With a 20 wins and nine losses, the Milwaukee Brewers have the best record in baseball. While the offense and starting pitching has been fine, the bullpen is the real reason for the highly unexpected success. What was expected, however, are the calls for regression. Many analysts, including Dave Cameron of fangraphs, have expressed concern that the bullpen will not continue to be as good as they have in April. This is most likely true and not worth much debate.

However, calls for steep negative regressions also mean that the player or players have thus far done something amazing. The natural question is: how does the Brewers’ bullpen performance in April stack up against recent history?

**The Big Three**:

Baseball is a strange team sport. While the team win is the most valuable part of the box score, it is unclear how teammates affect each other on the field. In fact, statistical evidence has shown that team concepts such as chemistry, pitching to the score and line up protection have minimal if any effect on the outcome of a season.

To that end, instead of looking at the Brewers’ bullpen as a wholesome entity, here are three individuals with amazing April performances that have most affected and defined Milwaukee’s relief work:

**Best Brewer Relievers in April 2014**

IP | ER | BB | SO | |
---|---|---|---|---|

Francisco Rodriguez | 16 | 0 | 4 | 23 |

Tyler Thornburg | 14.2 | 1 | 5 | 17 |

Will Smith | 12.1 | 1 | 8 | 18 |

We can see that the performance has been outstanding and, likely, nowhere close to the true talent level of the three hurlers. The question worth asking here is, if bullpen pitchers act independently AND if one month performances are subject to large random fluctuations, “what are the chances that three relievers had the same fantastic month on the same team?”

**Baseball-reference play index**:

The query:

From 2004-2013, how many relief pitchers have had a month of pitching 12 or more innings while giving one or zero earned runs?

Answer: 485

485 out of 60 months means that you’d expect 8.23 relievers PER MONTH to have a month similar to the one the “big 3” are having.

In fact, this holds true for 2014 as well. In March/April of 2014 there have been six non-Brewers relievers that meet the criterion.

**Non-Brewers “ace” relievers from April 2014**

Name | Team | IP | ER | BB | SO |
---|---|---|---|---|---|

Chris Capuano | Red Sox | 14.1 | 0 | 2 | 15 |

Jean Machi | Giants | 13.2 | 1 | 3 | 10 |

Dale Thayer | Padres | 12.2 | 1 | 4 | 12 |

Chris Withrow | Dodgers | 12.1 | 1 | 10 | 18 |

Hector Rondon | Cubs | 12.1 | 1 | 4 | 14 |

Adam Ottavino | Rockies | 12.1 | 1 | 2 | 15 |

**Next Question**:

If roughly eight or nine pitchers are expected to have such numbers PER MONTH, what are the chances that three or more of them will be on the same team?

Assuming there are 210 relievers floating around at any given time (30 teams x 7 RPs) the probability that three of the 8.23 “ace” relievers are on one team is:

*drum roll*

**0.17%** (Per month)

**Conclusion**:

While the Brewers bullpen, especially the Big 3, have performed extremely well, one realizes that this occurrence isn’t super rare. The fact that roughly eight relievers a month have such performances certainly does not seem exceptional and while it’s more rare for three of them to be on the same team, a 0.17% chance isn’t rare enough for it to be a pure statistical marvel.

Still, that performance has been a big reason why the Brewers have a five game lead on the St. Louis Cardinals in the NL Central. And THAT’s special.

References:

Fangraphs.com

Baseball-reference, Play Index

http://stattrek.com/online-calculator/binomial.aspx

Follow me on Twitter: @vbarot87

h/t to Steve, @BrewersKeepTUTH , for the idea.

So… with 30 teams, and roughly 6 months of baseball per team, there are 180 opportunities for this to happen each year. If it is .17% chance each month, that is a 1.02% (.17% x 6) chance each year. If there are 180 opportunites and only a 1.02% chance, it means this happens only once every 100 years roughly or once every 18,360 opportunites (180 x 1.02%). How much rarer does it have to be for you to consider it a statistical marvel? I sure would consider the Cubs winning a World Series a statistical marvel… of course, that has been more than 100 years.

haha… touche.

First off, that is a great point and one I considered thoroughly before dropping. Here’s the reason why:

We’ve chosen a specific cut-off of 12+ IP and <1 ER for 3 RP's in one month. Is 11 IP and 2 ER THAT much worse? What about 14 Saves but 10 IP and 2 RP's?

What about a team having such a performance from April 15th to May 15th? It wouldn't show up in monthly splits but is just as good. There are an endless list of possible statistics that you could look at and see how rare it is. All of the tiny tiny probabilities add up to a 100% in a given season.

To me, "rare" in baseball is something like one in hunderds of thousands or millions. Having said all that, if you find 0.17% to be rare, that's great! It is after all a subjective point and if that impresses you then so be it.

Good argument with the idea of hundreds of thousands being a requirement for the argument of rarity. My counter-argument is that time should be a factor as well. A player hitting for the cycle, or a throwing no-hitter could be considered rare, but it would not be unreasonable to expect to see a couple every year. Pitching a perfect game would be closer to a statistical marvel, since it only has happened 21 times since 1900. And even better is the unassisted triple play, with 15 occurances since 1900. But even with as rare as those are, we don’t consider them to be unreasonably rare.

That being said, something that happens only once a century, in my opinion, is super rare. Something that is 6 times less likely than an unassisted triple play seems unlikely and a statistical marvel.

Stolen from Wikipedia (Sorry for not using a better source. My work blocks a lot of websites): “2] Cycles are rare in Major League Baseball (MLB), and have occurred only 304 times since the first by Curry Foley in 1882.[3] The cycle is about as uncommon as a no-hitter (282 occurrences in MLB history);[4][5] it has been called “one of the rarest”[6] and “most difficult feats”[7] in baseball. Based on 2009 offensive levels, the probability of an average MLB player hitting for a cycle against an average team in a game is approximately 0.00590%; this corresponds to about 2.5 cycles in a 162-game season with 30 teams” http://en.wikipedia.org/wiki/Hitting_for_the_cycle

21 times since the modern era began in 1900, most recently by Félix Hernández of the Seattle Mariners on August 15, 2012.

http://en.wikipedia.org/wiki/Perfect_game

it has happened only 15 times since 1900 at the major league level.

http://en.wikipedia.org/wiki/Triple_play