Fun With Pythagoras, pt. 4

added 5/23/2005 by Scott Barzilla

Columnist note: Last time we looked at the offensive side of the game. This time we will look at pitching. In particular, conventional wisdom tells us that a good bullpen is very important in a close game. We will test that conventional wisdom in this edition.

We have looked at a number of different factors dealing with Pythagorean records, but pitching might be the key. It certainly makes sense when you think about it. If we look at offensive efficiency in close games then the other half of that equation (pitching). In close games that usually ends up involving the bullpen. As we have seen over the last couple of decades, the bullpen has seen more and more work. This has had a profound effect on how managers manage their pitching staff.

When we look at Pythagorean record we can point our finger at two things a manager does that have overwhelming effects on this. A manager’s offensive philosophy (when to run, when to bunt, when to hit and run) has more effect on the outcome of the game than almost anything. However, there is no greater effect on the game then how a manager handles the pitching staff. How long do you go with the starter? Who do you bring into the game and how long do you give them? Do you utilize specialists? Yes, the days of leaving the starter out

In the 1970s, we saw teams steadily move towards limiting starters to seven innings, but typically they would employ only two or three pitchers to use in the interim. The likes of Goose Gossage, Rollie Fingers, and Mike Marshall routinely pitched more than 100 innings in relief. Older fans will remember the days of Big Red Machine with Pedro Bordon, Rawly Eastwick, and Clay Carroll. Even those days are long gone.

The implication is that a lot of fans miss the important members of the bullpen when the focus on the closer. The modern closer may pitch between 50 and 70 innings. This leaves the rest of the bullpen to take up the remaining 200 or 300 innings of important relief work. Even if we assumed that the closer could go 100 innings you would still have to find at least three more solid relief pitchers to have a solid bullpen.

Good teams like these teams usually have good starting rotations and a good closer. Nearly every team we have studied have had great closers throughout the last decade. The rest of the bullpens have been the key between success and failure. The Yankees embodied that in the mid 1990s with John Wetteland, Mariano Rivera, and Mike Stanton. Unfortunately, teams cannot keep pitchers this good together for a very long time. For instance, Billy Wagner, Brad Lidge, and Octavio Dotel were only together for one season.

For our purposes here, we want the relievers to have the same ERA or lower than the starters. A significantly higher ERA is a sign that a team is struggling in the late innings. That could be a huge symptom of a team that struggles in one run games. We will also take a look at how the team did in saves against the league average. Unfortunately, blown saves have not been recorded back to 1995, but a correlation can clearly be seen between low save totals and shoddy relief pitching. Even though the lower the ERA the better we will switch the ratio numbers to be consistent with the offensive variety. If you are below 100 you are below average and vice versa.


		   INN          ERA          SV        ERARatio      SVRatio	  Combo
Braves      5127      3.65      457       98      113      211	
Red Sox     6242      4.49      436       96      110      206
Indians     5855      4.58      404      100      102      202
Astros      5354      4.11      414       98      102      200
Yankees     5169      4.25      494      100      125      225	           
Cardinals   5307      4.05      417      102      106      208       

We can break these numbers down for further analysis. The first category should not be overlooked by any estimation. The number of innings a bullpen hurls can be a reflection of two things: the quality of the rotation or the philosophy of the manager. In this case, we should not be shocked that the Braves and Yankees come out looking very good here.

The Braves had Maddux, Glavine, and Smoltz for the majority of the last ten years while the Yankees have had guys like Andy Pettitte, David Cone, David Wells, Roger Clemens throughout the years. Fewer innings means fewer relief pitchers. Generally speaking, middle relievers are not reliable. If they were reliable they would be starting pitchers or closers. A quick math tutorial shows the following per season data on bullpen innings.


            INN PS     INN PG
Braves       512.2        3.0
Yankees      516.2        3.1
Cardinals    530.2        3.1
Astros       535.1        3.1
Indians      585.2        3.2
Red Sox      624.1        3.2

Notice how these numbers breakdown to look very similar on a per game basis. Yet, during a season, the Red Sox see an additional 100+ innings from their bullpen than the Braves and Yankees. Why does this happen? Well, the Red Sox have had a shaky rotation throughout those ten years. Certainly, when you look at them as a team you see offense as the focus over the last decade.

Those additional innings usually end up going to relievers that simply shouldn’t be relief upon. The Astros and Red Sox have one thing in common: Jimy Williams. When George Streinbrenner accused Williams of rubbing it in with the six pitcher no-hitter those of us in Houston laughed. He did that kind of stuff all the time. Grady Little was said to be cut in Williams image, so you almost had half of your games managed by a Captain Hook in Boston.

The second category we can look are the saves and save ratios. The save ratios were compiled by dividing the team’s total by the league average. Here we see the Astros in last place in saves ratio with the Cleveland Indians. This doesn’t seem to jive with the knowledge that Billy Wagner was the closer throughout much of that period. Wagner will likely go down as one of the top five closers in the last decade. Cleveland has had more closers than that roll through town.

It’s hard to pinpoint why this would be the case, but we could surmise that the Astros have had fewer save opportunities than their adversaries. This is because the Astros have been on the winning end of a lesser percentage of close games than four of the five other teams. Those of us that have watched the team know they win an inordinate amount of games by large margins. In fact, I did a study for addictsports.com last year showing this to be true.

As for the third category: bullpen ERA. We can also see the Astros finish near the bottom of the list. Only the Red Sox had a worse ratio than the Astros in this category. Of course, the difference is not enough to write home about, so like the other categories, relief pitching cannot explain all or even most of the inefficiencies for the Astros, but we are seeing a trend developing. This can clearly be seen when we combine the ratios. Even though the Astros did not rank at the bottom of either category they rank at the bottom of the combined rate. The Astros have ranked in the bottom half in every category we have introduced whereas the other clubs have been up and down.

Perhaps, the club is scuffling in Pythagorean record because they don’t excel at any of the categories we have seen so far. If we were to offer any advice to the Astros it would have to be two-fold. Either, the club could extend their starters that extra inning or they could sign one or two more productive relief pitchers. Certainly, the last off-season offered them many opportunities to sign productive middle relievers, but they declined to sign any of them.

Coming Up Next

Now that we have covered managers, one run games, offense, and pitching we have only one major category to look at. Certainly, a lot has been said about defense and its affect on the game. We will sort through some of it (we certainly can’t do everything) in the next edition to see if defense has any effect on our dilemma.

Scott Barzilla is the author of “Checks and Imbalances” and “The State of Baseball Management.”