Tracking Trends In Baseball History

By Gary Fletcher


“What were things like, Grandma, back in the old days?”

“Well, Gary, for one thing baseball used to be played by human beings. They played bare handed or else wore gardening gloves. Batters used magic tree branches for bats, and worked in coal mines 24/7, how the hell they ever managed to play games is beyond me, but they did. And went to church and begat children, too. It was a wonderful time.”

“Aw, Grandma…”

Yeah. Obviously we don’t know everything. Fortunately we do have a statistical record of what happened in games well over a century ago, and lots of written accounts from days gone by as well. The statistics don’t say anything about magic or clutch hitting, not then and not now, but…

Well, consider Errors. Here is a chart showing errors per game year by year 1876 through 2016:

View the chart, it’s just one click, it won’t hurt, and it opens in a separate tab

The decline of errors, at first precipitous and then steady, is no surprise to anyone reasonably well acquainted with baseball history. Let’s look at home runs per game:

Click here to view chart in separate tab

Kind of the reverse of Errors per game, though I don’t have any confidence of constructing a logical connection. (Well…more balls over the fence means less opportunity for a fielder to be credited with an error…no doubt that is a small part of a larger explanation).

But you might expect a relationship between HBP and HR, right? Old baseball books are full of stories of knocking down batters after a home run was hit, stuff like that. Let’s track both things in one chart (HBP is the blue line). This one starts in 1884 (no HBP data previously).

Click to view chart in separate tab

Not sure how to interpret this one. The lines cross at 1920, suggesting if anything an inverse relationship. I think this is just the home run explosion independent of HBP. There does seem to be some low level relationship between HR and HBP starting around 1950.

With some misgivings about charting things back to the cretaceous era and the often missing or suspect records of those days, I still thought it would be fun to track the frequency – the per game frequency – of actual in-game events. I’m talking about stuff like singles and doubles, errors and double plays (either turned or hit into),  stolen bases, sacrifice flies…but not batting average or WAR or Win Shares or anything that isn’t directly witnessed by fans at a ball park.

Why per game? Well, a few reasons, but mainly I am thinking in terms of what a fan might see during a game at any one point in history. Granted, how often a fan of the Houston Astros might see a home run in 1975 is quite a bit different from what a fan of the Cubs or Red Sox might see. But this is mostly fun, and I don’t want to get too tangled up in splitting decimal points, or even full digits in some cases.

All of this information can be downloaded from BBRef * in the team standings sections for standard hitting, pitching and fielding. With a few adjustments (Games for all MLB teams divided by two gets you actual games played) and some use of basic excel functions, you can get the year by year per game totals for individual events, or groups of events, and chart them either separately or combine them for contrast. Some people like to ski, some people like to drink, some people like to attend the opera. Man, they don’t know what they’re missing!

* That’s a lot of digging. And digging is work. And I never work. How can I trick somebody into doing the digging for me? Who’d be dumb enough? †

† Carl Barks, from the mouth of Gladstone Gander

I don’t see nearly as much baseball as I used to. As I have grown older I’ve become much less interested in the creatures in the sea or on the beach, more interested in the ocean and the tides. I don’t mean that I find no pleasure in watching a game, or in the accomplishments of its participants. (I do, when life, opportunity and being awake allow it).

I find that the sabermetric community seems mostly interested in the questions of, ‘how good?’ How good is this shortstop? What are his chances of getting into the Hall Of Fame? Is this Cubs team a great team? How do they stack up against great teams of the past? Who is the greatest player of all time? Who is the worst?

And that’s fine. That’s inevitably how it should be, anyway, because what drives the evolution of baseball (and sports, business, life, the universe) is competition, after all. And who wins (or loses) is surely a question of who is better (or worse). *

* Need to send out for more parentheses

Still… If there’s no audience there just ain’t no show.† Again, I’m just wondering about the chances for a fan at the ballpark seeing something…in this case, the chance of seeing some event that has been recorded statistically.

† from the song, Rain-O, by Chilliwack

These are my first baby steps in understanding the evolution of baseball in the statistical record, starting with what it looked like when MLB first took form. Of what things become more frequent, or less frequent. What things went away, then came back? What things appear to be related? And where are we going?

I believe that what fans like isn’t lots of runs being scored so much as the threat of runs being scored. The two must be related of course. Consider:

Runs Scored per Game:

Click to view chart in separate tab

This shows a pattern of oscillation, like a pendulum swinging back and forth but with less and less energy on each swing. It suggests that if MLB were to truly stabilize (and become relatively changeless) that runs would end up at around 9 per game…kind of like where it is right now.

Singles, on the other hand, are on a long downward trend. There are still a lot of them, but vs doubles, triples and homers, their share of hits isn’t as large as they used to be. In fact, that makes for a good chart, tracking the percentage of each type of hit. Well, let’s add walks in there as well. A rate of each of these things per game:

Click to view chart in separate tab

Interesting that the frequency of singles and walks appear to be converging.

Boy, we all thought it was just PEDs, but the trend has always been towards more and more power in the game. We are now at a virtual dead heat with the highest HR per Game rate in history, set in 2000. These extra home runs are not translating into more runs, however.

Just looking at walks:

Click to view in separate tab

Kind of stable since World War II, really, but overall trending downward post the PED era.

Okay. I can play around with stuff all day, but believe it or not I was leading up to something. In this next chart I will attempt to track the frequency of what we might call ‘small ball,’ the little things that people like to talk about. That is:

Stolen Bases and Caught Stealing, Hit By Pitch, Sacrifice Hits (Bunts), Sacrifice Flies, International Walks, Balks, and Double Plays. *

Click to view in separate tab

* Of course there much smaller things than these, but just not statistically recorded in a bulk way year to year for these purposes

This chart starts in 1955 for the simple reason that these categories, as a group, are not fully tracked until then.

Here’s the deal. While watching the post season games this year I was struck by a number of games that seemed to turn on little plays…I am not sure if this was just because of the broadcast guys who may have over-emphasized these things, or if I tended to do that for myself. Maybe both things. But in any case I felt the games were more fascinating as a result.

All these little things…the stolen base, place hitting…hitting behind the runner…caught stealing…all these things (again, as a group) being either one run strategies or at least one run results, were in retreat from the game since the Babe Ruth revolution. .

Whether my efforts here show this or not, I don’t know, but I believe that while the power game added something exciting to the game, it also began to whittle away at the little game.

And logically it should. The risk of losing a baserunner when the batter might well double, triple or (especially) hit a home run…when, in fact, just being at the plate means you have a runner in scoring position…that risk has become less and less reasonable as power hitting has gained more and more prominence.

But the hell with the logic. I don’t like it. I believe that the game was better when we had more balance. And according to the last chart, the game had terrific balance, the most balance, from about 1975 to about 1995. (Convenient for my old fogey tendencies).

I don’t blame PEDs for the whole thing; they just gave a big booster shot into the rear end of a trend that has always been there, the trend toward more power in the game. When MLB cracked down on the chemicals it provided a temporary relief, a temporary detour from that trend. But we are right back where we were headed, with home runs per game at essentially the highest point ever, and headed further.

What to do? Hell, I don’t know. But let’s not overreact, after all. There are still a lot of things going on ball games today.

Here’s Homers and Little Things in the same chart for comparison:

Click to view in separate tab

The trending isn’t looking good, but there are still a lot more things going on in the game than home runs.

You still here?

Good, hope you enjoyed some of this with me.  But, enough is enough for now. I could spend an unforgiveable amount of time examining these trends and producing these charts.

And you can do this yourself if you wish. I have uploaded the per-game data to the BJOL Stats Depository. At this time that file – Trends Data FSR – is at the top of the list, or was the last time I looked. There are about 40 categories, and I guess about 1600 one to one comparisons you can make, and I don’t know how many other multiple comparisons.

If you are not a BJOL member, and are interested, email me at and I  will send you the excel file.

I think its fun, and I think it’s interesting. I hope it’s a useful source of compiled data. If you agree, you are welcome to the information.




Gary Fletcher

December 19, 2016

Leave a reply