLB average which is slightly better than the norm.
If you want a very good pitching staff, let's assume 6 quality innings pitched per start. 32 starts times 6 IP = 192 IP/ starter. Therefore, you starters should be able to handle 960 IP during the season (192 IP x 5). That is 65.8 percent of the innings needed to be played. That leaves the rest of the pitching staff to consume 498 IP.
The next elements of innings pitched is the end of game. Modern baseball has become more specialized as the starters are no longer expected to pitch complete games. The roles of the set up man (8th inning) and the closer (9th inning) have become standard operating procedure. But you do not want to overuse either role player. If you plug in 60 IP for the closer and the set up man, that is 120 IP off the 498 IP remaining. That leaves a total of 378 innings for the middle relievers to consume during a course of a season.
Middle relievers are considered "failed" starters. They are used to pitching long stretches during their career and have to adapt to the up and down daily call. However, if your team has 5 middle relievers to divide the work load, that equates to only 75 2/3 IP per reliever. Over a 24 week season, that means an average relief pitcher would throw 3 1/3 IP per week.
There is no current stat besides the nebulous "hold" for middle relievers. Instead of a quality start, we can look to a "quality week" for middle relievers as a benchmark for success. If over the reliever's weekly work load of 3 1/3 IP, he gives up 1 ER, that would be a weekly ERA of 2.70, which would be excellent, a "quality" week. If he gives up 2 ER, his weekly ERA would be 5.40, which would be horrible. This demonstrates the paper thin margin of error for the middle relievers on a baseball staff. A general manager would need to find 5 pitchers who are each capable of giving up 24 earned runs or less to have a quality middle relief bullpen.
So a season's work load can be divided as follows:
Starters: 960 IP
Middle Relievers: 378 IP
Set Up Man: 60 IP
Closer: 60 IP
Total IP: 1458
In 2011, the Seattle Mariners scored the least amount of runs (556). That means the M's averaged only 3.43 runs/game in offense. The league averages for runs scored was 723 for the AL, 668 for the NL and 694 for MLB.
In 2011, the Cubs scored 654 runs, below all league averages. The Cubs runs per game average (RPG) was 4.03. The NL Central RPG average was St. Louis 4.70, Cincinnati 4.53, Milwaukee 4.45, Houston 3.80, Pittsburgh 3.76.
By contrast, average runs allowed were as follows: AL: 717 R (656 ER); NL: 673 R (615 ER); MLB: 694 (634 ER). (Note: 8.6 percent of runs scored in the majors last season were "unearned.")
In the NL Central, team ERA was as follows: Milwaukee 3.63, St. Louis 3,74, Pittsburgh 4.04, Cincinnati 4.16, Cubs 4.33, Houston 4.54.
The RPG/ERA differential can be calculated as follows:
St. Louis: 0.96
Milwaukee: 0.82
Cincinnati: 0.38
Pittsburgh: (0.28)
Cubs: (0.30)
Houston: (0.78)
The RPG/ERA ratio almost mirrors the final standings:
Milwaukee: 0.82 -- 96 wins
St. Louis: 0.96 -- 90 wins
Cincinnati: 0.38 -- 79 wins
Pittsburgh: (0.28) -- 72 wins
Cubs: (0.30) -- 71 wins
Houston: (0.78) - - 56 wins
The numbers show that a team really needs to have a RPG/ERA ratio of 0.40 + to get to 81 wins (.500). To get 81 wins, using the NL RPG of 4.12 less the .40 ratio, a pitching staff ERA needs to be 3.72 or 603 earned runs.
We already calculated that a quality middle relief corps should only give up 24 ER/season or 120 in total. 603- 120 = 483 ER remaining. The closer and set up man should have an ERA of 3.00 or less, which would equate to 20 ER/each or 40 total. 483 -40 = 443 ER remaining. If you divide that by the five starters, they each can give up 88.6 ER in the season. This calculates to a starter ERA of 4.15, which is above all the league ERA staff averages and above our quality start ERA.
So what is the bottom line in this exercise?
You can construct a viable pitching staff in the NL Central with:
Five starters with an ERA of 4.15;
Five middle relievers with ERAs of 2.85; and
Two end of game relievers with ERAs of 3.00.
It is interesting to note that a team can have a starting staff of average pitchers and closers and still be competitive. However, if each middle reliever adds one additional ER/week, their ERAs soar to 5.70. The RPG/ERA ratio changes from a .500 competitive ratio of .0.40 to only 0.16 which is an approximate 73 win season at best. The middle relievers are the key bridge to an overall pitching staff's success.
