Sixty Feet from Home

Maffie: Right Down K Street

The number of runs scored in Major League games has declined dramatically over the past few years. Runs/Team/Game peaked at 5.14 in the year 2000. In 2013, that number had steadily fallen, all the way to 4.13. That’s a 20% decline in less that 15 seasons.

This changing run environment has led people to say we have entered the Age of the Pitcher. But is that an accurate characterization of what’s going on?

First, let’s look at the decline in runs being scored.

Fewer Long Balls to Dig

Baseball is such a big game it’s easy to lose track of the small things. With thousands of games each year, sometimes we lose track of the big things, too. Take home runs as an example. During the mid-90s to early 2000s, it was commonplace for hitters to launch forty or more home runs out of the ballpark each year. What has happened to the forty home run hitter since 1990?


Year Home Runs HR-2000 Total 40+ HR Hitters
1990 3317 -2375 2
1991 3383 -2309 2
1992 3038 -2654 2
1993 4029 -1663 5
1994 3306 -2386 2
1995 4081 -1611 4
1996 4962 -730 17
1997 4639 -1053 12
1998 5063 -629 13
1999 5528 -164 13
2000 5692 0 16
2001 5458 -234 12
2002 5059 -633 8
2003 5207 -485 9
2004 5451 -241 9
2005 5017 -675 9
2006 5386 -306 11
2007 4957 -735 5
2008 4878 -814 2
2009 5042 -650 5
2010 4613 -1079 2
2011 4552 -1140 2
2012 4934 -758 5
2013 4661 -1031 2

During the height of the home run craze, almost half the teams in baseball had a forty home run hitter. Last year, only two did. There is a line in the film The Usual Suspects where Kevin Spacey describes the disappearance crime lord Keyser Soze by saying, “And like that…he’s gone.” You could almost say the same for the 40 home run hitter.

The middle two columns show the yearly home run total and the difference between each year and the height of the home run era (2000).  Since 2000, the league has witnessed a steady decline — 20% — in the number of home runs.

This decline brings us to the big question: Is baseball witnessing a normal rebalancing to a pre-steroid era or is there a more fundamental shift in the balance between pitching and hitting?

Other Statistics

The changing run environment goes beyond the decline of home runs. Since 1990, the three seasons of 2011-2013 saw the lowest league OBP (.321, .319, .318), the lowest batting average (.255, .255, .253), the highest strikeout percentage (18.6%, 19.8%, 19.9%) and the lowest walk-rate (8.10%, 8.0%, 7.90%).  The last three years have been the worst offensive years since the end of the Cold War.

Given those underlying trends, it’s not surprising that from 2007 to 2013, league FIP fell from 4.47 to 3.87.

From 1993 to 2013, the number of runs scored league-wide fell from 20,862 to 20,255 – and that includes there being two more teams playing in 2013 compared to 1993. The number of runs scored per team has varied from year to year, mostly due to luck-based variation in BABIP. But the general trend is clearly down. I’d like to see a few more years of data to determine if the trends in BA, OBP, BB% and K% continue to put downward pressure on runs/team, or if something else is keeping runs/team relatively constant.

With a stronger steroid testing regime, it was expected there would be fewer runs scored than during the height of the home run craze. But is relative PED-cleanliness the explanation for why batting average and on base percentage are falling? What happened to Moneyball?

Beyond smaller biceps, there are many factors that contribute to this result, including a larger strike zone, smarter defenses, and better bullpen management. But I want to focus on one trend that is not getting much attention: Pitchers are just throwing harder than they used to.

Fastball Velocity

Thanks to the PitchF/x system installed in all MLB stadiums, we are now able to track the speed, rotations, and movement of every pitch in every game. This system started in 2007, so we can’t go back to the early 90s. Yet, here is the average fastball velocity from 2007-2014:

Season FBv
2007 90.3
2008 90.7
2009 91.2
2010 91.2
2011 91.5
2012 91.6
2013 91.7
2014 91.6

That’s a pretty impressive trend. But who, exactly, is throwing harder?

Here are the numbers for starters:

Season FBv K/9
2007 89.8 6.3
2008 90.3 6.45
2009 90.8 6.65
2010 90.7 6.77
2011 91 6.75
2012 91 7.14
2013 91.3 7.19
2014 91.3 7.72

And here are the numbers for relievers:

Year FBv K/9
2007 89.5 6.94
2008 91.7 7.97
2009 91.7 7.3
2010 92.1 7.84
2011 92.8 7.67
2012 93.6 9.9
2013 94 9.36

Starters have picked up a little bit of giddy-up on their fastball, but the big change has come from relievers. In just seven years, relievers have picked up almost five MPH on their fastball and their average K/9 has skyrocketed to 9.36.

For reference, Stephen Strasberg posted a 9.39 K/9 in 2013.

We hear all the time that there are finesse pitchers who can strikeout batters based on their experience and control. That certainly is true, but it doesn’t rule out the argument that fastball velocity plays a very large part in determining a pitchers strikeout rate. Dave Cameron examined the relationship between fastball velocity and strikeouts and found an R^2 of .23. That means that 23% of what you need to know about a pitcher’s strikeout rate is given to you by a radar gun. This is a very large effect, and as Cameron argues at the end of the article, probably the single largest factor that is in control of the pitcher.

[Cameron's study doesn't include the possibility that there is a non-linear relationship between K% and vFB. If there is a non-linear relationship, then his findings would understate how important vFB is to strikeouts].

So how important are strikeouts at preventing runs? One article shows that K% was comparable to more complex statistics (FIP, xFIP, SIERA, etc.) at predicting how many runs a pitcher would give up in the future. In other words, strikeout rate appears to be the most important component at explaining how many runs a pitcher will give up in the future.

Take these two together and you start to see a strategy: the single most important statistic to explain how many runs a pitcher will give up is the number of times they record an out without the ball being hit into play. The single biggest factor in determining strikeouts is how hard you throw a baseball.

It looks like MLB front offices have noticed this relationship and are stocking their bullpens with flamethrowers.

So while it is accurate to say that the balance of power has shifted from home plate to the mound, it might be more accurate to say that the real power comes out of the bullpen. And that power is clocked at 94 MPH.

36 thoughts on “Maffie: Right Down K Street

  1. Great article. Nice research.

    Those statistics are truly amazing to me. The entire velocity gap between starters and relievers has occurred in the past six seasons. In 2007, starter velocity was actually even slightly higher than reliever velocity.

    That’s a sea change for the way organizations populate the bullpen.

    There are a number of explanations for the decline in offense in the past few years. Obviously stricter PED enforcement (and doesn’t all this data prove how much more PEDs benefitted hitters compared to pitchers) is part of it. It also wouldn’t be surprising if all the defensive shifting didn’t contribute to suppressing BABIP, as the study Mike cites above claims.

    To bring this back to the Reds – adding a couple MPH to his fastball probably has helped Homer Bailey hit that next level the past couple years. Chapman excepted (and that’s a big exception), the Reds bullpen runs against this grain. Broxton, Marshall and LeCure are not power arms. Broxton once was, but he isn’t now.

    • Cameron talks about K% and control in his article. He implies that the weak relationship between K% and vFB is due to harder throwing pitchers lacking the control they need to get a third strike. Ondrusek is a good example of this. Aging curves for pitchers show they gradually lose steam as they progress in their career — and this drop off (for starters) is pretty early in life. Those younger pitchers may not develop the control they need until later in their career — further complicating the vFB and K%.

      There are reports that young pitchers’ arms are being taken care of better than before, and due to this, they are able to develop a better fastball. This is an interesting theory and something I hope new data emerge on in the coming years.

  2. Along the lines of tracking how many 40+ HR guys there are in the league at any given time, I’d be curious to also track how many guys there are that can throw 100+ mph. I know back in 2000, there was Randy Johnson, Billy Wagner, and….. Is that it? Now, we have Verlander, Rosenthal, Chapman, Kimbrel, Strasburg, and probably more I’m forgetting.

    I remember when a guy throwing 100+ was a true rarity, but now it seems like quite a few teams have one, just like quite a few teams had 40+ HR guys back in the day.

    • Oh yeah, and Zumaya and Robb Nenn. Just remembered them. And didn’t Broxton used to throw pretty hard too? I seem to recall him topping 100mph…. Am I just imagining that?

      • I think you’re right about Broxton. His *average* velocity in 2009 was 97.5, so he almost certainly got above 100 mph regularly. Boy, that sure shows how he’s a completely different pitcher now. Even from last year (94 mph) to this year (92 mph), he’s dropped off. Although for the 2014 number you have to factor in that average velocities increase for many pitchers as the year goes on.

      • You are correct — I seem to remember Broxton hitting triple digits in the All-Star game awhile back (I think the echo’s Steve’s point about Broxton being a totally different pitcher now).

        • Yeah, and at the same time I remember the WLB’s having a guy in their pen who threw even harder than Broxton. Name slips my mind. I don’t even think he’s in the league anymore.

        • I can’t remember that guy’s name either. The Cardinals are certainly Exhibit A for loading their pen with power arms. They converted Trevor Rosenthal from a position player to reliever. Their pen is loaded with guys who throw 95-mph-plus.

    • I suppose you could use PitchF/x data to identify all the 100+ mph pitches since 2007 and see how many different pitchers were in the data. No rigorous way to know before that date, though. You could go by anecdotal reports in the press about speed gun readings, but that’s hit or miss, so to speak. Interesting question.

    • The max fastball velocity is an interesting stat because very few pitchers throw as hard as they can on their pitches. The numbers reported in the table are all average fastball velocities — which means that MANY fastballs are above those numbers. Straussberg has an average fastball velocity around 94 even though he can hit triple digits (same with Verlander).

      There was that argument that steroid hitters were just trying to keep up with juicing pitchers. The numbers don’t seem to vindicate that argument.

      • Broxton had been sitting at 91/92 yesterday, but when he was down to his last pitch he got it up to 94 (I think it was fouled off).

      • Baseball Savant (PitchF/x data) shows Ventura as the only pitcher to be at or over 100 mph this year (4 times) assuming I did the search right.

      • 17 pitchers in 2013 were recorded as throwing 100 mph or more. Chapman had the most instances (215) of anyone. Bruce Rondon had the highest percent of pitches at/over 100 mph (25.6%). Chapman’s percent (19%) was his highest since 2010 (34%!).

  3. My guess is that the relationship between velocity and strikeouts is non-linear. Is the difference between a 92 and 95 mph fastball the same as the difference between 95 and 98? or 98 and 101? or 101 and 104?

    The physical aspect of it might be linear (although if reflex-quickness is distributed on a bell curve, proportionately fewer people would be in the part of the tail who can turn on a 100+ mph fastball). But the familiarity aspect would almost certainly be non-linear, given normal distribution of pitcher velocity. If hitters only face Chapman-like fastballs from a handful of pitchers, their familiarity with how to barrel up wouldn’t be linear steps, would it?

    • I also think there is a relationship between called strikes off the plate and fastball velocity. Umpires have less time to judge a fastball at 95+ than a “normal” 90MPH one. The number of “missed” calls — I imagine — will go up.

      • It’s funny you should bring up umpires. I was wondering if the strike zone has evolved over the years to the point of affecting these numbers to a degree. I understand that each umpire has his preferencial zone he calls, but has a culture developed over the years of modifying the zone to be more hitter friendly, and then a few seasons later making it a pitcher’s zone? How much assistance in the fluctuation of numbers can the umpires be, and how much have they been? I’m not saying it’s a cognizant decision making process, but the boys in blue are fans as well. Maybe I’m overthinking it going from season to season, but I’m certainly willing to lay money that era to era it had a great affect on outcomes.

  4. Bottom line, the game has gotten more boring for the average fan. If the trend continues, I think you’ll see some teams move in their fences to help offset the decline in offense (looking at you Miami, Detroit, Mets, Oakland, Seattle).

    • Yeah, I was reading an article yesterday where the author was speculating the trend toward pitching was so dramatic that MLB would take steps to reverse it. There is historical precedent for that (lowering the height of the mound, for example). And that would be a league-wide solution, not dependent on individual teams taking action with their fences – although both of those could certainly take place.

    • The decline in offense is a REAL problem for the game. I’m am a huge fan and I watch and cheer regardless. But the game is out of balance now , like back in 1968. We can only lower the mound a little more.. That might help . Move the mound back ? .. Probably not going to happen ( more pitching arm injuries ) ..How about giving the hitter a little better break on the check swing ? That might help . There is no doubt pitchers are bigger , stronger and throw harder than ever .. especially kids from the US.. There are not many Harpers, and Trouts around these days . American kids just do not either play or stick with the game any more and allow themselves to become good hitters … Most all of the great hitters today are hispanic . American kids who do stay with the game are almost exclusively big strong pitchers as I see it .. The better US athletics play football and basketball . As a over the top baseball fan, its painful to watch. And that is a huge problem for the game ..

  5. Re: more hard throwing pitchers now

    I found this article:

    Here are some relevant parts:

    In the 2003 season, there was only one pitcher who threw at least 25 pitches 100 mph or faster (Billy Wagner). In 2012, there were seven, according to Baseball Info Solutions.

    In 2003, there were only three pitchers who threw at least 700 pitches 95 mph or better. In 2012, there were 17. There were 20 pitchers a decade ago who threw at least 25% of their fastballs 96 mph or faster. Last year there were 62, including Carter Capps, the Seattle Mariners’ 22-year-old right-hander, whose average fastball travels 98.3 mph, tying him with the Royals’ Kelvin Herrera for the top spot in the game.

    So yeah, it seems to jibe with all this data we have, and also answers my question if there are more guys now throwing 100mph with regularity.

    Very interesting. You’d think when pitchers were juicing, there’d be more hard throwers. Maybe the juiced pitchers got injured more since pitching it a lot more stressful on the body than hitting?

    • My sense from limited reading is that PEDs really didn’t help pitchers beyond recovery from injury. It might have aided in leg development which could contribute to velocity indirectly, but durability mostly. I’m not sure that pitchers reaching certain velocity has much to do with strength. I could be wrong, but I doubt Randy Johnson was physically stronger than most other pitchers. Maybe Mike can shed some light on what factors allow pitchers to throw the ball faster.

      We know that pitchers either didn’t use PEDs as much as hitters or it didn’t help them as much (or both) otherwise the HR/Runs rates wouldn’t have exploded during the Selig, I mean, Steroid Era.

    • Could be — but I think there is a greater emphasis on pitching metrics than before. Pitchers used to be evaluated baed on ERA, but FIP and other advanced metrics place a higher emphasis on strikeouts. With that, players selection will gravitate toward players with high K%, which coincides with throwing harder.

      The movement to protect young players’ arms probably also plays a role in it.

      It seems plausible to me that pitchers that quickly add muscle mass would be more likely to be injured than pitchers who gradually did the same. It is also worth noting that muscle recovery is a big part of performance enhancing substances.

      On the other hand, there has been a spike in TJ surgeries in recent years. That seems to indicate that there is a natural limit to how hard a pitcher can throw — regardless of how they got there.

      Health awareness in the coming years will be interesting.

      • Mike do you know if anyone has looked at the correlation between velocity and TJ surgery? I wonder if the pitchers (16?) who have had it this year represent disproportionately high-velo pitchers. You’d have to control for age cohort and the N isn’t big enough. But still, the explanation for the TJ “epidemic” may be the emphasis on throwing harder.

    • Just spitballing here, but could it be possible that strength is not the barometer it’s perceived to be when it comes to throwing fire? Perhaps flexibility and agility were sacrificed for strength and stamina. We really don’t know what sample size we are working with because we don’t know the percentage of users.

      Could more aggressive scouting, especially overseas (like Chapman), be turning up more guys with heat?

  6. Me, I prefer the chess match that takes place more often in low scoring games. I like fewer pitching changes, but more pinch runners, defensive replacements, etc. Fewer pitching changes means seeing higher quality arms longer. Also in low scoring games you always have the threat of a slugger instantly changing momentum with a single swing. I guess I don’t fall into the category of ‘casual fan’.

  7. Yes, it’s absolutely true that managers only want guys in the bullpen who are throwing heat. You don’t see guys like Leake coming into games in the 7th inning anymore. Kids are being groomed from an early age to throw hard. Fathers now sit in the stands with radar guns during little league games.

    Another factor is that pitching has fundamentally changed. Until Greg Maddux came along, you didn’t see guys throwing breaking stuff on both sides of the plate. Maddux effectively widened the plate on both sides, which has had an effect on umpiring as well. Maddux wasn’t a hard thrower, but hard throwers picked up on Maddux’s technique and began using it themselves. When you have a guy who can break the ball on either side of the plate and throw 97–you have a guy who is almost unhittable.

    Yet another factor is that the from a batter’s perspective, a K isn’t the humiliating AB it once was. There’s little shame in striking out 100 times. There’s contract money to be made hitting the ball hard–and if the price is a few more strikeouts–so be it. Plus, more than a few ballparks have short porches, which goad hitters into swinging from the heels.

    Evan Longoria once said: “I don’t have a two-strike approach. I mean, I could decide to shorten up and roll over and hit a ground ball. But on this level, if you roll over something because you were just trying to put the ball in play, you’re going to be out more than 95 percent of the time. It’s more about, what can I do to help the team? For me, it’s getting three healthy hacks and using them.”

  8. For Steve Mancuso and RLN staff. Do you think MLB should consider eliminating the extreme shifts promulgated in the game today? After all position players in football (i.e. lineman required to be in a stance) or basketball (3 seconds in the lane) are restricted by rule to be in certain positions during play.

    • Interesting. My initial reaction was to think no, never going to happen. But when you gave those examples of other sports, it made me pause. Baseball could pass a rule that says you have to have two infielders on each side of second base. Personally, I don’t feel strongly one way or another whether baseball should do something like that. If run suppression continues, they’ll do something, though. Maybe they’ll lower the pitcher’s mound again. Interesting idea, though.

      • I enjoy seeing the shifts and think they are clever. My opinion is that they should move the walls *back* 20 to 50 feet and flatten the mound. More speed guys in the outfield for me (and more triples and in-side the park home runs). The usual home run ball just doesn’t excite me.

  9. This was an excellent discussion. I’ve been saying all year that the Reds need to go find some guys who throw hard for their bullpen.
    Heck, I’d bring up Stephenson and let him pitch out of the bullpen for the year. He throws 97-98.

    Jumbo Diaz throws 98. Bring these guys up.

  10. This is a fantastic, thought provoking article and discussion. I really look forward to all the new features/articles this season with the information and discussions they bring.

    Thanks Mike.

  11. Is it possible to break down when in the baseball game home runs are hit? If there is a causal relationship between increased pitch speed and decreased home run frequency, and relievers throw significantly faster than starters, then more home runs should be hit off starters than relievers (adjusted per inning pitched by starters/relievers–not just because starters pitch more innings than relievers). Perhaps it’s possible to know, on average, when a reliever enters a game by determining the average innings pitched per game by a starter. Then compare that to when home runs are hit. If increased speed causes less home runs, and relievers throw significantly faster, then significantly more home runs per inning (possibly even broken down to a third of the inning) should be hit per starter-pitched inning than reliever-pitched inning. If this is not the case, perhaps speed is not a causal factor, or some other factor(s) is/are making it appear as though speed is not a causal factor.

    • Let me tweak that idea a bit, but it’s a great one. How about comparing whether fewer home runs are being hit in the 7-8-9 innings now compared to 2007, with the control group being innings 1-6 from 2007 to now. That would isolate the effect of the velocity more. The problem with comparing the bullpen innings to the starter innings is that all things aren’t equal. Starters are usually better pitchers, that’s why they are assigned more innings (Cincinnati Reds excepted). But your thought is a sound one, to see if there are measurable differences in late inning home run rates. I’m sure that data is out there somewhere.

      • Thanks for responding. So how does one go about obtaining this info? Also, I think the immense difficulty of isolating the causes of success in baseball (whether at the macro or micro level) is a testament to the game’s greatness.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s