(The following is being syndicated from The Captain’s Blog).
Offense has been down across the majors this season, continuing a trend that began at the start of the decade and accelerated last season.
In the National League, the per-team average has been 4.13 runs, which would be the lowest output since 3.88 runs in 1992. Meanwhile, if the American League’s current average output of 4.26 runs per team remains constant over the rest of the year, it will be the junior circuit’s lowest offensive display since the strike-shortened season of 1981.
Runs Scored by League, 1901-2011
Source: Baseball-reference.com
One of the main factors that seems to be driving this downward offensive trend is a similar decline in the number of home runs. In 2000, when run production in the major leagues reached 5.14 per team, 3% of all plate appearances culminated in a homer. This year, the homerun rate has fallen all the way to 2.3%. During that span, runs per game has almost correlated in lock step with the homerun rate (R =0.896), so there can be no denying that fewer balls over the fence have meant fewer runners crossing the plate.
The connection between homeruns and run production is an easy one, but what is at that root cause of the decline in long ball power? Countless theories abound, but most center on the gradual increase in testing for both steroids and amphetamines. Although part of the overall trend could be attributable to that dynamic, it seems likely that many other variables have also played a role.
On ESPN’s recent Sunday Night Baseball telecast, Bobby Valentine relayed a conversation with Chipper Jones in which the future Hall of Famer lamented the increased use of cutter. According to Jones, the refinement and proliferation of the pitch has been a key factor in swinging the balance back in favor of the pitcher.
Cutter Above: Pitchers Who Most Rely on the Cutter, 2011
| Pitcher | Team | CT% | CTv |
| Mariano Rivera | Yankees | 88.3% | 90.6 |
| Andy Sonnanstine | Rays | 67.4% | 84.2 |
| Scott Atchison | Red Sox | 56.7% | 85.5 |
| Rich Thompson | Angels | 55.6% | 87.8 |
| Mike Adams | Padres | 54.9% | 87.2 |
| Lance Cormier | Dodgers | 52.1% | 86.3 |
| Roy Halladay | Phillies | 44.5% | 90.6 |
| Kerry Wood | Cubs | 44.1% | 89 |
| Brian Wilson | Giants | 41.9% | 89.5 |
| Brandon Lyon | Astros | 41.5% | 89.1 |
Note: CT%=percentage of pitches that are cutters; CTv=average velocity of cutter.
Source: fangraphs.com
Although it would be incredibly difficult to verify Jones’ claim, the increasing popularity of the cutter is undeniable. According to fangraphs.com, in 2006, only 25 pitchers threw the cutter at least 15% of time. This year, that number has jumped all the way to 64. In total, the percentage of cutters thrown in the majors has increased by almost 100% since 2006. While it’s true that the pitch/FX data used to make pitch classifications is not infallible, such an exponential increase should easily mitigate any potential margin for error.
Emergence of the Cutter: Pitch Selection from 2006 to YTD2011
Note: SF=splitter, CH=changeup, CB=curveball, CT=cutter, SL=slider, FB=fastball
Source: fangraphs.com
More than anyone else, Mariano Rivera has become synonymous with the cutter. Considering his incredible success, which has largely been built on the pitch, it’s not surprising that many others would seek to follow his lead. However, just because more people are throwing the cutter doesn’t necessarily mean they have been as successful doing it.
Unfortunately, there isn’t enough reliable data upon which to base a definitive conclusion about the relationship between the cutter and the current offensive malaise. However, there are a few hints floating around. One of the pioneers in developing benchmarks for pitch types has been The Hardball Times’ Harry Pavlidis, and as luck would have it, he recently published data updated through the end of the last week. Using Pavlidis’ benchmarks, it’s possible to examine Chipper Jones’ theory a little further.
Pitch Type Benchmarks
| Type | GB% | LD% | FB% | PU% | HR/FL% |
| All | 45% | 19% | 29% | 7.40% | 7.10% |
| CH | 49% | 18% | 26% | 6.60% | 7.30% |
| CU | 50% | 19% | 25% | 5.80% | 7.20% |
| FA | 36% | 21% | 34% | 9.40% | 7.20% |
| FC | 44% | 20% | 27% | 8.50% | 6.30% |
| FS | 53% | 17% | 24% | 6.20% | 6.30% |
| KN | 43% | 17% | 30% | 9.90% | 9.10% |
| SB | 56% | 15% | 26% | 3.70% | 9.10% |
| SI | 52% | 19% | 24% | 4.40% | 6.70% |
| SL | 45% | 17% | 29% | 8.50% | 7.80% |
Note: Click on source link for an explanation of data and related abbreviations.
Source: Harry Pavlidis/www.hardballtimes.com
The ratio that practically jumps out of the chart above is the homerun percentage for fly balls hit off the cutter. Although more of a contact pitch than the curve ball and slider, the cutter ranks as the most effective breaking pitch when it comes to keeping a batted ball in the park. Without having access to more granular data, we can’t say that the cutter actually results in fewer homeruns per times thrown (especially compared to the curve, change and slider), but it does seem to be superior to the fastball in this regard. Considering that much of the cutter’s growth has come at the expense of the fastball, it’s probably safe to assume the pitch has played at least a small role in the lower homerun totals.
Exactly how much influence has the cutter had on offensive output? With more data, we should be able to answer that question. Until then, it might be best to defer to Chipper Jones.
If you enjoyed this article, please consider sharing it!
11 Responses to Is the Cutter Responsible for the Decline in Offense?
Leave a Reply
TIQ IQ
I wouldn’t draw any large conclusions from this, as there are likely large classification issues — especially with the data from pre-pitchf/x era where the pitches were classified by eye rather than cameras.
It’s an interesting theory but I don’t know how much it is coming ‘at the expense of the fastball’ so much as the PITCHf/x algorithm is classifying a few more cutters as they continue to refine the system. There’s still large error bars — even at hte league-wide level — that make it unreliable without combing through the data like Harry does. A large amount of cutters get labeled as just fastballs or sliders as I believe it (along with the splitter) is the hardest pitch for the PITCHf/x cameras and neutral net to properly identify during the game or after the fact.
I wouldn’t (and didn’t) draw any large conclusions either, but there is also anecdotal evidence suggesting that the cutter is exploding in popularity. Also, even the jump from 2009 to 2011 alone is significant, which mitigates concern over the more flawed data from before that period.
I think we can definitively conclude that more pitchers are throwing more cutters, and according to the benchmarks, it seems like cutters are less likely to leave the ballpark. If both assumptions are true, the cutter would bear some responsibility for the offensive decline. I’d like to know, for example, how many HRs are hit per each pitch thrown. That would at least allow us to calculate theoretical changes in HR totals based on historical data. Until then, I think it’s safe to stick with the simple conclusion and continue to keep an eye on the growth of the cutter.
I still contend it’s the size of the strike zone. Questec was introduced because a lot of umpires were calling pitches in the opposite batter’s box strikes. Since it’s been removed (a bad idea replacing a good one), the strike zone has continually gotten larger.
As always happens when scoring is down, so is attendance. Outs are failures-too many failures vs too many successes bores most fans and they stop watching or attending games.
The larger strike zone is an interesting theory, but I am not sure offense drive attendance as much as you suggest. Perhaps that will the topic of my next post.
Is the strike zone continuing to grow, or are is the increased ability to measure pitches (and it becoming publicly available data) just made it easier to see how big the strike zone has been all along? I’d lean towards the latter.
Will,
Where are you getting your pitch selection data for the league? You say, “In total, the percentage of cutters thrown in the majors has increased by almost 100% since 2006. While it’s true that the pitch/FX data used to make pitch classifications is not infallible, such an exponential increase should easily mitigate any potential margin for error.”
Pitch f/x did not exist before 2007, and I don’t think it started classifying cutters reasonably until about 2009 (as in the algorithm changed). If you are getting that data from fangraphs, it’s probably BIS data and not pitch f/x. It’s hard to say how accurate the BIS stringers are, so that may not provide very good evidence.
Also, it seems unlikely that a small increase (if it is small) in cutter percentage could have such drastic effect. However, we also need to consider how increased cutter usage effects the effectiveness of other pitches. Perhaps it’s not so much the cutter, bot how the cutter makes a fourseam less effective? And it’s hard to say how much the lower HR/FL means for cutters, because not all pitches are thrown in the same counts.
The data is from fangraphs, I only went back to 2006 because I thought they were using Pitch/FX since that time (thought that was when pitch/FX began). Does fangraphs disclose where they get their data? If they aren’t using pitch/FX, I would be less likely to use that data in the future.
Also, maybe I didn’t make more clear, but I wasn’t suggesting it had a drastic effect…for many reasons, including ones you cited, we can’t know that with the data we have right now.
I just read over your post again, and I must have been a little hasty the first time before making that comment. You definitely don’t suggest a drastic effect, so sorry about that!
And Fangraphs has disclosed where they get their data, and they carry both pitch f/x and BIS data. If you look at any player page, under the main page you will see “pitch type” info. That’s BIS data. To get to pitch f/x data, you need to go to the pitch f/x tab. They only show aggregate BIS data for some reason
Ah…thanks for that bit of info. I wasn’t aware of that. Thought the aggregate data was pitch f/x.
Also, i don’t want to seem like I’m trashing the article, either. Well done and just a singular step to what could possibly be a slice of the answer to run scoring drop-off. I enjoyed it and just wanted to clarify that I wasn’t trying to solely rip on it.
I didn’t come away with that impression, but if anyone ever feels a rip is justifiable, feel free to fire away!