Whiffing more and scoring less: how big is the impact?

On Wednesday’s NLCS game, Tom Verducci remarked on this factoid.

Rk Tm Year Games W L PA AB R H 2B 3B HR RBI BB SO BA OBP SLG OPS SB CS
1 SFG 2014 5 4 1 Ind. Games 189 169 15 41 5 1 0 12 13 18 .243 .301 .284 .585 2 2
Provided by Baseball-Reference.com: View Play Index Tool Used
Generated 10/16/2014.

Above is a list by team of the number of 2014 post-season games with 4 or fewer batting strikeouts. So, where is the rest of the list, you ask? Actually, that is the whole list. The Giants have had no more than 4 batting strikeouts in 5 of their 10 post-season games. The other 9 playoff teams – nada.

Verducci has expressed how impressed he is by the Giants’ ability to score without the need of a base hit, a knack he attributes to their low strikeout total. The rationale is that fewer strikeouts mean more balls in play, more pressure on the defense and, therefore, more runs scored. Is he right? Let’s find out.

Before we leave the Giants’ exploits, consider that they have averaged 3 runs in the 5 games above, four of them wins. This despite no home runs and a paltry .585 OPS. Obviously a very small sample size, but considerably better than this season’s 651 NL team games against NL opponents with no home runs and team OBP and SLG both under 0.350. The median runs scored in those games was only 1 with median strikeouts of 8, resulting in a 162-489 record for a 0.249 winning percentage (in other words, the 1962 Mets).

The prevailing trends in the majors are, of course, declining run production and ever increasing strikeouts. Thus, it would be reasonable to suppose that more strikeouts would, in fact, lead to less scoring. But, are the strikeouts a cause and are runs scored an effect? Or, are higher strikeouts and lower scoring both merely symptoms of increasing pitcher dominance?

To find out, I looked at all 2340 major league games in 2014 using data provided by Baseball-Reference.com. To encapsulate offense in each game, I calculated Net Bases as TB + BB + HBP + SB – CS – GDP. In other words, bases gained by batters plus net stolen bases (missing, of course, are bases gained or lost due to baserunning prowess). That metric looks like this:

Games by Net Bases and Strikeouts 2014

Those are the 2340 games of the 2014 season (or, rather, the 4680 team games) broken down by Strikeouts and Net Bases. Most games (about 82%) have between 4 and 12 team strikeouts and between 6 and 30 net bases. To look at the range of runs scored for the same strikeout levels, a metric of Runs per Net Base is derived. To make it more meaningful, I’ve called that Runs per 10 Net Bases. That result is shown below.

Runs per Net Base 2014
The center line is the average Runs per 10 Net Bases in all games with the indicated number of batting strikeouts (the columns show how many of those team games there were). The companion lines are +/- one standard deviation. While the standard deviation is substantial (i.e. there can be a wide range of runs scored from the same number of net bases), there is nevertheless a clear and consistent trend to declining scoring with increased strikeouts. A better illustration may be the chart below.

Runs per Game by Net Bases and Strikeouts 2014

Above are the average runs per game for the subset of games with the indicated Strikeout and Net Base levels. The highlighted numbers on the right side of the chart show how many runs per game are lost at each Net Base level between the 4-6 Strikeout level and the 10-12 Strikeout level. Those are significant reductions and a clear indication that a team will do much better if it can reduce its strikeouts while maintaining the same Net Base level. But, can teams do that?

While there are runs to be gained by reducing strikeouts, there is a much greater increase in runs that will result from increasing Net Bases. So, if a team reduces its strikeouts at the expense of its Net Bases, that result will likely be to a team’s detriment. Hence, the argument for striking out a lot – yes, it reduces balls in play (a lot) but that’s the price to be paid if you want more extra-base hits and, especially, more home runs.

The much stronger correlation between Net Bases and scoring than between Strikeouts and scoring can be gleaned from the table below.

Average Runs by Net Bases and Strikeouts

Above are the average runs scored in 2014 games for each combination of Net Bases and strikeouts (the averages are subject to small sample distortion as some cells will represent small numbers of games). The bordered area is where 82% of games occurred. The relation between strikeouts and scoring is most pronounced in the corners of this area, showing the highest (red) and lowest (blue) scoring games. Looking at the center of the bordered area, though, will show only modest changes in run scoring scanning along a row (especially below 10 strikeouts), but much more significant changes scanning along a column (especially below 20 net bases).

So, what have we learned? I would summarize as follows:

  1. increased scoring can result from fewer strikeouts provided that offensive output (as represented by Net Bases) does not suffer as a result
  2. therefore, let your sluggers slug and accept the strikeouts that result – they’re likely worth it (up to a point)
  3. but, get your would-be sluggers who aren’t to focus on making more contact

Who falls into the category for point 3? Well, these guys for starters. They had ISO of 0.100 or less in 2014 despite striking out in more than 15% of their 400+ PAs.

Rk Player SO ISO PA Year Age Tm G AB R H 2B 3B HR RBI BB BA OBP SLG OPS
1 Chris Johnson 159 .098 611 2014 29 ATL 153 582 43 153 27 0 10 58 23 .263 .292 .361 .653
2 Austin Jackson 144 .090 656 2014 27 TOT 154 597 71 153 30 6 4 47 47 .256 .308 .347 .655
3 Jackie Bradley 121 .068 423 2014 24 BOS 127 384 45 76 19 2 1 30 31 .198 .265 .266 .531
4 Leonys Martin 114 .090 583 2014 26 TEX 155 533 68 146 13 7 7 40 39 .274 .325 .364 .689
5 Allen Craig 113 .100 505 2014 29 TOT 126 461 41 99 20 1 8 46 35 .215 .279 .315 .594
6 Dee Gordon 107 .089 650 2014 26 LAD 148 609 92 176 24 12 2 34 31 .289 .326 .378 .704
7 Robbie Grossman 105 .100 422 2014 24 HOU 103 360 42 84 14 2 6 37 55 .233 .337 .333 .670
8 Jason Kipnis 100 .090 555 2014 27 CLE 129 500 61 120 25 1 6 41 50 .240 .310 .330 .640
9 Brock Holt 98 .100 492 2014 26 BOS 106 449 68 126 23 5 4 29 33 .281 .331 .381 .711
10 DJ LeMahieu 97 .081 538 2014 25 COL 149 494 59 132 15 5 5 42 33 .267 .315 .348 .663
11 Joe Mauer 96 .095 518 2014 31 MIN 120 455 60 126 27 2 4 55 60 .277 .361 .371 .732
12 Emilio Bonifacio 85 .086 426 2014 29 TOT 110 394 47 102 17 4 3 24 26 .259 .305 .345 .650
13 Daniel Nava 81 .091 408 2014 31 BOS 113 363 41 98 21 0 4 37 33 .270 .346 .361 .706
14 Jon Jay 78 .075 468 2014 29 STL 140 413 52 125 16 3 3 46 28 .303 .372 .378 .750
15 Ruben Tejada 73 .073 419 2014 24 NYM 119 355 30 84 11 0 5 34 50 .237 .342 .310 .652
Provided by Baseball-Reference.com: View Play Index Tool Used
Generated 10/17/2014.
0 0 votes
Article Rating
Subscribe
Notify of
guest

11 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Arsen
Arsen
9 years ago

Great work. In this run scoring environment a team should have a couple of big boppers and a bunch of guys who make contact. I’m sure a bit more speed would help teams put pressure on the defense and score some runs. Sounds like the 1974 to 1985 seasons. That’s the baseball I grew up on. It does seem like teams have been quite slow to adjust to the pitching dominance we’ve seen over the past couple of years.

tag
tag
9 years ago

Doug, Exceptional job and really interesting data. Just on a subjective level, what I’ve noticed in the playoffs is the value of making even bare-minimum contact with a runner on third, i.e. how difficult it is for teams to make a force play at the plate on easy, routine grounders. In the clinching KC-Baltimore game, Oriole 1B Pearce gloved a bouncer and made a perfect throw to the plate, but the ball was kicked out of the catcher’s glove and resulted not only in the runner scoring but a subsequent runner coming around for the difference in the 2-1 win.… Read more »

Richard Chester
Richard Chester
9 years ago

Nice work Doug. I found the chart with the red and blue shading easiest to follow. My feeling is that a larger number of games played would make the chart more meaningful. Perhaps you could combine several seasons into one analysis. Also I would like to see a graph with SO as the X-axis and Runs as the Y-axis, one plot for each number of Net Bases but that would make for a very crowded graph with overlapping lines.

brp
brp
9 years ago

One of the sports shows had a breakdown recently of the Royals and how they’ve been able to get runners from 3rd w/less than 2 outs almost every time in the postseason.

Interesting stuff here.

Daniel Longmire
Daniel Longmire
9 years ago

There seems to be an odd up-tick in runs scored between the 13 and 14 strikeout categories, but only in the upper half of the chart. Between 5 and 17 Net Bases, only four lines show a decline in production. What could account for this? A small sample size? Random noise?

Daniel Longmire
Daniel Longmire
9 years ago
Reply to  Doug

That makes sense, Doug; thanks for the clarification. Strange how there were MORE games with a high number of Net Bases AND a high amount of strikeouts than the expected low NB/high K combo.

Doug
Doug
9 years ago

In my first chart, the proportion of low strikeout games (under 10) that were also low NB was fairly high (~75%). But, low NB games (10 or less) overall were less than 25% of all games.

Ken
Ken
9 years ago

Speaking of high strikeout rates, I got to wondering how often a pitcher gets a run home with a runner at 3rd and less than 2 outs. So I did some research. For opportunities, I used total PA in those situations minus walks and HBP unless the bases were loaded. A success was any PA where an RBI occurred. I’ve got the yearly numbers back to 2004, and so far, the worst percentage was in 2014. Anyone want to guess what it was? The rate for non-pitchers was 55.7%.

Camila Auvil
9 years ago

Good article, thanks.