Rob Bradford on the proliferation of home runs

[charlieoscar]

Bradford has an article on the WEEI site today that begins:
http://www.weei.com/articles/column/baseballs-bigger-swings-examining-why-were-middle-historic-home-run-pace

"Entering Tuesday night Major League Baseball players were hitting more home runs than ever before in its history, averaging 1.23 homers per game. That's ahead of the No. 2 most prolific season, 2000...."

And it goes on to quote players about the differences they've noticed between major and minor league balls, then follows with some statements from an MLB spokesman:

"As a quality control effort, we routinely conduct in-season and off-season testing of baseballs in conjunction with our consultants at UMass-Lowell to ensure that they meet our specifications. All recent test results have been within the specifications. In addition, we used a third-party consultant to test whether the baseball had any impact on offense in recent years, and he found no evidence of that.

"It's no secret that the balls in the Majors and the Minors have slight differences. The Minor League ball is a good ball, just like the ones used at the collegiate level are good. The Major League balls are made under a different manufacturing process and have a more rigorous quality control process. Again, balls have to meet our testing specifications (and they have)."

To which I would respond that it is also no secret by their very own words that major and minor league balls are different and that the former have more stringent quality control. There is a measurement (I want to say co-efficient of restitution but I'll just call it X) that they use to grade the baseballs. This would actually be X +/- a certain percentage. It would be simple to arrange matters so that major league balls are manufactured so the majority of them come out on the percentage side favoring liveliness but still fall within the bounds. The ones at the other end of the spectrum could be sold in souvenir shops, etc....i.e., I trust MLB about as much as I do politicians.

 

[jon abbey]

Terrible to write an article like that in 2017 without mentioning the phrases 'exit velocity' and 'launch angle'. Hitters are using the widely prevalent Statcast info to tinker with their swings and try to hit more HRs, strikeouts are also up, and I believe that singles/doubles/triples are all down, and runs scored overall are about the same.

 

[Cesar Crespo]

Terrible to write an article like that in 2017 without mentioning the phrases 'exit velocity' and 'launch angle'. Hitters are using the widely prevalent Statcast info to tinker with their swings and try to hit more HRs, strikeouts are also up, and I believe that singles/doubles/triples are all down, and runs scored overall are about the same.

Runs are up about .1 per game over last year. There should be roughly 530 more runs scored in 2017 than 2016. There were 1,000 more runs scored in 2016 than 2015. There were 900 more runs scored in 2015 than 2014. HR's are a huge reason why there is more than half a run scored in 2017 than 2014. HR per game from 2014 to 2017. 0.86>>1.01>>1.16>>1.23. Runs Per Game 4.07>>4.25>>4.48>>4.59. HR's have made up 71.2% of the increase in runs since 2014.

This reverses a trend where we saw offenses get worse from 2012 to 2014. We are going to have to start readjusting what is a good slash line again. Slash lines from 2014 to 2017. Strikes out are up 1.2% since 2014, while OPS is up about 6%. 251/.314/.386, 20.4% K rate >> .254/.317/.405, 20.4% K rate>> .255/.322/.417, 21.1% K rate >> .252/.322/.419, 21.6% K rate. Average and OBP mostly stay the same, slugging sees huge increase.

For reference: In 2000, there were 5.14 runs scored per game, slash line of .270/.345/.437 and a K rate of 16.5%. Quite different than 2017 but the same exact .167 ISO.

 

[Vermonter At Large]

Good discussion here. For me the most interesting part of all of this is that HR Runs have become a higher percentage of overall run production than ever before. Throughout the Expansion Era, we saw that HRs accounted for about 36% of overall run production and that number was very consistent from year-to-year - even during the Steroid Era. Last season, HR Runs accounted for 47% of all runs produced, and this season those numbers could be over to 50%.

Neither the ball nor the players are juiced. During past eras when that was true, we saw spikes in overall batting rates - not just in HRs, and HRs still maintained that 36% share of (higher) overall run production. These numbers point strongly in the direction of the more complex factors that have been discussed: 1. Overall hitting approach by players and coaches geared toward the long ball; 2. Roster and lineup selection favoring players with HR power over other skills; and 3.) The prevalence of defensive shifting taking away some of the in-park batting events which produced runs without the long ball. Probably it is a combination of all three factors.

 

[Wake's knuckle]

I was just reading on fangraphs (article on Justin Bour: http://www.fangraphs.com/blogs/justin-bour-same-as-he-ever-was/ ) that the seams on the ball seem to be lower this year, leading to less drag and thus more distance with the same EV. Interesting.

 

[MyDaughterLovesTomGordon]

Yeah, 538 has a similar piece on the air resistance being lower:

https://fivethirtyeight.com/features/in-mlbs-new-home-run-era-its-the-baseballs-that-are-juicing/

There's some persuasive evidence in there for the ball being a factor.

Still, Tzu-Wei Lin is exhibit A: coaches have truly embraced the Ted Williams approach of a slight uppercut (the pitcher is slightly elevated, after all, as Ted argued). That's clearly a factor.

 

[soxfan121]

Rob Bradford, carrier of Josh Beckett's water, yesterday said the "Celtics beat reporters in this town are too close to the players."

It was one of the least self-aware statements I've heard recently.

Bradford went on to praise Nick Cafardo as a hard worker who tells it like it is.

 

[Vermonter At Large]

Yeah, 538 has a similar piece on the air resistance being lower:

https://fivethirtyeight.com/features/in-mlbs-new-home-run-era-its-the-baseballs-that-are-juicing/

There's some persuasive evidence in there for the ball being a factor.

Still, Tzu-Wei Lin is exhibit A: coaches have truly embraced the Ted Williams approach of a slight uppercut (the pitcher is slightly elevated, after all, as Ted argued). That's clearly a factor.

Meh. That 538 Article was the basis for the Bour article, btw. Arthur throws out a lot of names, but the basis for his thesis in the 538 article is pretty sketchy, IMHO. First of all, the 2017 numbers are the continuation of a three-year trend of HRs increasing. Granted, the increase so far this season over last has been significant, but from Arthur's own graph, there was a significant increase in the HR rate in 2016 without any corresponding decrease in the drag coefficient. A half season is not a large enough period of time in my opinion to isolate other factors from the changes in drag coefficient, assuming that those numbers are as accurate as he claims to begin with.

More significantly, any changes to the ball that increase HR performance should also be reflected in the overall offensive performance. As I stated earlier, the increase in run production from HRs has not been matched by increases in overall run production. We should expect to see more doubles and triples as we did in the 1930's and again during the Steroid Era if the ball were significantly juiced.

 

[MyDaughterLovesTomGordon]

I see your point there, for sure.

A couple further questions:

Is the distance from the mound to home plate enough to be impacted by air resistance and the ball? Should we see corresponding increase in pitch velocity, and are we? (I think we are, for sure, but I'm not sure how you'd separate out increased performance and the ball, since velocity has been increasing steadily at least since 2000.)

If it's the ball, should we be seeing a similar DECREASE in foul outs, as more of the fouls are making the stands?

I still lean toward hitting philosophy, swing mechanics, and shifts, for sure, but I think it's interesting to use the data that Statcast provides.

 

[charlieoscar]

In 1968 Danny Cater finished second in the AL batting race with a .290 average.

What was MLB's solution to the league only batting .230? They dropped the height of the mound from 15 inches to 10 inches. Are we seeing the effect of an increase overall in height of pitchers nowadays, which essentially lowers the mound even more? If the average height of pitchers was 6'1" a few years ago and 6'4" today (to just choose a couple of numbers), this could have an effect. It might be worth a study but in my experiences of being among players on the field, I'm not sure you could get sufficient accurate measurements of heights to be conclusive.

 

[Vermonter At Large]

I see your point there, for sure.

A couple further questions:

Is the distance from the mound to home plate enough to be impacted by air resistance and the ball? Should we see corresponding increase in pitch velocity, and are we? (I think we are, for sure, but I'm not sure how you'd separate out increased performance and the ball, since velocity has been increasing steadily at least since 2000.)

.

I wondered this myself. The older pulse doppler radar guns certainly would not have been accurate enough for this sort of fine-grain analysis. From 2006-2016, MLB calculated pitch velocity using Pitch F/X which would not have been accurate enough to calculate the begin and end velocities that Arnold used to calculate his drag coefficient numbers. STATCAST, which became the primary measurement tool for pitch velocity in 2017, claims to be able to accurately measure that, but here is a link https://fivethirtyeight.com/features/baseballs-new-pitch-tracking-system-is-just-a-bit-outside/ to an article he published back in April on 538 that states that STATCAST measurements of pitch velocity were not as accurate as Pitch F/X. At this point, I am extremely skeptical of this entire analysis.

 

[charlieoscar]

...If it's the ball, should we be seeing a similar DECREASE in foul outs, as more of the fouls are making the stands?...

I just did a real quick look at Retrosheet Game Logs by Year, Event Type=2 (Generic Out according to their code), and Foul Flag="T" and the results were:

2000 -- 3926 foul outs
2005 -- 4100
2010 -- 3598
2015 -- 3287 (I did look at some adjacent years and they were similar)

Note: this does not take into consideration any changes in ball parks, new or re-design.

 

[Vermonter At Large]

I like your thinking here, but I think that foul outs are so rare and so varied in nature that the data is a bit too grainy for this type of analysis.

 

[Vermonter At Large]

In 1968 Danny Cater finished second in the AL batting race with a .290 average.

What was MLB's solution to the league only batting .230? They dropped the height of the mound from 15 inches to 10 inches. Are we seeing the effect of an increase overall in height of pitchers nowadays, which essentially lowers the mound even more? If the average height of pitchers was 6'1" a few years ago and 6'4" today (to just choose a couple of numbers), this could have an effect. It might be worth a study but in my experiences of being among players on the field, I'm not sure you could get sufficient accurate measurements of heights to be conclusive.

We had this discussion here, I believe, a few years back. The Doppler radar gun readings are definitely skewed based on the geometry, so a downward or upward (or left/right) difference in angle could produce a slightly faster or slower reading. However, since they aren't terribly accurate to begin with, those effects probably aren't consistently detectable.

 

[Sampo Gida]

Meh. That 538 Article was the basis for the Bour article, btw. Arthur throws out a lot of names, but the basis for his thesis in the 538 article is pretty sketchy, IMHO. First of all, the 2017 numbers are the continuation of a three-year trend of HRs increasing. Granted, the increase so far this season over last has been significant, but from Arthur's own graph, there was a significant increase in the HR rate in 2016 without any corresponding decrease in the drag coefficient. A half season is not a large enough period of time in my opinion to isolate other factors from the changes in drag coefficient, assuming that those numbers are as accurate as he claims to begin with.

More significantly, any changes to the ball that increase HR performance should also be reflected in the overall offensive performance. As I stated earlier, the increase in run production from HRs has not been matched by increases in overall run production. We should expect to see more doubles and triples as we did in the 1930's and again during the Steroid Era if the ball were significantly juiced.

The study by MGL and Lindbergh looked at and tested balls from 2016 and 2014/pre Asb 2015) . The 2017 balls were not tested.

https://theringer.com/2017-mlb-home-run-spike-juiced-ball-testing-reveal-155cd21108bc

HR rates began to increase in the 2nd half of 2015 along with an increase of 1 mph in EV due to higher COR along with reduced drag from lower seams. Together the effect is equivalent to an increase in EV of 1.5 mph on the pre 2015 balls which Alan Nathan said would be enough to cause current HR rates.

As to your doubles and triples argument. Some of thats due to short fences and smaller parks relativevto the 1930's. Lot more ground rule doubles. However, since 2014 Singles have dropped 5.5% per game and XBH of all kinds have increased 16.6% per game. The lack of a corresponding run production may be due to a 7% increase in K Rate from an already historic high now.

So we have historic HR rates exceeding anything seen in the steroid era which may also been due to ball changes matched by historic K rates that partially offset the runs produced from increased HR production. BABIP despite all the shifting , smaller parks and emphasis on defense remains unchanged and at or close to historic highs

The evidence seems clearly to point to the ball.

 

[charlieoscar]

The study by MGL and Lindbergh looked at and tested balls from 2016 and 2014/pre Asb 2015) . The 2017 balls were not tested.

I mentioned this once before but I think it worth repeating. A baseball must meet certain measured required measurements, plus or minus a certain degree. There is nothing that I know of that prevents MLB from choosing those balls which are more "hitter-friendly" but still within the required parameters. For example, a ball that is wound more tightly will probably have a diameter that is slightly less than the specified exact diameter, but still within that specification. I would think a more tightly wound ball to be more lively. Also, if MLB is supplying balls to labs for checking there is no reason they couldn't send a sample from the full spectrum. I have no way of proving this but does anyone have a way of disproving it?

 

[Vermonter At Large]

The study by MGL and Lindbergh looked at and tested balls from 2016 and 2014/pre Asb 2015) . The 2017 balls were not tested.

https://theringer.com/2017-mlb-home-run-spike-juiced-ball-testing-reveal-155cd21108bc

HR rates began to increase in the 2nd half of 2015 along with an increase of 1 mph in EV due to higher COR along with reduced drag from lower seams. Together the effect is equivalent to an increase in EV of 1.5 mph on the pre 2015 balls which Alan Nathan said would be enough to cause current HR rates.

As to your doubles and triples argument. Some of thats due to short fences and smaller parks relativevto the 1930's. Lot more ground rule doubles. However, since 2014 Singles have dropped 5.5% per game and XBH of all kinds have increased 16.6% per game. The lack of a corresponding run production may be due to a 7% increase in K Rate from an already historic high now.

So we have historic HR rates exceeding anything seen in the steroid era which may also been due to ball changes matched by historic K rates that partially offset the runs produced from increased HR production. BABIP despite all the shifting , smaller parks and emphasis on defense remains unchanged and at or close to historic highs

The evidence seems clearly to point to the ball.

The evidence does not clearly point to any single factor.

I think I've been pretty nice in suggesting that Arthur's piece on drag coefficient was probably flawed. I don't know him personally and have only read a few of his pieces so I refrained from what might have been harsher criticism of that work. It was certainly an ambitious and creative approach, but my main problem is with the input data and with the sensational claims that may or may not have been his alone.

Exit Velocities and Launch Angles are nice tools, but again, we don't really know if they are accurate enough to make the kinds of assumptions that are being put forth and we can only go back a few years with these tools. I can't really speak to the quality of the work in the ball analysis, but MGL and Lindbergh are well-respected folks in the field. Again, it would be nice to have a larger body of ball test analysis to compare it to, and there is no anecdotal evidence that the manufacture of the balls has changed significantly since 2015 and suggesting that MLB is tinkering with the balls is a conspiracy theory worthy of Alex Jones.

The hit rate data you and I and others have provided does not support any single factor. The ball may or may not be a factor, but it certainly isn't the only or even probably the main factor in the HR boon. Certainly, the influx of bigger, stronger guys like Thames and Judge, the use of STATCAST data to optimize HR strokes, the prevalence of defensive shifts, and the overall approach to hitting are also part of the equation.

 

[charlieoscar]

This is a breakdown of slugging percentage by hit type (1B/2B/3B/HR) for 2010 through now in 2017. The overall SLG is in the last column. As can be seen, there is a huge jump in the portion of SLG derived from home runs starting in 2015 compared to 2014, which was the low mark of the eight years covered. Doubles are slightly higher and triples slightly lower.

Year SLG1 SLG2 SLG3 SLG4 SLG
---------------------------------------
2010 .173 .103 .016 .112 .403
2011 .171 .101 .016 .110 .399
2012 .169 .100 .017 .119 .405
2013 .171 .099 .014 .112 .396
2014 .172 .098 .015 .101 .386
2015 .169 .100 .017 .119 .405
2016 .166 .100 .016 .136 .417
2017 .162 .102 .014 .148 .425

 

[grimshaw]

Eno Sarris attempted a projection of home run leaders based on last year's vs this year's baseballs.
http://www.fangraphs.com/blogs/power-leaderboards-with-the-old-ball/#more-258788

He based it on the average increase of exit velocity since May 2015 being 1.02 mph and noted that players on average are hitting the ball 5.6 feet further.

What would be interesting is how much that translates to an increase to BABIP. It's already increasing due to all the home runs and strikeouts, but curious how it affects singles, doubles and triples. I would think a mid 300 BABIP nowadays wouldn't necessarily mean a big regression is coming relative to a few years ago.