Walks & strikeouts diverging in the past 2 years

As shown in the graph below, over the past 20 years, walk and strikeout rates (as a percentage of plate appearances) have usually gone in the same direction from year to year, and have usually been within 5 percentage points of each other. A couple of notable exceptions:

  • In 1999, the year after McGwire and Sosa shattered the season HR record, the BB rate surged by almost 8%.
  • In 2001, MLB directed umpires to enforce the real strike zone instead of their own creative interpretations. The result was drop of almost 12% in walk rate — the largest one-year drop of the past 40 years (at least). This might be considered a reaction to the 1999 spike in BB%, which was consolidated in 2000.
  • In 2002, some of the extreme changes from 2001 were reversed. At first blush, this seems like regression toward the mean, or what Bill James called the “Plexiglass Principle.”

For every year from 2003-09, BB% (blue) and SO% (red) went hand-in-hand, whether up or down. For 2006-09, both rates rose about 2% per year.

But in 2010, walks fell by over 4%, while Ks continued their inexorable climb. And last season’s walk rate fell another 4.8%, while Ks rose just a tad. The 2011 walk rate, 8.10% of PAs, was the lowest since 1968, while the 20.8% SO rate was the highest ever.

The last two years are the first time in modern history that the walk rate fell at least 4% in consecutive years. The only other consecutive drops of over 3% was 1951-52, which followed a historic high in walk rates from 1948-50.

Theories on the 2010-11 trend?

0 0 votes
Article Rating
Subscribe
Notify of
guest

55 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
DaveR
DaveR
12 years ago

Strikeouts are boring. Plus, they’re fascist.

Neil L.
Neil L.
12 years ago
Reply to  DaveR

Strikeouts aren’t boring if your closer has just stranded three runners to win the game, but I get your point. They don’t make good SportsCenter highlights.

Hartvig
Hartvig
12 years ago
Reply to  DaveR

Strikeouts can be exciting if Mitch Williams is your closer and walks 3 & strikes out 3 in the 9th with a 1 run lead.

Otherwise, not so much.

Lawrence Azrin
Lawrence Azrin
12 years ago
Reply to  DaveR

…and ground balls are democratic, so sez Crash Davis.

bstar
12 years ago

I think there may have been another directive to the umps about enforcing the strike zone, albeit an unofficial, behind-the-scenes one. Watching the Braves-Mets this weekend has done nothing but reinforce my thought that umps are calling the high strike with more and more frequency(as well they should).

Just a hunch.

Doug
Editor
12 years ago
Reply to  bstar

Agree. Last year or two, pitches between the belt and letters seem to have a better than even chance of being called strikes. Five or ten years ago, for most umpires, very seldom a pitch more than a smidge above the belt would be called a strike.

Or, at least, that’s my imperfect recollection from my observation of the tiniest fraction of big league games. 🙂

kds
kds
12 years ago

Year to year changes are probably mostly noise.
Note that when there is a change in the strike zone, or the ump’s enforcement of it, that you would expect the K rate and the BB rate to move in opposite directions. You are just switching balls ton strikes or vice versa. If the change is due to batter behavior, taking more pitches looking for a good one to hit hard, then you would expect the rates to move in parallel.

bstar
12 years ago

2010 looks extremely similar to 2001. 2011 is more perplexing.

Neil L.
Neil L.
12 years ago

I’ve been puzzling over the meaning of your study for a while, JA. Thanks for doing it. However, I haven’t been able to make sense of the data. On a very simple level one would expect the BB% and K% to correlate slightly negatively. The graphs should go slightly in the opposite directions over the previous year, barring any unusual circumstances like a redefined strike zone. The data span 20 years: in 13 of the years the rates correlate positively, one year (1996) the BB% rate was effectively zero, and in the other six years (1997, 1999, 2001, 2002, 2010… Read more »

Neil L.
Neil L.
12 years ago
Reply to  John Autin

It’s a big project, but I would like to wrestle one concept to the ground and prove it right or wrong.

You make some good points about the change in approach with two strikes, but, if we are entering anothe “little ball” era, post=steroids, then wouldn’t more walks be expected?

Neil L.
Neil L.
12 years ago
Reply to  John Autin

I’m not thinking you’re wrong, JA, but why question is why is there a positive correlation? Why trhe fewer walks, the fewer strikeouts and vice versa?

Mark in Sydney
Mark in Sydney
12 years ago
Reply to  John Autin

JA,

The correlation coefficient is 0.387, so there is a link, though mild (this based on per year/per team since 1962).

As for the trend, I don’t see it right now. On a per-team basis, there are outliers for last season like the Giants with more K and less BB than anyone, and the Oriels who gave up a load more walks than anyone else.

Certainly the numbers say that what we have seen in the last two years is within the realms or normality.

Neil L.
Neil L.
12 years ago
Reply to  John Autin

Mark, thanks for weighing in and crunching the numbers. I’m going to take your coefficient of 0.387 between BB and K as being gospel, going forward.

I’m surprised it is not a little higher, especially considering you went back so far into baseball history.

JA’s mention of pitches taken per plate appearance over the years is clearly relevant as is the correlation of these outcomes (walks and strikeouts) to home runs.

kds
kds
12 years ago

If we look widely over baseball history one thing that is clear is that K’s and K rates have risen almost continuously for a many decades. Walks have been much more stable. 1949 was a high point for walks, about 10.5% BB/PA compared to last years 8.1%. In 1949 the K rate was 9.3%, last year, 18.6%. I think it would be interesting to look at correlations among K rate, walk rate, HR rate, and scoring. Iirc HR rate and K rate went up as scoring went down in the 1960’s. What did walk rate do? One run strategies are… Read more »

Hartvig
Hartvig
12 years ago
Reply to  kds

That is somewhat along the lines that I have been thinking about this but far more logically thought out and better put than I was able to come up with. Both runs and home runs are down since the early 00’s. It seems at least possible that either hitters are adjusting their approach or management is looking for hitters who can put the ball into play more in a more run scarce environment. I’d have a hard time pointing to any one team or picking out a group of players to support my argument however. In fact the only team… Read more »

Mike L
Mike L
12 years ago
Reply to  Hartvig

It’s also possible that the eclipse of the steroid era is essentially a feedback loop; as home run and run scoring levels decline, the perception that the three run bomb is always around the corner may lead to a reversion to a more traditional strategy. I don’t thing this will affect the approach of true power hitters, although their overall output will be down. But it will have a perceptible impact on the rest. It’s interesting that in both 2010 and 2011,there were only two players with 40+ HR. From 1995-2006 MLB averaged more than twelve 40+ HR seasons her… Read more »

Lawrence Azrin
Lawrence Azrin
12 years ago
Reply to  Mike L

In a lower-run environment, I’m theorizing that power would be MORE valuable, and walks LESS valuable. With a lower OBP, it will be harder to string together longer sequences of walks/hits, so an individual walk would on average have less value. OTOH,
extra-base hits would score practically all the baserunners, and would be more “efficient” at scoring runs.

Neil L.
Neil L.
12 years ago
Reply to  Lawrence Azrin

So, Lawrence, in your opinion are we entering a low-run environment, are we already in it and have yet to adjust, or are the last couple of years a statistical blip offensively?

And, as a related question, is runs per game or home runs per game the better measure of a low-run environment?

Can we generate any evidence, from Play-Index data, about the 2010’s decade is headed?

Pardon me, there more questions than answers here, but they are questions I have wanted to look into.

Mike L
Mike L
12 years ago
Reply to  Lawrence Azrin

@20, Lawrence, that’s a really interesting point, but presumably you could quantify that by comparing the ratio of runs scored to runs scored on home runs over time. As to other EB hits, I’d expect that doubles would have increased during the steroid era, but the runs scored on doubles probably not at the same rate. Triples I think are too park dependent. The older, more cavernous stadiums would have been a lot more triple-friendly

Lawrence Azrin
Lawrence Azrin
12 years ago

#22/Neil L., I think that runs/game tells you more about the run environment, much more than HRs. In the 1920s/1930s, HR rates were a third or half of what they were the last 15 years, but R/G were as high or higher than recently. I don’t think we’re in a low-run environment now, more like a historically “average” level the last couple years (2010- 4.38 r/g, 2011 – 4.28). The high-scoring era ended in 2009 (4.61 r/g). I think the next few years will resemble the 1970s/1980s, about average scoring levels, but with some random fluctuations. Power, speed, and defense… Read more »

Richard Chester
Richard Chester
12 years ago
Reply to  Lawrence Azrin

Here are R/G and HR/G rates by decade. These averages do not take into account shortened seasons so the accuracy is not 100% but should be close enough to make a judgment.

The Decade….R/G….HR/G
2001-2010….4.68….1.05
1991-2000….4.76….0.99
1981-1990….4.29….0.82
1971-1980….4.14….0.73
1961-1970….4.06….0.83
1951-1960….4.39….0.85
1941-1950….4.32….0.55
1931-1940….4.85….0.55
1921-1930….4.93….0.44
1911-1920….3.98….0.18
1901-1910….3.94….0.13

Neil L.
Neil L.
12 years ago

Richard, thank you so much for the data.

The problem with using decades as the divider is that it may cut across key events in the history of baseball such as a change in the manufacture of the baseball, rule changes such as outlawing the spitter, changing the mound height or introducing the designated hitter.

There are extraneous factors such as World War 2, which took good players out of the game.

Also, expansion with its dilution of pitching, may have occurred at any time in a decade.

I think it almost requires year-by-year data to discover the bigger trends.

Lawrence Azrin
Lawrence Azrin
12 years ago

#24/ Mike L., Sorry, I’m not a big B-R P-I guy, I’llleave that to you guys that are more proficient in that area. As with many areas of baseball theory, I am just paraphrasing old BillJames writings; if it’s harder to get runners on base, I would think extra-base hits are more valuable and walks relatively less valuable. There is one HUGE difference between the 20s/30s and now; back then, maybe one ot two guys in a lineup was a HR threat; nowadays ALL BUT one or two guys is a threat to hit it out. That’s why you don’t… Read more »

Lawrence Azrin
Lawrence Azrin
12 years ago
Reply to  Lawrence Azrin

Sorry, I hit “send” before finishing. Last line:
“That’s why you don’t see lines like:
3 HR, 124 RBI or 9 HR, 119 RBI”

Neil L.
Neil L.
12 years ago
Reply to  Lawrence Azrin

Lawrence, I don’t think that it matters whether you are a big “Baseball Reference or Play Index” guy or not. Don’t be intimidated, there is a place for you here.

If you currently have eyeballs on a ML team and are willing to be objective than help the rest of us pursue a deeper understanding of baseball, either from a statistical or anecdotal perspective.

The connection between anecdotal impressions of a player and his cold, hard statistics has always intersted me.

Lawrence Azrin
Lawrence Azrin
12 years ago
Reply to  Neil L.

Neil L.,

I use B-R constantly to back up my arguments here; what I meant was that I do not use the Play Index to compile specific lists, or look for specific events. I’ll leave that to the P-I virtuosos here, such as John Autin.

I’m equally interested in dissecting oddball lists of unusual events, or arguing “who was better than who(m)??”.

For instance:
Mantle or Mays?
Kaline or Clemente?
Clemens or Maddux?
Doerr or Gordon?
Pedroia or Cano?
Moonlight Graham or Larry Yount?

Neil L.
Neil L.
12 years ago
Reply to  Neil L.

Lawrence @28, excuse me for being condescending.

I see you as being proficient in Player-Index searches.

I am laughing out loud at your provocative list of pairs? My gosh! Seriously.

Lawrence Azrin
Lawrence Azrin
12 years ago
Reply to  Neil L.

Neil L.,
That’s OK, we all bring different backgrounds and interests here.

That last pair of comparisons was just an inside joke. I do think that the other comparisons would make for some very intesting discussions.

Mike L
Mike L
12 years ago
Reply to  Lawrence Azrin

@25, Lawrence, as I have amply demonstrated on multiple prior occasions, what I know about math barely fills a tea-cup. When folk on here talk about standard deviations and Fibonacci Numbers, I think “cool site, now, what the heck did they mean by that?” I’m probably a lot more like you-I memorized a few volumes of Bill James.

Neil L.
Neil L.
12 years ago
Reply to  John Autin

JA, you were munching on cheese wheels last night.

Now nachos…. you are way too clever …. 🙂

This is a lot more like the old B-Ref free-for-all.

Lawrence Azrin
Lawrence Azrin
12 years ago
Reply to  John Autin

“Fibonachos” – In New England, they’re called “jimmy’s” or “sprinkles” at Friendly’s.

I’ll call your “Fibonacci”, and raise you a Small Sanple Size.

Neil L.
Neil L.
12 years ago
Reply to  Lawrence Azrin

All right…. pop quiz here.

Who can type the first 10 Fibonacci numbers without Google? Or do you know the formula?

Lawrence, you are revealing your math background!

Neil L.
Neil L.
12 years ago
Reply to  Mike L

I don’t know about memorizing Bill James, Mike L, but your comments in here make a lot of sense, math proficiency or not.

Mark in Sydney
Mark in Sydney
12 years ago
Reply to  Lawrence Azrin

Lawrence I have been having a but of “fun” with the new Lahman database and some stats software I have. Mostly to check for myself some of BJ’s assertions (know for yourself, right?). Some of the obvious: (1) Wins and Runs are very strongly correlated (0.594) (2) Losses and Hits are too (0.567) (3) HR Allowed correlate to Losses more so than BBs but not as much as Hits (0.567 v 0.528 v 0.314) All pretty obvious really. You want to score runs, you want to stop the opposition scoring runs. If they are going to get on base, then… Read more »

Neil L.
Neil L.
12 years ago
Reply to  Mark in Sydney

Mark, I assume BJ = Bill James. What the heck time is it in Sydney when it is 9:00 P.M. in N. America?

Mark in Sydney
Mark in Sydney
12 years ago
Reply to  Neil L.

1125 tomorrow 🙂 Tuesday morning here so we already know who won .. ha! If only

Neil L.
Neil L.
12 years ago
Reply to  Mark in Sydney

Whaddya mean? You’re on the other side of the International Dateline?

I know, I know …..

Mark in Sydney
Mark in Sydney
12 years ago
Reply to  John Autin

JA Runs to Hits = 0.836 Runs to 2B = 0.740 Runs to 3B = 0.054 Runs to HR = 0.779 Wins to Hits = 0.543 Wins to 2B = 0.312 Wins to 3B = 0.152 Wins to HR = 0.387 Not too much in it John. Doubles and homers come out about the same. Though all this really says is that most games are won, and most runs are scored, through singles rather than extra base hits. There is also a correlation of 0.572 between doubles and HRs. On average, teams recorded 252 doubles against 142 homers a year… Read more »

bstar
12 years ago
Reply to  Mark in Sydney

Mark,

Really interesting numbers, but I’m not understanding what you mean exactly by ‘most games are won and most runs are scored through singles.’ Do you have Runs to 1B and Wins to 1B numbers?

Neil L.
Neil L.
12 years ago
Reply to  Mark in Sydney

I like the numbers you have given us, Mark. The correlation coefficents between various offensive events are, to me, thought-provoking.

Bear with me, but, If am understanding your posts correctly, you are saying the frequency of an offensive event has a bearing on its correlation to winning a game?

So a triple has a lower coefficient than a double because of its relative rarity, despite the fact that a triple has more potential ro contribute to an individual run scoring than a double?

Thanks.

Mark in Sydney
Mark in Sydney
12 years ago
Reply to  Mark in Sydney

Neil, Exactly. Triples are getting rarer therefore their influence is lower. (As an aside, DPs have almost no impact on anything. Go figure.) Basically, batting is hard. At the elite level, getting a hit every 3 at-bats is very good. Doubles are harder still, requiring a ball hit to the right part of the field. Arguably, a 2B is twice as easy to achieve than a HR but has about the same impact on run scoring and wins. I guess that you can also theorize that two players, with the same total bases from the same at-bats, one a double… Read more »

Mark in Sydney
Mark in Sydney
12 years ago

Hey, bstar.

Sorry I was cryptic. Breaking down then numbers. Over the last 50 years, the average team has won 80 games a season, scoring 703 runs.

So, on a per-game basis, that average team would score 4.340 runs per game to get the win. Again, on average, they would do this by making 8.704 hits, 1.556 2B, 0.2108 3B and 0.883 HR. Basically 1HR per 9Hits.

Meaning, to me, that you need to make hits to win 🙂

bstar
12 years ago
Reply to  Mark in Sydney

Thanx for the reply, Mark, but how did you go from that conclusion to ‘most games are won and most runs scored through singles’? It’s the singles part I’m confused about, not the overall hits.

Mark in Sydney
Mark in Sydney
12 years ago
Reply to  bstar

Right. I was attempting to be funny by stating the obvious. There are about 10 singles for each homer which means that they are more important. More a play on bad statistical analysis.

bstar
12 years ago
Reply to  Mark in Sydney

I gotcha now. I get confused easily. I thought you were maybe trying to say that singles were more important than extra-base hits as a whole. Just eyeballing 2011 stats, it looks like more than 50% of all total bases are from extra-base hits, which means less than fifty percent are from walks and singles. Of course, two singles leading to runners on first and second is two total bases but a double is only one runner on second. I would like to know what percentage of runs are driven in by extra-base hits vs. walks/singles, as well as the… Read more »

Neil L.
Neil L.
12 years ago

“•In 1999, the year after McGwire and Sosa shattered the season HR record, the BB rate surged by almost 8%.”

I’m not quite ready to let this blog die yet, John.

Are you implying a cause and effect between the HR duel and the walk rate the following season? If so, I don’t quite follow. I may missing be something obvious.