Living in Dangerous Times
Let’s start with the surprising conclusion: Batters are getting hit by pitches at near-historic rates. For all that you hear about pitchers who won’t pitch inside, and umpires issuing warnings that make it impossible to throw at hitters, and batters being unwilling to take one for the team, we’re seeing batters get hit by pitches at the highest rate since the turn of the last century.
I looked at each decade since 1901, the first year there were two leagues. Using FanGraphs’ Leaders page, I calculated the number of hit batsmen per 100 games played:
1901-1910 6.5
1911-1920 4.9
1921-1930 3.6
1931-1940 2.6
1941-1950 2.4
1951-1960 3.1
1961-1970 3.4
1971-1980 3.1
1981-1990 3.0
1991-2000 4.7
2001-2010 5.8
2011-2014 5.3
Baseball was kind of a wild game in the early days, with all sorts of shenanigans on the ball field, including throwing at batters. Hit batsmen were already in decline when, on August 16, 1920, Carl Mays hit Ray Chapman in the head with a pitch, killing him. Hit batters declined through the next three decades, bottoming out at 2.14 per 100 games in 1946. They stayed around 3 or so per 100 games through the 1980s, and then they took off. Here are the 30 years with the most batters hit by a pitch per 100 games:
1. 1901 8.0 11. 1907 6.1 21. 2008 5.5
2. 1903 6.9 12. 2004 6.1 22. 2011 5.3
3. 1905 6.9 13. 1909 6.0 23. 2009 5.3
4. 1902 6.8 14. 2003 6.0 24. 2013 5.2
5. 1904 6.6 15. 2006 6.0 25. 1913 5.2
6. 1911 6.5 16. 2005 6.0 26. 2010 5.2
7. 2001 6.2 17. 2007 5.8 27. 1999 5.1
8. 1908 6.1 18. 2014 5.7 28. 1998 5.1
9. 1910 6.1 19. 2001 5.7 29. 2000 5.1
10. 1906 6.1 20. 1912 5.5 30. 2012 5.1
Isn’t that strange? Every year from 1901 to 1913 and every year since 1998. Nothing from the intervening 84 seasons. It raises two questions:
- What’s going on? Why have hit batsmen increased despite efforts to cut down on beanball wars? It really has turned on a dime. There were 3.8 hit batsmen per 100 games in 1992, the 68th straight year below 4.0. It hasn’t been below that level since.
- When will it change? Andrew McCutchen’s plunking was a big story but didn’t lead to any calls for change. Amid laudable efforts to improve player safety, from batting helmets to neighborhood plays to home plate collision rules, hit batters are returning to levels not seen since the year before Babe Ruth’s rookie season. There have been some pretty terrible beanings, like Jason Heyward’s last year. Let’s hope it doesn’t take something worse than that to reverse the trend.
Writer for Baseball Prospectus
I’d be careful to attribute this entirely to pitchers and especially to pitcher intent. If we’re really interested in the increase in HBP you have to look at HBP by location to see if batters are maybe being stubborn about getting out of the way or are standing closer to the plate. Without some analysis like that it’s almost impossible to say what’s driving this, and I really doubt it’s some kind of increase in headhunting.
Maybe it has something to do with the armor that players sometimes wear to the plate now? Or the emphasis of OBP and not BA?
That’s a good point – batters are taking more pitches nowadays, so even if the rate of pitches that hit batters isn’t higher, the number of hit batsmen per 100 games would increase>
I checked HBP/1000 batters instead of HBP/100 games. There aren’t data before 1916, so the 1901-13 period I cited in the article fall away. The top 9 years are still contemporary (since 2001) followed by a mix of recent years and the Deadball II era. Specifically, the top 32 seasons for HPB/BFP are every season between 1962 and 1973 and every season from 1996 on. So we’re still in record high territory.
I agree that pitch location is pretty key here, xsturmin8, but of course the inflection point was years before PITCHf/x. There really hasn’t been that much variance in recent years. Good point by you and vslyke that this may not be a pitcher thing exclusively. I do recall hearing complaints about the protection guys like Bonds and Bagwell wore but it seems that I don’t hear as much about that anymore. Of course, that’s not data.
I’m going to ask around about this at the SABR Seminar, if anyone’s going. I’m the guy who looks like Groucho but grayer; say hi if you’re there.
well, for one thing, pitchers are throwing harder than ever before. Thus, it’s more difficult to get out of the way. VERY interesting article.
I agree with this reasoning. The average fastball has gone from the low- to mid-80s in the 1980s to almost 93 this season. Add in the fact that everyone has much sharper breaking stuff instead of the slow looping curveballs, and hitters have to stand in longer to make sure the ball isn’t breaking back into the zone.
It still presents something of an issue if it means hitters get hurt more often as a result, but the solutions are pretty complicated. Either pitchers have to throw slower (perhaps by making pitchers throw the whole game, like Bill James and others have advocated), or the hitters have to be given more time to react in some way. Another point in favor of backing up the mound, maybe?
I asked around at the Saber Seminar, got several interesting hypotheses, none of which panned out – will post them soon. I agree with Dan – if hitters get hurt more, something’s going to have to give, and it’s hard to figure out what. There isn’t a clear solution, or even a muddy one, like Rule 7.13 for home plate collisions.
Hey all, I’m submitting a followup post today, but in preparing it, I noticed a problem with the data. I downloaded League Stats from the FanGraphs Leaders tab for my historical data. What I didn’t realize is that for each season, FanGraphs totals games played by players rather than games played by team. By that I mean, the 1902 Senators played 138 games, but FanGraphs lists 1,279: 123 for Ed Delhanty and Bill Coughlin, 120 for Max Carey and Jimmy Ryan, 119 for Bill Kiester, 87 for, yes, Boileryard Clarke, etc. So I way over-counted games and therefore way under-counted HBP/game. (It seems to me that a team’s games should be the number of games it played, not the total games played by all its players, but maybe that’s just me.) Fortunately, the conclusions not only hold, they’re stronger. HBP/100 games were 63.7 in 1901-1910, 52.9 in 1911-1920, 40.9 in 1921-30, 29.4 in 1931-1940, 28.1 in 1941-50, 39.1 in 1951-60, 43.9 in 1961-70, 37.7 in 1971-80, 36.7 in 1981-90, 58.2 in 1991-2000, 72.1 in 2001-2010, 63.8 in 2011-14. The 30 seasons with most HBP/game were 1901-1911 and 1995-2013. The top ten are 2001, 2004, 2003, 2006, 1901, 2005, 2007, 2002, 2008, 1911. The maximum was 77.8 HBP/100 games in 2001, the minimum 25.1 in 1941. Sorry for the inaccurate figures, but I really was just going by the csv file I downloaded. I didn’t realize the unusual Games calculation.