Oh boy, here we go...

Global warming is or isn't happening, and anyone who disagrees should be killed. Slowly.

PETCO, here we come!

Oh boy, here we go...

No, we don't.

Yeah. No. We don't.

Somehow, this made me think of this:

You may be interested to know that global warming, earthquakes, hurricanes, and other natural disasters are a direct effect of the shrinking numbers of Pirates since the 1800s. For your interest, I have included a graph of the approximate number of pirates versus the average global temperature over the last 200 years. As you can see, there is a statistically significant inverse relationship between pirates and global temperature.

Even the graphs FTFA are similar. Stunning!

The article is actually quite interesting. The linearity of the plot of home run distance vs temperature is a cool finding. (I'd like to see Nathan confirm that this isn't an artifact of park effect - if there is any correlation between fence distance and climate, that could create a similar effect. I expect it's a climate effect rather than a park effect, but I think it is worth checking.)

Whatever you think of the theory of anthropogenic global warming, it certainly is true that the last two decades in America have both been hotter than previous decades. They have also featured a higher-offense and much more homeriffic style of baseball. As Nathan further shows, the temperature effect on home runs is way, way too small to be the cause of the run explosion of the 90s.

I remain mostly convinced it's the baseballs. The effect doesn't take the shape you'd expect if the environmental explanations were correct (ie, climate, steroids, opposite field home runs, parks). It's a two-year spike in 93-94, a plateau, and a two-year plummet in 2010-2011. The baseballs seem like the best explanation.

I came here hoping Nathan was asserting that home runs were causing global warming, but at least TWWMH salvaged this debacle at #5. Well played, sir.

I hope everyone knows that it's illegal to connect Carolina League home runs to global warming. If you've got the data, please destroy it.

The last 5-10 years, homerun rates have dropped. Temperatures have not. Any coincidental link is history.

I agree with 6.
Personally I think the main statistical/playing effect of the Ped usage in the 80s-00s has been on career curves/shapes.

We've seen longer careers/primes than before, we've seen sharper peaks and/or misplaced peaks and/or both peaks...
Of course there have always been players who have had sharp peaks, or primes at the "wrong" times, but when before we had maybe 1/10 do something we now have 2/10... maybe 1 of those guys did PEDs, maybe not...

The homeruns? I have no doubt that there were effects on the individual level, but the simultaneous league-wide shifts suggest something more universal- such as the baseballs themselves...


Of course it could be some random combination of all of these:
the baseballs
umpiring- both change in individual personnel and directives given to umpires as whole
park changes
weather
PEDs/nutrition
training
scouting (if you look for 3 true outcomes guys you will find them, if you look for Juan Pierres/Ichiros, you will find them too)
the bats (trees used, shape, weight, weight distribution, etc)
medical advances (lasix...)

I'm sure I'm forgetting some possible factors as well

I'm sure I'm forgetting some possible factors as well

One that I always found interesting was Bill James' suggestion - based on an off-the-cuff observation by Greg Maddux - that hitters changed their approach on outside pitches. The theory, which is at least plausible, runs that hitters in amateur leagues with aluminum bats learned to crowd the plate and hit outside pitches for power to the opposite field. This turned out to be a strategy that also worked in the big leagues with wood bats, and opposite-field power increased significantly as guys learned the new strategy.

I'll leave the climate change topic be, but note that the most obvious culprit in the spike in HRs in the 1990s were magic, invisible flying monkeys.

If we're investigating global warming and its effect on baseball (or lack thereof), we need to study the comparative movement of glaciers over the past 25 to 30 years. I'd suggest looking at the baserunning of Cecil and Prince Fielder.

I'd like to see a study of HRs correlated with this dataset:

Is music getting more profane?

EDIT: For TLDR people the answer to the question is yes.

The article is actually quite interesting. The linearity of the plot of home run distance vs temperature is a cool finding. (I'd like to see Nathan confirm that this isn't an artifact of park effect - if there is any correlation between fence distance and climate, that could create a similar effect. I expect it's a climate effect rather than a park effect, but I think it is worth checking.)

This is pretty much what I thought. Very interesting article and it's amazing just how linear the data actually is. Like most Alan Nathan articles, worth a read.

While there's a correlation between higher temperatures and more HRs for sure, I'm thinking that global warming had less to do with it than ROIDS.

1980 is the year McCarver retired. Interesting....

Yeah, you'd think there would be a cooling effect with all that hot gas no longer expelled at the ballpark.

It's a two-year spike in 93-94, a plateau, and a two-year plummet in 2010-2011. The baseballs seem like the best explanation.

It would certainly make sense, but my only problem with this explantion has always been that it would seem to be the easiest theory to prove or disprove, and as far as I've heard it never has been. I mean, it's been 19 years now since offense first jumped in 1993 and there still hasn't been any definitive study convincingly proving that the ball was "juiced." (or has there?)

I've always thought the main cause of the sillyball era may have been something as simple as hitting talent far outdistancing pitching talent for a 15 or so year stretch. Even adjusting for era, there were A LOT of great hitters peaking from 1993-2007. There were several all time great pitchers too, but the AVERAGE pitcher seemed to be worse than normal during this time. It's been balancing out the last few years as overall pitching quality has risen and the number of great hitters has decreased a bit.

But that's all just from observation. I haven't compared numbers or anything, so I could be dead wrong.

Have there been any changes in the coefficient of restitution for baseballs and also bats in this period? I believe that UMass-Lowell does testing.

Where is the temperature measured? I remember that on really hot days at Riverfront, the announcers would say how hot it was on the turf, which added a few degrees. I wonder if the fact that virtually every stadium now has grass changes things.

For anyone who won't read the article, the gist is that they found that 1 degree F of temperature change increases home runs by about 0.6%. In the last 10 years, mean temperatures in the US have been, what, something like 2 degrees F above long-term averages? So that would mean the higher temperatures have increased home runs by 1.2% (assuming the change in mean is evenly distributed across MLB stadiums, which I'm sure it isn't).

It would certainly make sense, but my only problem with this explanation has always been that it would seem to be the easiest theory to prove or disprove, and as far as I've heard it never has been. I mean, it's been 19 years now since offense first jumped in 1993 and there still hasn't been any definitive study convincingly proving that the ball was "juiced." (or has there?)

I don't think it requires the balls to be "juiced", precisely. Rawlings made MLB baseballs in Haiti until they moved the operation to Costa Rica in 1990. My understanding is that the baseballs made in Haiti were handwoven, while the Costa Rican baseballs were machine woven. The argument (and pitcher I can't remember which pitcher I first saw make it) is that the Haiti-manufactured baseballs were more variable, some softer, some harder. The Costa Rican ones are all on the harder end of the official tolerances. This means that the ball isn't necessarily "juiced" -- it's still within spec -- but because it's at the high end of spec it's still more hitter-friendly than the old ball usually was, and there aren't softer baseballs that the pitcher can covet and try to keep in play as long as possible.

So the timeline would be: the Costa Rican balls came into use in 1991, it takes the hitters a year or two to start changing their approach, then you have the high post-1993 offense. I have no idea if this is correct or not, but I think there's something interesting in the narrative.

Read the article and checked out the link. The 1880-on graph of temps vs HR was interesting. However, the fit was pretty bad from 1940 thru the early 1960s. HRs lagged in the early 40s, but that's probably WW2. But HRs were about 20% above the mean with temps near/slightly below the avg during the 1950s-early 60s, so something else was at work. Not a good fit the past several years, either.

One more skunk in the woodpile: The warming hasn't been seasonally even, but has been greater in winter than in summer. If so, the temp change component (assuming there is any) would be even lower than shown in the article. IIRC, the largest increase has been in Feb/Mar.

I hope everyone knows that it's illegal to connect Carolina League home runs to global warming. If you've got the data, please destroy it.

This deserves to be acknowledged as RDF.

So the timeline would be: the Costa Rican balls came into use in 1991, it takes the hitters a year or two to start changing their approach, then you have the high post-1993 offense. I have no idea if this is correct or not, but I think there's something interesting in the narrative.

The only part of this that wouldn't seem to fit is why offense started to go back down again in 2008 (or whenever). Did they start manufacturing MLB balls in Haiti (or elsewhere) again? If the balls were the largest factor for the offensive explosion and they haven't changed them back to the old ones, I don't see why the sillyball era was a temporary thing rather than ongoing.

What a great article! (no seriously). Although I think using 1degC/30 years is not necessarily the right number. You want the average increase in daily North American temperatures between 1980-2011 in April-Sept. Actually... he should be able to extrapolate this from actual game time temperatures and times etc. Actually you wouldn't have to extrapolate. I am pretty sure you can get T and Humidity reports hourly for major US cities going back 30 years.

Not worth the bother, at most it would double the effect (and as @23 points out, may even be in the other direction) so there are still 25x increase in HRs to go. Another effect is that he should really be looking at HR/FB rates not raw HR rates (again not likely to influence main conclusion).