Things are getting pretty crazy what with the season about to start,
BP book signings, season previews, and some other behind the scenes
projects. But before we get into the season, I’m going to dig into the
mailbag and clear up a few outstanding issues from some previous

First one is from reader J.E. regarding the price
of strikeouts

I found it interesting that while you
adjusted higher-contact guys downward for GIDPs, you don’t seem to have
taken into account the fact that a certain percentage of balls put in
play also become basehits. Then again, everyone treats the numbers that
way, which is why it drives me so crazy that I can’t stop focusing on
one thing.

If Adam Dunn strikes out 50 fewer times, and gets a single 10 of
those ABs, he’s at 283/402/586. Why do we always assume that 100% of
ABs that would be converted to non-SO outs would still be outs? As you
say, reaching base a couple handfuls more per season is small potatoes,
but not meaningless.

This point was raised by quite a few readers. The main problem is
that we can’t say what Adam Dunn would have done if he
hadn’t struck out all those times because we have to assume that a
change in outcome is a result of change in approach. Because power and
strikeouts are positively correlated, we would have to assume that
Dunn’s change in approach would decrease his power numbers while
decreasing his strikeouts, but we cannot say with any accuracy how much
his power numbers would decrease. Thus, we can’t reliably estimate the
cost of reducing strikeouts and then compare it with the expected

Hitters aren’t as susceptible to the severe regression to the mean
that pitchers experience with regards to balls in play. (If they were,
Ichiro Suzuki wouldn’t be, well, Ichiro!) Instead,
successful players tend to follow either a power + strikeouts approach
or a slap + run approach; there are obvious pros and cons with each

Are higher power players costing themselves singles by striking out
more often? Sure, but we’re forced to focus on the results rather than
the process to accurately value the cost of those strikeouts. If Dunn
did everything else the same, but struck out 50 fewer times and had 10
more singles he would certainly be more valuable than his current
approach. But we don’t have that data; we can’t say what he would have
done. So we’re left with evaluating the results of the process.

It’s a tough situation because we can mentally say Dunn would be
valuable if he put the ball in play more often, but we don’t have that
data. Instead, we’re left to analyze the numbers we do have and while
speculation about what Dunn would have done is nice, it is speculation
and not analysis.

Next one, same column, but a different issue:

When Red Sox fans discuss Mark Bellhorn,
the term ‘productive strikeout’ addresses the larger number of pitches
he takes when striking out. Those pitches give Manny and Papi better
looks at the starting pitcher’s ‘out pitch,’ but they also fatigue the
starting pitcher.

At the level of difference you’re discussing–hundredths of
runs–does the speed with which the starting pitcher is fatigued come
into play? A strikeout must, on average, take at least one more pitch
than a ball-in-play out, and starting pitchers have run averages
significantly higher than middle relievers. Should this impact the

— Bill Wellman

Interestingly, several Red Sox fans wrote discussing Mark
and his whiff issues (whiffues?). Last year the
discussion was between Bellhorn and Pokey Reese for
starting second base job; this year it appears to be between Bellhorn
and Edgar Renteria for the second spot in the lineup.
We haven’t been able to quantify any effect of previous batters taking
more pitches to inform subsequent hitters, but look for something along
those lines in the future.

With regards to seeing more pitches, the correlation between
strikeout rate and pitches per PA is very small; the coefficient of
correlation (r-squared) was only .1816 among players with at least 400
PA in 2004, meaning strikeout rate only explains 18.16% of the variance
in P/PA. In short, there are better ways to see a lot of pitches than
simply by striking out.

Regarding higher strikeout hitters fatiguing the opposing pitcher:
getting the starter out of the game isn’t necessarily a good thing as
relievers are now putting up comparable
numbers to starters
. Getting the starter tired earlier certainly
gives teams a slightly better chance of facing the lower rungs of the
bullpen or that the other team will keep the fatigued starter in for
more pitches, but because strikeouts aren’t necessarily a result of
patience at the plate, it’s tough to call those extra pitches a benefit
of striking out.

Going a little further back, among the several articles on park
, many readers pointed out that I neglected to include
weather information in a discussion about predicting park factors based
on park dimensions.

It seems to me that weather conditions
would play a role in park factors since it’s clearly demonstrated that
humidity and temperature affects the action of the ball in play.

Given local weather oddities of the last few years and variations of
the time of the year when the games are played at the ball parks, I’d
expect to see higher correlation if those types of factors were taken
into account.

There are other effects also, such as number of day or late afternoon
games that may affect batters’ vision that could affect park factors.

Taking into account such environmental factors would obviously take
rather labourious work to try to demonstrate, but I suspect the
additional data might clear up the picture quite a bit.

— D.K.

When I was in Chicago last summer for a couple Cubs games, it was
obvious that the weather plays a significant role in the park factor
there. Nate Silver pointed out to me later that he believes Wrigley’s
park factor is based almost solely on the prevailing winds during a
season and after taking in a few games there, I would certainly think
plays a big part.

Weather information is available for games going back quite some
(once again, all hail Retrosheet) and it would
be interesting to go back and see if prevailing weather information has
an effect on park factors. However, the focus of this article was
attempting to predict park factors going forward and – despite the best
efforts of Clay Davenport – the forecast for April-October 2005 isn’t
available yet. Maybe in a few seasons.

Keeping with the Cubs theme, several Cub fans wrote to me about my
assertion that the Cubs’ Spring
program was the worst in the business:

Regarding the Cubs spring training
program, I do know that they had an insert for NRIs and minor leaguers
at the start of Spring Training. I did not get to go this year, but my
father did, and he returned with the insert autographed by the new Cubs
announcers, Bob Brenly and Len Kaplan. To the annoyance of my wife, the
insert remains on the dining room table where it has been for over a
week now, awaiting its inevitable journey to the folder in my attic
where I keep my Buzz Capra and Jose Cardenal autographs.

— C.R.

I was able to find the aforementioned “Vine Line” that had more
information (as well as Sammy Sosa on the cover), but
no one has cleared up whether this was the phantom NRI inserts people
keep mentioning.

In defense of the Cubbies, I dropped by a Brewers game later that
week and found their provided scorecards also lack a pitching section.

Finally, digging way back to the idea of tired
and how they perform the next day:

Interesting article, but I don’t think
you addressed another key factor: whether the tired team was playing at
home in both games or had to travel between games. A swing in winning
percentage by 50 points is significant, but in the ballpark of what you
expect the home field advantage to be. I’d like to see you expand this
study taking into account expected W/L based on win percentage and
home/road. I imagine you’ll still see the same drop, but it may offer
further enlightenment, since I’d expect a team that played an extra
inning game and then had to travel would do a whole lot worse than a
team that played at home for both its extra inning game and the
following game.

— Matthew Knight

I was going to get into travel in another article, but since you
asked, these are the year by year numbers for teams that played at
11 innings and traveled to play the next day:

1995  11  5     6   .500    .455
1996  14  6     8   .491    .429
1997  20  10   10   .494    .500
1998  20  11    9   .488    .550
1999  17  8     9   .520    .471
2000  14  6     8   .492    .429
2001  14  6     8   .484    .429
2002  12  5     7   .487    .417
2003  21  6    15   .483    .286
2004  19  8    11   .511    .421

The TOTW% column is their expected winning percentage for the
EXTRAW% is their winning percentage in the games after the long
As you can see, every year except 1997 and 1998, the winning percentage
decreases substantially, more so than if the teams do not have to
travel. There’s some concern about sample size in each individual year,
but with 10 years worth of data, it’s pretty convincing that your
suspicions are correct.

There are quite a few other emails still sitting in the old Inbox
that I haven’t gotten to yet, so I apologize to those of you to whom I
haven’t responded. Several are very good questions that deserve a depth
of response that would fill a full column, so look for them in the
future. To the rest of you, thanks for writing in. Your questions are
always insightful and engaging and they allow us at here BP to do what
we love to do: talk about baseball until we can talk no more. Or at
least until we hit our column word limit.

You need to be logged in to comment. Login or Subscribe