View Full Version : The Value of Getting Ahead of Hitters with a Glance at the Rotation

04-19-2006, 02:19 AM
Earlier in the year before spring training, I took a look at how Reds hitters were able to effectively – or ineffectively as in the case of Wily Mo Pena and Tony Womack – work themselves into better batting counts to help them produce more offense at the plate.

Any hitters found to get themselves in more hitting counts and less pitching counts than the league average were stated to be exhibiting positive plate discipline. Guys such as Pena and Womack, who were failures at working the count, were shown to have horrible plate discipline. In a way, the data stated the obvious for players on the extreme ends of the spectrum while also providing some more information on players not on the extreme ends.

Mixed in with the discussion of how better batting counts helps hitters was the question of it also working in reverse; how much do better pitching counts help pitchers? Finally, I decided to take a quick look.

In 2005, I found that 34.834 percent of all plate appearances ended with the batter having the advantage in the count. Likewise, I also found that 30.993 percent of all plate appearances ended with the pitcher having the advantage in the count. Throughout various seasons those percentages have likely fluctuated slightly, but they should be good enough to use as a starting point to determine league average. I’ll also use the same methodology of examining five outstanding pitchers in recent seasons and compare them to the current five guys in our rotation. While the analysis isn’t exceptionally thorough to determine the high-end goals of batter count results for pitchers, it’s enough to provide a decent rough idea.


Roger Clemens 4608 .290 .466 82.29 15914 28.956% 93
Greg Maddux 5496 .281 .467 98.14 16702 32.906% 106
Pedro Martinez 3475 .264 .419 88.81 9880 35.172% 114
Randy Johnson 5105 .269 .425 69.93 14515 35.171% 114
Johan Santana 1284 .329 .516 51.36 3504 36.644% 118

Right away, the glaring figure belongs to Roger Clemens. Throughout his career Clemens has been below average in ending opposing batter plate appearances in a pitcher’s count. Maddux, Martinez, Johnson and Santana, while still above average by a decent amount, are by no means crushing the league average in getting ahead of batters. The important column to pay attention to is PA/HR, as Clemens, Maddux and Martinez were exceptionally stingy giving up the long ball when they were ahead in the count. Johnson and Santana weren’t quite in the others’ territory, but their PA/HR numbers are still pretty solid. What these five pitchers also have in common are the miniscule SLGA numbers when ahead in the count as everyone except for Santana is holding opposing batters to a slugging percentage under .300 when they are ahead in the count.


Aaron Harang 798 .382 .595 46.94 2292 34.817% 112
Bronson Arroyo 761 .419 .644 38.05 2556 29.773% 96
Eric Milton 2034 .378 .605 40.68 5861 34.704% 112
Brandon Claussen 361 .465 .701 60.17 1109 32.552% 105
Dave Williams 391 .345 .537 43.44 1423 27.477% 88

Immediately, the first thing that jumped out at me was Arroyo. His overall HR/9 rate in all counts throughout his career has actually been slightly above average, but when he’s ahead in the count he’s having major problems avoiding the long ball. When the Arroyo/Pena trade first occurred, vermonter provided some analysis of Arroyo and also stated that he tends to make a few mistake pitches each outing, and it’s those mistake pitches that get rocked. The data above could definitely support that theory, since when Arroyo gets ahead in the count he may be more apt to finish a hitter off with a sweeping breaking ball out of the zone. If it’s a mistake pitch, however, that breaking ball suddenly becomes a hanger, and we all know what big league hitters can do to hangers.

Another item that caught my eye was the PCPA+ numbers of Harang, Milton and Claussen and how they compared to the top line starters. Each of those three Reds arms have a career history of getting ahead of hitters at much the same rate as guys such as Martinez and Johnson. The two key columns to look at, however, are SLGA (slugging percentage against) and PA/HR. With the exception of Claussen, each of our Reds arms have been giving up twice as many home runs as the fab five when ahead in a pitcher’s count. The SLGA numbers show the same trend as Reds pitchers are giving away about 100 more points in slugging percentage compared to the fab five in pitcher count plate appearances.

Troublesome? You bet, but it doesn’t end here.


Roger Clemens 5575 .472 .906 50.23 15914 35.032% 99
Greg Maddux 4789 .473 .869 48.87 16702 28.673% 122
Pedro Martinez 2981 .480 .899 35.49 9880 30.172% 116
Randy Johnson 4821 .501 .972 38.88 14515 33.214% 105
Johan Santana 1069 .504 .947 28.13 3504 30.508% 114

Again, much like the pitcher count figures, Clemens comes in at near a league average HCPA+ rate while the other four arms come out significantly above average. Not surprisingly, PA/HR figures are drastically different, however, as there appears to be some correlation that the five arms above gave up a rate of home runs twice as high when behind in the count compared to being ahead in the count. Likewise, SLGA numbers start to creep up as the fab five hovers around the high .400s to low .500s when behind in the count. For Clemens, he’s able to offset the disadvantage of being average in relation to getting ahead/behind hitters, and that's largely because he’s still able to keep the ball in the park more often than everyone else.

When we pull up Reds figures, it gets ugly in terms of PA/HR and SLGA.


Aaron Harang 754 .638 1.098 31.42 2292 32.897% 106
Bronson Arroyo 636 .599 1.059 37.18 2556 32.003% 109
Eric Milton 1914 .681 1.118 19.14 5861 32.657% 107
Brandon Claussen 358 .636 1.170 21.06 1109 32.281% 108
Dave Williams 561 .595 1.057 23.38 1423 39.424% 88

Two words: batting practice.

Harang, Milton and Claussen all avoid getting behind in the count a bit better than the league average, but when they do get behind, lookout. They’re giving up home runs at a terrible rate, and all five have an opponents slugging percentage nearly matching and/or exceeding .600. Arroyo is by far the best of the group as his PA/HR rate compares well to Johan Santana and Randy Johnson, and it is very likely his high SLGA can be attributed to giving up Fenway doubles.

The career numbers for the others, however, do not look promising at all, though it should be stated that if Aaron Harang’s 2005 season – in which he gave up less than one home run per nine innings – was extracted from his career then we would obviously notice a considerable difference. With the exception of Arroyo who actually improved his rate of PA/HR while behind in the count, Reds pitchers largely started giving up twice as many home runs when they fell behind in the count, a rate increase that correlates well with the rate increase by Clemens, Maddux, etc.

Arroyo’s PA/HR figures by count is still a massive aberration, and I’m still not sure yet what to make of it other than perhaps a tendency to hang breaking balls when trying to put a hitter away.

Here’s the two-strike data for both sets of pitchers, and I’ll present them side-by-side.


Roger Clemens 8347 .258 .480 103.05 15914 52.451% 114
Greg Maddux 7853 .250 .482 106.12 16702 47.018% 102
Pedro Martinez 5513 .227 .417 108.10 9880 55.800% 121
Randy Johnson 8246 .226 .425 76.35 14515 56.810% 123
Johan Santana 1913 .271 .485 76.52 3504 54.595% 118


Aaron Harang 1207 .417 .697 52.48 2292 52.661% 114
Bronson Arroyo 1148 .334 .595 47.83 2556 44.914% 97
Eric Milton 2976 .347 .599 43.13 5861 50.776% 110
Brandon Claussen 577 .390 .692 57.70 1109 52.029% 113
Dave Williams 612 .355 .627 34.00 1423 43.008% 93

We now start to see a bit of separation in the frequency of reaching a specific count as the five legendary starters, Maddux notwithstanding, were able to work themselves into devastating two-strike counts more often than Reds arms have. Again, notice the discrepancy in PA/HR and SLGA. When Clemens, Maddux, etc. get two strikes on you, you’re just about already out at the plate. For Reds pitchers, however, they’re still giving opposing hitters far too many total bases even after establishing a two-strike count on hitters.

One more table, this time comparing the two groups when they’ve established an even count, either 0-0, 1-1 or 2-2.


Roger Clemens 5731 .404 .650 46.98
Greg Maddux 6417 .414 .691 51.34
Pedro Martinez 3224 .398 .633 48.91
Randy Johnson 4589 .433 .676 33.50
Johan Santana 1151 .462 .703 44.27


Aaron Harang 740 .605 .925 27.41
Bronson Arroyo 977 .539 .821 40.71
Eric Milton 1913 .565 .851 22.77
Brandon Claussen 390 .508 .769 32.50
Dave Williams 471 .550 .822 26.17

Thus far in his career, Arroyo has fared very well in avoiding the long ball when even in the count, and Claussen’s figures aren’t terrible. Eric Milton’s numbers, as most would guess, are absolutely abysmal, and Dave Williams isn’t much better. Once again, if Harang’s 2005 season was extracted from his career totals we’d likely see a PA/HR rate closer to Arroyo’s than to Milton’s.

One way this data can be even more useful for analyzing our pitching here in Cincy is to break it up by season, which I hope to do for Harang’s 2005 season and Milton’s last two seasons. The sample size is only 10 pitchers total, but there appears to be a correlation of giving up twice as many home runs per plate appearances when behind in the count than when ahead in the count.

Just by looking at walk rate, it’s probable that Harang’s improved HR/9 rate last season was due to both an improvement in his stuff and his command of the strike zone, but I am curious as to how much he improved in commanding the strike zone and getting ahead of hitters last season. As others have pointed out, Milton’s disastrous career turn in HR/9 starting in 2004 is likely due to a loss in velocity because of his knee condition rather than losing command of the zone.

Suffice to say, one quick way to drastically improve home run rate is to improve overall command of the strike zone. By getting ahead of hitters and getting two strikes on them, a pitcher severely lessens the potential of giving home runs and overall total bases. None of this should be surprising to anybody, though I am curious if the rate of home runs allowed increases at a rate two times higher across the board for all pitchers when they’re behind in the count. At least in 2005, the league average was a 31.50 PA/HR rate when the hitter had the count advantage compared to a 54.88 PA/HR rate when the pitcher had the count advantage.

One thing is for certain and can never be repeated enough: Identify pitchers who can keep the ball in the yard who also have a tendency to get ahead of hitters in the count, and the results will definitively speak for themselves.

04-19-2006, 09:20 AM
I am not convinced that getting ahead in the count is all its made out to be.

One way to get ahead of the hitter is to throw a strike that the batter cannot hit. Guys like Clemens can do that while guys like Williams cannot.

Thus it looks like getting ahead in the count is the key, while the real key is throwing pitches that guys cannot hit.

Of course I am not convinced of this either, but it does cross my mind from time to time.

Thanks for the info above, interesting.


Johnny Footstool
04-19-2006, 09:30 AM
One way to get ahead of the hitter is to throw a strike that the batter cannot hit. Guys like Clemens can do that while guys like Williams cannot.

Thus it looks like getting ahead in the count is the key, while the real key is throwing pitches that guys cannot hit.

Exactly. That's the difference between "throwing strikes" and "pitching to contact."

04-19-2006, 09:38 AM
Yep, GL, it was ungodly late when I finally posted that, but the same thoughts crossed my mind too. Great pitchers become merely "human" when behind in the count, but they're still great pitchers, and when they're ahead in the count they're nearly impossible to hit.

It's a small sample size, but at least with those 10 guys it doesn't appear that any extremes exist like we see with hitters. There are a handful of hitters who get ahead in the count nearly half of all their PAs, but it doesn't look like any pitchers come close to doing that. The batter count factor seems to be much more stated, and quite a bit more important for the hitter than the pitcher.

Having the stuff that guys either cannot hit out of the park or cannot hit at all is the key ingredient, with the ability to get ahead of more hitters than other arms just being a slight bonus to add on.

04-19-2006, 10:50 AM
Here are the seasonal splits for Harang and Milton the last few seasons:

Player/Year K/9 BB/9 HR/9 BABIP ERA HCPA% PCPA% 2SPA%

Harang 2004 6.99 2.96 1.45 .315 4.86 32.51% 34.25% 51.30%
Harang 2005 6.93 2.17 0.94 .310 3.83 28.82% 39.95% 53.50%

Player/Year K/9 BB/9 HR/9 BABIP ERA HCPA% PCPA% 2SPA%

Milton 2002 6.37 1.58 1.26 .292 4.84 32.15% 35.85% 52.49%
Milton 2004 7.21 3.36 1.93 .274 4.75 31.48% 33.02% 52.66%
Milton 2005 5.94 2.51 1.93 .329 6.47 29.66% 34.52% 47.33%

This is very interesting, IMO.

In 2005, Harang got into less hitter counts, more pitcher counts and more two strike counts than the previous season, and the differences are rather substantial. His K/9 and BABIP was largely unchanged, but he showed vast improvements in his BB/9 rate and especially his HR/9 rate. This makes sense, and it's possible to conclude that some of his overall improvement in ERA by a full run was due to just getting himself in better batter/pitcher counts while on the mound in 2005. It's also possible to conclude that some of his overall improvement in ERA was making more bats miss, though his overall K/9 rate was virtually identical.

Eric Milton's seasonal splits show no relation to batter/pitcher counts, however. The only noticeable change by count is Milton's two strike counts in 2005 were reduced, which is likely because he got tagged in quite a few PAs before he could even establish a second strike. The only explanation for Milton's demise the last few seasons seems to be the knee injury that several others have already pointed out, as Milton has not really declined in losing control of the strike zone in individual batter matchups.

Arroyo coming up ...

04-19-2006, 11:09 AM
And the Arroyo seasonal splits for 2004 and 2005:

Player/Year K/9 BB/9 HR/9 BABIP ERA HCPA% PCPA% 2SPA%

Arroyo 2004 7.15 2.37 0.86 .299 4.03 29.67% 31.39% 48.48%
Arroyo 2005 4.38 2.37 0.96 .285 4.51 30.92% 31.61% 46.67%

Like Milton, the count differences for Arroyo in each season were neglible, at best, with the largest difference being a reduction in two-strike counts, and that is understandable given Arroyo's sudden drop in K/9 rate. Arroyo didn't lose the ability to get himself in primo counts last season compared to his 2004 campaign, which points to his drop in K/9 and increase in ERA likely being a full result of him just missing fewer bats.