There is something satisfying about starting a research project. Even more so about completing it, even in parts. I have been delaying most larger numbers-based queries until the offseason, when the pressing need to respond to the here and now gives way to the more solid introspection necessary for the task. However, my wife is going on vacation without me, and, if you haven’t noticed between botched vaults and arguments over botched vaults, August is incredibly boring. So I rewarded myself with a tedious and questionable study of something with no predictive value and of only marginal and tangential interest to most baseball fans. ENJOY!

            I imagine that most people tend to focus on the pitching side of no-hitters: How many strikeouts did the pitcher record? Did he get lucky? How close to perfect was he? And so forth.

            I suppose that is all very interesting, but I can’t help but focus on the guys around the pitcher: How did they fare in their secondary roles? Most specifically, did they hit better than usual? Worse? No differently? And so I endeavored to answer that question, albeit in the second or third crudest way possible.

            Before I get into the parameters of the study, I would like to point out that I know that there are better ways to do this exercise. More correct ways, more precise ways, and god willing, simply easier ways. However, because the first two from that list could not be combined with the third, I chose to do it my way: a combination of accuracy and ease, for the sake of completion and because I wanted to see if there was anything there to investigate before I sent in the big guns.

            So I took all the no-hitters from 1956 until 2007 because, well, for the same reason almost all studies start in 1956. Retrosheet is a godsend; please see their disclaimer bit at the bottom of the article. This list includes no-hitters pitched in losses so long as the pitcher(s) threw 9 innings of no-hit ball, but not gems shortened by weather or by being the visiting team in a loss. Sorry, Matt Young and Andy Hawkins.* But if you’re looking for holes in the study…well, there will be more, so probably best not to focus on that one. The list does not include Lester’s no-no, for a variety of reasons, most of which you can probably figure out, given that…

            The next step was to take the actual number of runs and hits for the team whose pitcher no-hit the other and comparing those to the expected numbers of hits and losses. There are many terrible ways of calculating those, and probably one supremely complicated way of doing it correctly. I chose a middle path. I calculated the expected hits and runs for a team by finding that particular year’s team and its home and away splits and dividing the runs or hits at home or away by the number of games that year. Ways to do it worse: simply use average runs or hits for the league, use average runs or hits for an entire season without regard to ballpark. Ways to do it better: take out the stats from the no-hit game itself and divide by 76 or 80 (depending on the year) instead of 77 or 81, adjust for starting pitcher faced, figure hits and runs per inning and then multiply by number of actual innings, and many more I am sure will pop up in the comments below.

            From there, it’s math. Compare expected hits and runs (or average hits and runs, given the loaded nature of the word “expected” in these circles) with actual hits and runs in various ways. Look for anomalies, trends, happenstances, holy grails, etc.

            There are enough moving parts involved that any sweeping statements would be unsupportable and, frankly, idiotic. Still, the journey is the thing, and there are quite a lot of things you can learn just by traveling from end to end. The “results” are below, and anyone who wants the spreadsheet can have it (email me at demedici@gmail.com), but for me, the most interesting things are:

            - There were 127 no-hitters from 1956-2007. That’s 127 in 52 years, about 2.44 a year. From 1956 to 1968, there were 35 no-hitters. There were then 10 in the two years following the “Year of the Pitcher.”

            - In 1971, there were three no-hitters, all in the National League, and all in away games. In 1972, there were three no-hitters, all in the National League and all in home games. In 1973, the first four no-hitters were all in the American League and all in away games. Phil Niekro broke up this little streak by throwing a National League home no-hitter in August. Knuckleballers are always screwing things up with their zany antics.   

            - 83 of the no-hitters occurred at home, 44 away. Presumably, this has something to do with comfort levels and familiarity with the mound, and maybe even some home cooking by the umpires or official scorers.

            - From 1996-1999, every team that won a World Series had a pitcher throw a no-hitter in that year. The only other years that this happened were 1956 (Larsen’s perfecto in the Series itself), 1965 (Koufax), 1984 (Jack Morris), and 2007 (AA Pitcher Clay Buchholz). The pitchers involved in the streak were Gooden, Brown, Wells and Cone. There’s something poetic about four guys with serious issues and wonderful stuff coming together in a list like that.

            - The 2004 Diamondbacks were really, really bad.

            - I somehow always manage to forget that the American League switched to an 162-game schedule a year before the National League did.

RESULTS  

            Teams whose pitchers threw no-hitters scored 539 runs and gathered 1047 hits in 127 games, for per game averages of 4.244 runs and 8.244 hits. From their splits, they should be expected to score 549.282 runs and compile 1103.804 hits, or 4.325 R/G and 8.691 H/G.

            So they underperformed by 1.87% when it comes to runs, but 5.15% when it comes to hits. However, it’s important to note that these numbers are skewed because of a false assumption regarding ceilings and floors. A team has a floor on how much it can underperform offensively. That is, it cannot score fewer than 0 runs, nor have fewer than 0 hits. Thus, underperforming is capped. Overperforming is not. This plays a small part in driving the overall numbers, and demands a more thorough statistical analysis, only some of which I will do.

            On a game-by-game count, teams underscored their expectations 72 times and outscored them 55 times. They recorded fewer hits than their expectations 77 times and recorded more hits 50 times.

            In the 55 games where the teams outperformed their expected run total, they did so by an average of 2.426 runs. In the 72 underperformances, they came in an average of 1.996 runs short.

            When they exceeded their expected hit total, they did so by an average of 2.547 hits, while the underperformances clock in at an average of 2.393 hits under the expected total.

            So, as one would expect, teams overperformed at a higher rate (but less frequently) than they underperformed, hiding a little of the underperformance in the bigger picture. Still, I’m not terribly convinced there is anything supporting a claim that teams tend to hit worse when their pitcher is dealing filth.

            Alternative avenues of addressing this issue (in addition to the ones noted above) would be: using high game score games in addition to no-hitters to broaden the sample, using all of the no-hitters in history or determining on an inning-by-inning basis where the over- and underperformances occur. If the underperforming occurs in the later innings, maybe batters really do take it easy. I cannot explain, however, why the hit differential is out of proportion with the run differential, or why it should be. Still, very fun all the way around, even if it’s totally unsound and 25 people have done it before me.

* And Silver King, but he’s not in the time period anyway.

The information used here was obtained free of charge from and is copyrighted by Retrosheet.  Interested parties may contact Retrosheet at "www.retrosheet.org".