Runs Scored and K's are Correlated

[LouGehrig]

A strong correlation exists between N.L. strikeouts and run production from 1950-2008. Does this mean that the more a team strikes out the more runs it scores?

http://baseball.suite101.com/article...are_correlated

[SamtheBravesFan]

I didn't look at the article, but the idea does seem plausible, because a team that strikes out a lot tends to swing at pitches a lot. If that is the case, surely they'd score a few runs. ;P Maybe looking at walks would help too?

[RuthMayBond]

A strong correlation exists between N.L. strikeouts and run production from 1950-2008. Does this mean that the more a team strikes out the more runs it scores?

http://baseball.suite101.com/article...are_correlated

If that was true, I'd be revered more than the Babe AND Cobb put together I once struck out in the on-deck circle, after the game had ended

[LouGehrig]

Maybe looking at walks would help too?

It would and I might. It is time consuming to set up the raw data.

[Honus Wagner Rules]

If that was true, I'd be revered more than the Babe AND Cobb put together I once struck out in the on-deck circle, after the game had ended

Did you strike out swinging or on a called third strike? It makes a big difference in how we revere you.

[The Dude]

After reading the article, I agree it looks as if there's a strong correlation between league totals for strikeouts and runs scored.

But the writer slips up when he states:

Does this mean that the more a team strikes out the more run it scores? It sure does, but this is the problem. Data must be interpreted, and the interpretations must withstand intense scrutiny. Teams that hit home runs usually strike out a lot but score a lot of runs.

Sure, he says that for the deadball era this is not true, but he basically states it is true for modern baseball. I don't comprehend standard deviations and the like, so I figured I'd tackle this issue with a little bit of 8th grade math.

I wanted to keep my sample size consistent, so I decided to look at the NL from 1998-2008, which would give me 16 teams every year.

I made lists for each season, one with the teams ranked 1-16 in runs scored, and 1-16 in K's. I then proceeded to look at the top 5 in Runs scored, and where they sat in the rankings of the 1-16 in K's. THis is what I got:

There were 5 teams in each season that were in the top 5 in Runs scored. 
   5
x 11
----
  55 Total teams that were in the top 5 in their league in runs scored. 
This would also mean there were 55 teams in the league that were 
in the top 5 in strikeouts. 

There were only 17 teams that were in the top 5 of their league in 
runs scored, and in strikeouts. That equates to .31 percent of the 
teams who led their league in runs scored, also lead their league in 
strikeouts.

So we can see that the author was wrong when he stated that modern day teams that score the most runs also strike out the most. In fact, there was only one instance where the top team in runs scored was the top team in strikeouts:
2005-Reds

However, it occurred 3 times that the team with the most runs, struck out the least:
2000-Rockies
2001-Rockies
2003-Braves

And, there were 3 instances were the team that struck out the least was 2nd in their league in runs scored:
1999-Rockies
2002-Cardinals
2004-Giants

So now that I'd proven the teams that scored the most hardly struck out the most, I thought I'd check the teams that scored the least.

There were 5 teams in each season that were in the bottom 5 in Runs scored. 
   5
x 11
----
  55 Total teams that were in the bottom 5 in their league in runs scored. 
This would also mean there were 55 teams in the league that were in 
the top 5 in strikeouts. 

There were only 19 teams that were in the bottom 5 of their league in 
runs scored, and the top 5 in strikeouts. That equates to .35 percent
 of the teams who were at the bottom of their league in runs scored,
 and in the top of the league in strikeouts.

If it is the case that a higher percentage of teams in the bottom 5 of their league in runs scored, are also in the top 5 of their league in strikeouts than the percentage of teams in the top 5 of their league in runs scored and also in the top 5 of their league in strikeouts, then it is not the case that teams who strike out more score more runs.

[csh19792001]

First, wouldn't this be much more appropriate for the stats forum?

If it is the case that a higher percentage of teams in the bottom 5 of their league in runs scored, are also in the top 5 of their league in strikeouts than the percentage of teams in the top 5 of their league in runs scored and also in the top 5 of their league in strikeouts, then it is not the case that teams who strike out more score more runs.

Exactly. Anyone who states that higher K totals (in and of themselves) lead to higher run totals, lacks even a rudimentary background in research and/or statistics.

Why? They're not controlling for extraneous and/or confounding variables. This is a false positive and leads to the assumption causal (but not correlative) relationship. You can't infer that B results from A without considering everything that contributes to the impact of variables C,D,E, etc....

With today's conditions (namely, everything being setup to favor sluggers and home runs), vastly more players are going to benefit from the all or nothing approach to the plate. HR hitters tend to K more, and with these conditions, teams with more HR hitters tend to score more runs.

Incidentally, in the long run, K's are almost 10% worse in very low run scoring environments. My gripe with player who K a ton has little to do with long run statistics- or statistics in general.

[Honus Wagner Rules]

Incidentally, in the long run, K's are almost 10% worse in very low run scoring environments. My gripe with player who K a ton has little to do with long run statistics- or statistics in general.

Is it more of an aesthetic or stylistic issue?

[csh19792001]

Is it more of an aesthetic or stylistic issue?

It's a skill issue, and part style/aesthetics. It all comes back to greatness vs. value.

I think DiMaggio is one of the greatest hitters ever in large part because of his HR/K ratio. Williams is in the vicinity of Ruth as a hitter, IMO, in part because of his ridiculous BB/K ratio. Speaker was clearly one of the greatest hitters ever because he couldn't be struck out despite being one of the top few power hitters in baseball.

To me, Adam Dunn will never be a truly great hitter even if he is very valuable on paper. A tremendous amount of K's indicates some combination of selfishness, impatience, and a general lack of diversity of skills at the plate. Singles and doubles shouldn't be mishit attempts at homers. Going up trying to pull the ball every time for a homer is artless and it isn't real hitting.

[1905 Giants]

It's a skill issue, and part style/aesthetics. It all comes back to greatness vs. value.

I think DiMaggio is one of the greatest hitters ever in large part because of his HR/K ratio. Williams is in the vicinity of Ruth as a hitter, IMO, in part because of his ridiculous BB/K ratio. Speaker was clearly one of the greatest hitters ever because he couldn't be struck out despite being one of the top few power hitters in baseball.

To me, Adam Dunn will never be a truly great hitter even if he is very valuable on paper. A tremendous amount of K's indicates some combination of selfishness, impatience, and a general lack of diversity of skills at the plate. Singles and doubles shouldn't be mishit attempts at homers. Going up trying to pull the ball every time for a homer is artless and it isn't real hitting.

What about guys who routinely ground into double plays?

Would it be better to have a guy who struck otu not a ludicrous amount of times but a high amount of times, or a guy who routinely grounded into double plays but didn't strike out much?

Just curious.

[Sultan_1895-1948]

You cannot get a hit if you don't make contact. If, when making contact, you happen to hit into a double play, then so be it. But first thing is first. Giving yourself the chance for a hit. A guy like Ripken grounded into about 20 double plays a year. As many as Cecil Fielder. I'll take that. I'll take him making contact and letting a hit be possible.

Adam Dunn-----------BA--------K------PA

Man on 2nd----------.184-----105-----422
Men on 2nd&3rd------.235-----23------93
Man on 3rd lt 2 outs--.265-----57------230
Man on 3rd, 2 outs---.188-----49------205

[LouGehrig]

After reading the article, I agree it looks as if there's a strong correlation between league totals for strikeouts and runs scored.

But the writer slips up when he states:



Sure, he says that for the deadball era this is not true, but he basically states it is true for modern baseball. I don't comprehend standard deviations and the like, so I figured I'd tackle this issue with a little bit of 8th grade math.

I wanted to keep my sample size consistent, so I decided to look at the NL from 1998-2008, which would give me 16 teams every year.

I made lists for each season, one with the teams ranked 1-16 in runs scored, and 1-16 in K's. I then proceeded to look at the top 5 in Runs scored, and where they sat in the rankings of the 1-16 in K's. THis is what I got:

There were 5 teams in each season that were in the top 5 in Runs scored. 
   5
x 11
----
  55 Total teams that were in the top 5 in their league in runs scored. 
This would also mean there were 55 teams in the league that were 
in the top 5 in strikeouts. 

There were only 17 teams that were in the top 5 of their league in 
runs scored, and in strikeouts. That equates to .31 percent of the 
teams who led their league in runs scored, also lead their league in 
strikeouts.

So we can see that the author was wrong when he stated that modern day teams that score the most runs also strike out the most. In fact, there was only one instance where the top team in runs scored was the top team in strikeouts:
2005-Reds

However, it occurred 3 times that the team with the most runs, struck out the least:
2000-Rockies
2001-Rockies
2003-Braves

And, there were 3 instances were the team that struck out the least was 2nd in their league in runs scored:
1999-Rockies
2002-Cardinals
2004-Giants

So now that I'd proven the teams that scored the most hardly struck out the most, I thought I'd check the teams that scored the least.

There were 5 teams in each season that were in the bottom 5 in Runs scored. 
   5
x 11
----
  55 Total teams that were in the bottom 5 in their league in runs scored. 
This would also mean there were 55 teams in the league that were in 
the top 5 in strikeouts. 

There were only 19 teams that were in the bottom 5 of their league in 
runs scored, and the top 5 in strikeouts. That equates to .35 percent
 of the teams who were at the bottom of their league in runs scored,
 and in the top of the league in strikeouts.

If it is the case that a higher percentage of teams in the bottom 5 of their league in runs scored, are also in the top 5 of their league in strikeouts than the percentage of teams in the top 5 of their league in runs scored and also in the top 5 of their league in strikeouts, then it is not the case that teams who strike out more score more runs.

This is the author and your post is fantastic -- well, maybe not fantastic, but great. That is just my point at the beginning of the article. The statisticians can do what they want with the data, and often they do just that. I love your eighth grade math, because it can be understood.

We can try to correlate anything. The problem becomes cause and effect. Two things can vary directly but may be unrelated.

It is ridiculous to think that merely because strikeouts and runs scored are higly correlated, strikeouts produce runs (I'm playing with words.).

Your analysis is great because it exposed the fallacy that teams that strike out a lot score a lot of runs.

[csh19792001]

What about guys who routinely ground into double plays?

Would it be better to have a guy who struck otu not a ludicrous amount of times but a high amount of times, or a guy who routinely grounded into double plays but didn't strike out much?

Just curious.

What about them?

A guy like Joe Torre and Ernie Lombardi were rally killers, sure.

But at least they're putting the ball in play. Giving their team a chance, cutting down with two strikes, taking the ball the other way on the outside pitch.

Double plays always score guys from third. IIRC, the Yankees won the 1960 World Series on a GIDP- only run scored in the game. K's produce nothing, at least a GIDP has the potential to produce something.

Great major league hitters adjust and try to hit the ball to the right side with a man on 2nd and less than 2 outs. They try to hit the ball into the air with a player on third and less and 2 outs. There are productive, adaptive hitters and productive outs. Adam Dunn types produce a HR every 18 AB's and not much else. All those walks, but clog the basepaths once they're on, thus largely devaluing their own bases on balls.

In any case, GIDP's are much more rare than K's, which negates the comparison anyway. Far more outs in today's game are attributable to K's than GIDP's.

[csh19792001]

You cannot get a hit if you don't make contact. If, when making contact, you happen to hit into a double play, then so be it. But first thing is first. Giving yourself the chance for a hit. A guy like Ripken grounded into about 20 double plays a year. As many as Cecil Fielder. I'll take that. I'll take him making contact and letting a hit be possible.

Adam Dunn-----------BA--------K------PA

Man on 2nd----------.184-----105-----422
Men on 2nd&3rd------.235-----23------93
Man on 3rd lt 2 outs--.265-----57------230
Man on 3rd, 2 outs---.188-----49------205

Now I see that you beat me to it, Ran.

[Wade8813]

Yes, there are all those caveats (which the author mentioned, but maybe should have gone into more detail on). But still, a "Pearson's r" of 0.97 seems RIDICULOUS. From what I've heard, it virtually doesn't happen. I'm not saying the author's wrong, since I have no idea, but it'd be interesting to have verified. And if it is accurate, I'd love to see research looking more into it.