Extreme Seasons

October 7, 2016
Where Have They Gone?
 
Where have all the errors gone,
Long time passing?
Where have all the sac bunts gone,
Long time ago?
Where have the complete games gone?
Gone to bullpens, every one.
When will they ever learn?
When will they ever learn?
 
My apologies to Pete Seeger for butchering his lyrics.  And, while we’re at it, might as well apologize to Bob Seger, and Corey & Kyle Seager, too, just to be safe. 
 
Before we leave the music part of the article, though, I thought I’d pass along a little something I stumbled across….
 
"Where Have all the Flowers Gone?" is certainly one of the most iconic folk songs ever recorded.  It has been recorded by many famous artists, including The Kingston Trio, Peter Paul & Mary, Bobby Darin, Roy Orbison, Johnny Rivers, Richie Havens, and….Bernie Sanders
 
Yep, that Bernie Sanders. 
 
In 1987, Sanders, who was the mayor of Burlington, Vermont at the time, recorded and released a folk/spoken word album called "We Shall Overcome".  The album cover credits Sanders and "30 Vermont Artists".  The tracks included such well-known songs as "Oh Freedom", "This Land is Your Land", "We Shall Overcome", and "Where Have all the Flowers Gone?"
 
The Wikipedia entry states that "Once word of the project spread among the Vermont music community, musicians began lining up to be part of it. The studio sessions were reminiscent of the Michael Jackson - Lionel Richie recording, We Are the World."  
 
Say what? 
 
Let’s see…….instead of Michael Jackson, Stevie Wonder, Bruce Springsteen, Tina Turner, Bob Dylan, and Ray Charles, Bernie was accompanied by the likes of Howard Mitchell, Ginny Peck, Emily Wadhams, David Weaver and Steve Rainville.  I suppose those were big names in the Vermont music scene in the ‘80’s, but I think that’s getting a little loose with the term "reminiscent".  Unless, of course, you’re the kind of person who considers Rosanne Barr’s version of "The Star Spangled Banner" to be "reminiscent" of Whitney Houston’s……
 
Moving on….
 
Extreme Seasons
 
As you are probably aware, in 2016 strikeouts established a new record, with the average team striking out just over 8 times a game (8.03).  This broke the previous high of 7.71 in 2015, which in turn broke the mark set in 2014 (7.70), and so on.  In fact, this was the 9th consecutive season we have seen a new high set for strikeouts per game.  Strikeouts per game have been steadily increasing for the last few decades. 
 
Average team strikeouts per game, by decade:
 
Decade
Team K’s per Game
1900's
3.55
1910's
3.65
1920's
2.81
1930's
3.32
1940's
3.55
1950's
4.40
1960's
5.70
1970's
5.15
1980's
5.34
1990's
6.14
2000's
6.56
2010's
7.52
 
As you can see, strikeouts, after dipping a bit in the ‘70’s, have been steadily increasing for the past 40 years. 
 
In addition, you are probably also aware that, even though strikeouts are at an all-time high, 2016 also had a near-record # of home runs, just missing the mark set in 2000.
 
Seasons with the highest average # of home runs per team per game:
Year
HR
2000
1.17
2016
1.16
1999
1.14
2001
1.12
2004
1.12
2006
1.11
1996
1.09
2003
1.07
1987
1.06
2009
1.04
 
So, 2016 ended up as the 2nd highest HR/game season of all time.  It’s received a fair amount of attention for that, but it seems different in many ways than some of the other seasons listed, in part because no individual player posted an unusually high total.  The MLB HR leader was Mark Trumbo of Baltimore with 47, a good but hardly an overwhelming total.  No 73 HR for Bonds, no 70 or 65 for McGwire, no 3 seasons of 60 for Sosa.  Not even a single 50.  So, in 2016, we saw a high # of HR’s in total, but with a lack of individuals who posted unusually high totals.
 
Here’s an interesting fact…..in 2016, there were 111 players hitting 20 or more HR’s, which is the highest figure ever.  That seems to be one of the keys for this year….no ridiculously high totals for any individual hitter, but with a lot more batters achieving solid totals.
 
Top 10 seasons with players hitting 20 or more HR’s:
Year
# Players with 20+ HR
2016
111
1999
103
2000
102
2004
93
2008
92
2006
91
2001
90
2009
87
2007
86
2003
86
 
So, in 2016, we had some exceptionally high figures in both K’s and HR’s.  However, we also saw some exceptionally low results as well, as we had very low instances of sacrifice bunts, intentional walks, complete games, and errors.  So, that got me to wondering about "extreme" seasons, seasons where we see categories on one end of the spectrum or the other.
 
Was 2016 an extreme season?  Certainly in some categories, it was.  But how could we measure that, and how does it compare to other seasons?  What makes an "extreme" season?  Does one or two extremes qualify…or does it need to be a season with numerous categories of extremes?  I thought I’d take a crack at analyzing it.
 
Approach
 
I thought about going all the way to the first season for which we have stats, 1871.  Talk about "extreme"….teams averaging 10.5 runs per game, teams averaging around 8 errors a game.  Yeah, that’s certainly extreme.  However, a lot of things were quite different then, so I decided not to go back that far.
 
The "Dead Ball Era" of the 1900's and the 1910's?  Sure, those were extreme years as well.  Few HR’s, lots of errors, lots of complete games, low batting averages, few runs scored.  The game being played in the Dead Ball Era was certainly "extreme" compared to today, but none of us (well, very few, anyway) that are alive today personally witnessed those years.
 
Ultimately, I decided I was more interested in how some of the basic indicators of the game have changed in our lifetimes.  My baseball experience dates back to the late 60’s, but I thought I’d go back a little further than that. 
 
For a fan who’s 70 years old (born in 1946), I figure that that person’s earliest memories might be from the mid-1950’s.  That seemed like a decent place to start.  In addition, one of the categories I was interested in was Intentional Base on Balls (IBB), and that has been tracked since 1955.  So, although I certainly could have gone back further, I decided to use 1955 as my beginning point, giving us 62 seasons to evaluate.
 
What I wanted to evaluate was how seasons have changed over time.  I wanted to compare, at a basic level, what a game in the 50’s was like vs. what a game today is like in terms of how often you tend to see certain things occur.  What did you see "a lot" in the 50’s when watching a ball game?  What did you tend not to see?  How about today?  How have some of the basic events changed in their frequency?  And, what seasons tended to be extreme, and which ones tended to be more in the middle?
 
I decided to take a look at several basic categories, measured in average occurrence per team per game (except for batting average and OBP).  The ones I decided on were:
 
Runs
Doubles
Triples
Home Runs
Stolen Bases
Walks
Intentional Base on Balls
Strikeouts
Batting Average
On Base Percentage
Sacrifice Bunts
Complete Games
Errors
Double Plays
 
For every season between 1955 and 2016, I ranked each season in each of those 14 basic categories, from high to low.  That gives us 62 figures for each category.
 
If a season had the "best" figure in a category, it was ranked as "1".  The lowest, "62".  "Best", for most of these categories, meant having the highest figure, with 2 exceptions:  errors and intentional walks (IBB), since those are really both defensive-based categories.  I considered the lowest error/game figure to be the best, so that would get a rank of "1".  Same for IBB. 
 
Although, really, in this approach, the direction being measured doesn’t matter a whole lot, because of the next step.  I took each season’s distance from the midpoint as a measure of how extreme it was.  In other words, finishing first or last in a category was equally extreme.  For example, the highest HR/game figure in the time frame was 1.17 in 2000, while the lowest was .58 in 1976.  In the context of the study, those were both considered to be equally "extreme"….they were just on different ends of the spectrum.
 
So, a rank of 1 and a rank of 62 were equally extreme.  To get distance from the middle, I took the absolute value of 31.5 minus the rank.  So, in this case, the year 2000 had a "score" of 30.5 (31.5 minus 1) for HR’s, and 1976 also had a score of 30.5 (absolute value of 31.5 minus 62).  So, the 2 years were equally extreme in that category.  The higher the "score", the more extreme. 
 
For example, here’s how the year 2000 looks in table form, with the raw data being expressed in #/team/game (except for batting average and OBP, which are just their normal expressions):
 
Category
 Raw Data  (#/Game)
Rank
Extreme Score (distance from middle)
R
         5.14
1
30.5
2B
         1.83
5
26.5
3B
         0.20
36
4.5
HR
         1.17
1
30.5
SB
         0.60
32
0.5
BB
         3.75
1
30.5
IBB (Fewest)
         0.25
9
22.5
K
         6.45
18
13.5
BA
.270
2
29.5
OBP
.345
1
30.5
SH
         0.34
44
12.5
CG
         0.05
45
13.5
E (Fewest)
         0.71
17
14.5
DP
         0.97
9
22.5
 
So, 2000 was very extreme in terms of runs, home runs, walks, batting average, and OBP.  On the flip side, it was middle of the road in stolen bases (ranked 32 out of 62, just about right in the middle), and triples.
 
If you average the 14 figures, it comes out to a figure of 20.14.  Among seasons since 1955, it’s a pretty extreme season….but not the most extreme.  It is among the top 10, however.
 
Results
 
Hopefully, that approach makes sense.  It treats all 14 categories equally, and tries to measure each season by how far from the middle it is across that spectrum of categories.
 
OK….enough talking.  let’s cut to the chase.  Here is the list of seasons sorted from most extreme to the least extreme since 1955:
 
Year
"Extreme Season Number" (XSN)
1968
22.86
1967
22.00
2014
21.93
1999
21.00
1955
20.93
1956
20.86
2016
20.64
2015
20.57
2013
20.29
2000
20.14
1963
19.86
2007
19.29
1972
18.86
1976
18.57
1966
18.14
1965
18.14
2012
18.07
1971
17.93
1996
17.86
2008
17.64
1961
17.64
2006
17.36
2004
17.14
1998
16.93
1997
16.86
2005
16.64
1975
16.57
1964
16.43
2011
16.36
2009
16.07
1980
16.07
2001
15.93
1974
15.79
2003
15.29
1995
15.21
1981
15.14
1969
15.00
1960
14.86
1962
14.71
1994
14.29
1978
14.29
1958
14.21
2010
13.93
1957
13.93
1979
13.86
1959
13.64
1973
13.07
2002
12.86
1970
12.14
1988
12.14
1987
11.36
1989
11.29
1983
10.86
1977
10.57
1992
9.57
1984
8.86
1993
8.43
1990
8.21
1982
7.93
1991
7.79
1985
6.93
1986
6.57
 
Now, I realize those numbers by themselves don’t mean a whole lot without digging in.  So, let’s look a couple up close:
 
1968 – "The Year of the Pitcher"
 
Here’s 1968 in table form:
Category
 Raw Data
Rank
Extreme Score
R
         3.42
62
30.5
2B
         1.19
62
30.5
3B
         0.21
29
2.5
HR
         0.61
61
29.5
SB
         0.47
48
16.5
BB
         2.82
62
30.5
IBB (Fewest)
         0.38
60
28.5
K
         5.89
28
3.5
BA
.237
62
30.5
OBP
.299
62
30.5
SH
         0.46
7
24.5
CG
         0.28
6
25.5
E (Fewest)
         0.85
40
8.5
DP
         0.87
60
28.5
Total Score
 
 
22.86
 
In a result that may not surprise many of you, 1968 ranks by this method as the most extreme season since 1955.   It is, after all, still known as "The Year of the Pitcher" (although some people also tried to hang that same label on 2010). 
 
In 1968, runs/game ranked last in the study, doubles were at the bottom, as were walks, batting average and OBP.  HR/game were the 2nd-lowest, DP/game were at the 3rd-lowest, and even IBB were extreme (since I measured IBB as "fewest", the rank of 60 actually meant that they were the 3rd "highest" in the study.  In other words, even though walks were low, there were a lot of intentional walks being issued that year).
 
So, most of those ranks were extremely low.  On the high side, there were a lot of sac bunts (ranked 7th among the seasons) and complete games (ranked 6th).    Basically, 10 of the 14 categories registered as fairly extreme.  The only categories that were relatively ordinary were triples, strikeouts, and errors.
 
Again, not a surprise.  I suspect that many of you, if you had to think of an extreme season off the top of your head, might very well select 1968.  It was quite a year.
 
1968 was, in many ways, the polar opposite of 2000, which we looked at earlier.  Compare the two years side-by-side.  For the most part, where 2000 was high, 1968 was low:
 
Category
2000 Figure
2000
Rank
 
1968 Figure
1968
Rank
R
5.14
1
 
3.42
62
2B
1.83
5
 
1.19
62
3B
0.20
36
 
0.21
29
HR
1.17
1
 
0.61
61
SB
0.60
32
 
0.47
48
BB
3.75
1
 
2.82
62
IBB (Fewest)
0.25
9
 
0.38
60
K
6.45
18
 
5.89
28
BA
.270
2
 
.237
62
OBP
.345
1
 
.299
62
SH
0.34
44
 
0.46
7
CG
0.05
45
 
0.28
6
E (Fewest)
0.71
17
 
0.85
40
DP
0.97
9
 
0.87
60
 
The only categories that were at least somewhat close to each other on the scales were triples, strikeouts, and stolen bases.  In most other categories, they were on opposite ends of the spectrum.
 
1986
If 1968 was the most extreme of the last 6 decades, what about the least extreme?  Invert the 6 and the 8, and you get 1986:
 
Category
Raw Data
Rank
Extreme Score
R
4.41
26
5.5
2B
1.55
27
4.5
3B
0.20
36
4.5
HR
0.91
27
4.5
SB
0.79
2
29.5
BB
3.38
16
15.5
IBB (Fewest)
0.31
34
2.5
K
5.87
29
2.5
BA
.258
26
5.5
OBP
.326
25
6.5
SH
0.36
39
7.5
CG
0.14
33
1.5
E (Fewest)
0.82
33
1.5
DP
0.93
31
0.5
Total Score
 
 
6.57
 
1986 really didn’t have any "signature" category, other than stolen bases, which were the 2nd highest in the data (1986 had Vince Coleman, Rickey Henderson, Eric Davis, and Tim Raines all stealing between 70 and 107 bases) .  Basically everything else was middle-of-the-pack, with virtually all categories ranked in the 20’s and 30’s. 
 
It wasn’t alone, either.  Of the bottom 10 seasons on the "extreme" list, 5 were in the 1980’s, and 4 were in the 1990’s.  That result did make me wonder if perhaps there was a bias in the data, in that some categories have had consistent directional trends over this time span, and the 1980’s and 1990’s, which would be essentially in the middle of the time spectrum I looked at, would naturally end up in the middle.  That led to this next section.
 
Trends
 
I thought it would be interesting to see how the various categories have been trending over the past 62 years.  I decided to look by decade, as well as looking at high and low marks during this time frame.
 
Runs/Game
Runs have meandered a bit this time frame, dipping in the ‘60’s, increasing over the next 4 decades, but then taking a step back in the current decade.
 
High: 2000 – 5.14
Low:  1968 -  3.42
 
Decade
Average
1950's
4.38
1960's
4.05
1970's
4.15
1980's
4.29
1990's
4.67
2000's
4.76
2010's
4.28
Overall
4.37
 
 
Doubles
Double have had a similar pattern as runs since the ‘50’s – a dip in the ‘60’s, then increasing over the next 4 decades, then dipping back down again, although they remain relatively high.
 
High: 2007 – 1.89
Low:  1968 -  1.19
 
Decade
Average
1950's
1.36
1960's
1.30
1970's
1.38
1980's
1.52
1990's
1.69
2000's
1.83
2010's
1.71
Overall
1.55
 
 
Triples
Triples have been steadily becoming more and more rare.
 
High: 1956 – 0.29
Low:  2013 -  0.16
 
Decade
Average
1950's
0.27
1960's
0.24
1970's
0.23
1980's
0.23
1990's
0.20
2000's
0.19
2010's
0.18
Overall
0.22
 
Home Runs
Home runs dipped for 2 decades after the 50’s, then started increasing over the next 3 decades before coming back down some again, but they remain relatively high.
 
High: 2000 – 1.17
Low:  1976 -  0.58
 
Decade
Average
1950's
0.91
1960's
0.82
1970's
0.75
1980's
0.80
1990's
0.95
2000's
1.07
2010's
0.99
Overall
0.89
 
Stolen Bases
Stolen bases were quite low in the 50’s, but climbed steadily over the next 4 decades before coming back down some over the last 20 years.
 
High: 1987 – 0.85
Low:  1955 -  0.28
 
Decade
Average
1950's
0.30
1960's
0.42
1970's
0.62
1980's
0.77
1990's
0.73
2000's
0.58
2010's
0.59
Overall
0.59
 
 
Walks
Walks have meandered a bit over this time frame…down, up, down, up, down, down, and at the moment they’re relatively low.
 
High: 2000 – 3.75
Low:  1968 -  2.82
Decade
Average
1950's
3.44
1960's
3.14
1970's
3.30
1980's
3.22
1990's
3.43
2000's
3.34
2010's
3.04
Overall
3.27
 
 
Intentional Base on Balls
The use of the intentional walk became more popular over the first couple of decades it was tracked.  Since the 1970’s, however, it’s been on the decline, and with each passing year is getting used a little bit less.  In 2016, the intentional walk was used less frequently than it ever has been since it they started tracking its usage.
 
High: 1967 – 0.40
Low:  2016 -  0.19
 
Decade
Average
1950's
0.29
1960's
0.32
1970's
0.34
1980's
0.32
1990's
0.29
2000's
0.27
2010's
0.22
Overall
0.30
 
 
Strikeouts
As mentioned before, strikeouts have generally been on the increase over the past several decades, and 2016 marked the 9th consecutive season that a new high mark was realized.  In 2016, strikeouts per game was 83% higher than it was in 1955.
 
High: 2016 – 8.03
Low:  1955 -  4.39
 
Decade
Average
1950's
4.78
1960's
5.70
1970's
5.15
1980's
5.34
1990's
6.14
2000's
6.56
2010's
7.52
Overall
5.89
 
 
Batting Average
Batting average has seen a couple of different trends over this time span.  It dipped in the ‘60’s, then increased over the next 4 decades, but has since retreated to its lowest figure since the ‘60’s.
 
High: 1999 – .271
Low:  1968 -  .237
 
Decade
Average
1950's
.258
1960's
.249
1970's
.256
1980's
.259
1990's
.265
2000's
.265
2010's
.254
Overall
.258
 
 
On Base Percentage
OBP reflects the same pattern as batting average, and in the current decade is down to its lowest levels since the ‘60’s.
 
High: 1999 – .345
Low:  1968 -  .299
 
Decade
Average
1950's
.327
1960's
.315
1970's
.323
1980's
.324
1990's
.334
2000's
.335
2010's
.319
Overall
.325
 
Sacrifice Bunts
Sacrifice bunts have steadily been on the decline, and reached a new low in 2016.  From 1956 to 2016, we have seen a 60% decrease.
 
High: 1956 – 0.51
Low:  2016 -  0.21
 
Decade
Average
1950's
0.46
1960's
0.45
1970's
0.45
1980's
0.39
1990's
0.37
2000's
0.33
2010's
0.28
Overall
0.39
 
Complete Games
Complete games have gone the way of the dodo. 
 
  • In 1956, an average major league team had nearly 50 complete games.  Robin Roberts led with 22.
  • In 2016, an average major league team had 3 complete games, with no team having more than 10 (San Francisco).
 
High: 1956 – 0.31
Low:  2016 -  0.02
 
Decade
Average
1950's
0.30
1960's
0.25
1970's
0.26
1980's
0.16
1990's
0.08
2000's
0.04
2010's
0.03
Overall
0.15
 
Errors
Errors held pretty steady through the ‘70’s, but have been dropping ever since.
 
High: 1975 – 0.96
Low:  2013 -  0.56
 
Decade
Average
1950's
0.88
1960's
0.88
1970's
0.88
1980's
0.82
1990's
0.73
2000's
0.64
2010's
0.60
Overall
0.78
 
 
Double Plays
Double plays fell off from their 1950’s level, and stayed pretty steady from the ‘70’s through the 2000’s, but have dropped off some in the current decade.
 
High: 1958 – 1.05
Low:  2014 -  0.87
 
Decade
Average
1950's
1.01
1960's
0.92
1970's
0.95
1980's
0.94
1990's
0.93
2000's
0.94
2010's
0.89
Overall
0.94
 
 
Wrapping it Up
 
In conclusion, 2016 does rate as a fairly extreme season, the 7th most extreme since 1955 by this method.  Here’s 2016 in table form:
 
Category
Raw Data
Rank
Extreme Score
R
4.49
20
11.5
2B
1.70
19
12.5
3B
0.18
54
22.5
HR
1.16
2
29.5
SB
0.52
42
10.5
BB
3.11
50
18.5
IBB (Fewest)
0.19
1
30.5
K
8.03
1
30.5
BA
.256
40
8.5
OBP
.322
43
11.5
SH
0.21
62
30.5
CG
0.02
59
27.5
E (Fewest)
0.59
3
28.5
DP
0.90
48
16.5
 
 
 
20.64
 
Some of it was attributable to the dramatic increase in HR’s that we saw this year, and some of it was attributable to the ongoing trends of increasing strikeouts as well as generally decreasing numbers of complete games, errors, sacrifice bunts, and intentional walks that have been in motion for many years.   On the other hand, in terms of runs, batting average, OBP, doubles, and steals, 2016 was kind of middle-of-the road.
 
So, regarding the experience of watching a typical game in 2016?  You’re bound to see lots of strikeouts and lots of home runs.  However, since walks and OBP are relatively low, the high number of home runs doesn’t necessarily translate into a lot of runs.  Strategies such as sac bunts and intentional walks are being used less often than ever before.  And if you’re hoping to see a complete game pitched by your favorite team….well, good luck!
 
Thanks for reading,
Dan 
 
 

COMMENTS (6 Comments, most recent shown first)

grising
Thanks for the enlightening article, Daniel!

I've done some (amateur) research into different eras, so I found this article very interesting. I'm not surprised by 1968 and the years in the 2000-2016 era as being outliers. But I was surprised by 1955 and 1956. Without looking, I'm guessing that those years had very low SB and high HR and BB.

By the way, I use year and league (AL or NL) in my research. However, it looks like you used MLB for this article. I think that's okay because AL and NL stats correlate pretty closely year to year. But there are differences, so you might want to dis-aggregate by league as well as year for future research.

Thanks again for the article!
7:06 PM Oct 8th
 
mauimike
I'm pretty sure I f**ked Ginny Peck. I was in Vermont in 1974. Traveling across the Country. I was heading north to Canada, I always got lucky in Canada. I saw a sign, New Hampshire, two miles. I thought here's an easy way to get to New Hampshire. I hung a right. Drove two miles. Crossed the border. Made a U-turn. When I got back into Vermont down the road apiece. There was a vegetable stand. The girl was pretty. I was hungry. I stopped. One thing led to another. It was the 70's. I had a van. I spent a couple days in Vermont and Ginny.
3:32 AM Oct 8th
 
Gfletch
Dan, I think you may have just scratched the surface of a more interesting project, an analysis of the ongoing trends of the game, of where the evolution of baseball may be taking us. You chose to focus on looking for extremes (which is fine) but I am not sure that this is the most instructive thing that can be done.

I am thinking of the NHL's Plus/Minus statistic...an example of a slightly sophisticated statistic that is less interesting than the raw data that it is composed of.

I took your method results and graphed them year by year. The results are choppy, but they clearly show that the first 15 years in your study were very extreme, that the least extreme years were in the middle years, that baseball play is becoming more extreme (trending to be more and more so) since about 25 years ago. It does seem to me that the most interesting and entertaining baseball is being played during the least extreme seasons.

I'm thinking a great study would be to simply look at the raw counting stats and look at the per game incidence of those things occurring season by season, going right back to 1870. I think we might use the observations to see where the game is headed, might also see where the indications are that the game is going to change directions.

Thanks for a thought provoking article.
2:41 AM Oct 8th
 
DMBBHF
Correction - I referred to the Dead Ball Era as the "teens and twenties". I meant to say the 00's through the teens. I will correct as soon as possible. Sorry about that.
10:06 AM Oct 7th
 
DMBBHF
I should clarify that, while OBP and walks in the current decade are relatively low, they have trended up the last couple of years. But, for the 2010's as a whole, they're relatively low.
9:08 AM Oct 7th
 
rtallia
Anyone else surprised that OBP is going down, given its importance? Or is this simply a function of pitching being better?
7:12 AM Oct 7th
 
 
©2019 Be Jolly, Inc. All Rights Reserved.|Web site design and development by Americaneagle.com|Terms & Conditions|Privacy Policy