Modern media controls how the average fan views the game of baseball.
When they say that, "90 percent of the game is pitching," we tend to believe them. When they say that a player "is on a hot streak," we accept that as fact.
When they say, "this guy is great at working the count," we assume they are implying that walks are greatly beneficial to batters and disadvantageous to pitchers.
Is that really the case?
In this mini-series I will take a look at how bases on balls effect both a pitcher and a batter, and determine if they really are as important as the media would have us believe.
Up first, how important are walks in regard to the success of a pitcher?
(A disclaimer: The following pages include numbers, graphs, and other things some may or may not find harmful to their health. While I'm by no means a mathematics expert (I'm just a junior in high school, after all...what could a punk like me possibly know?) I do use (without proper explanation) a few terms that some may not know; such as r-squared, for example. That said, I again don't claim to be an expert. I'm nothing more than a passionate baseball fan.)
There are many ways to view this question. Let's start off modestly, though not exactly small, with every starting pitcher from the last five years, with at least 162 innings logged per season, and take a look at how walks issued correlates with ERA, the grand-daddy of pitching metrics.
This chart tells you exactly what you'd think a blob of blue squares would tell you: virtually nothing. While there is a slight increase in ERA when walks per nine innings increases, the correlation (a r-squared of just 0.08) is almost non-existent.
This is very puzzling. A pitcher with an ERA over 4.00 generally allows as many walks per nine innings as a pitcher with an ERA under 2.00. Small sample size isn't a major issue, as this is five years' worth of data; over 400 individual seasons.
If I were to stop here, the obvious conclusion would be that walks play no role in the success of a pitcher. But let's take a step back and take a look at things with a broader perspective.
Here is the relationship between the total number of walks allowed by a team over the course of a single season and the resulting team ERA, from 2004 to 2008.
Now we're starting to see some correlation. While 0.32 is hardly a solid enough number to accurately predict, it's enough for the kind of data we're dealing with.
So now we know that the number of walks a team issues has at least a marginal say in the overall success of a pitching staff.
Team wins, however, are the ultimate measure of success. Although there are dozens of other factors that play into the total number of wins a team has per season, pitching is perhaps the most important.
Here is the relationship between total number of walks allowed by a team's pitching staff and the number of wins that team had that season. (Again, this is data from the five years between 2004 and 2008.)
Just what you'd expect: a negative relationship between walks allowed and team wins. The r-squared is slightly higher than when comparing walks allowed to team ERA, meaning the number of walks allowed by a team has a greater, albeit minuscule, impact on team wins than team ERA.
Let's see if another form of evaluating pitchers (WHIP) stands true to form when compared to walks allowed. Of course, virtually half of the entire WHIP metric incorporates walks, so, if all goes according to plan, we should see a huge jump in positive correlation.
Sure enough, the r-squared is 0.58 - about as high as I can reasonably expect any statistic to take it. This doesn't really tell us much that shouldn't already be obvious, though, unless we take a look at how WHIP correlates to team wins...
That negative correlation is somewhat significant as well. So, via the ever-helpful transitive relation, the allowance of walks plays a role in the overall success of a team. Less walks equals better pitching equals more wins.
Let's take a quick look at another statistic that is widely accepted (at least in the sabermetric universe...good luck getting a broadcaster to use this) as being capable of evaluating a pitcher's true talent level is FIP. This metric eliminates elements out of a pitcher's control and places more emphasis on those in his control.
A league factor is also added to the total. It is usually around 3.2, so that is what I added in all cases. These FIP totals won't be perfectly exact, but they are close enough for our purposes.
Because such a heavy emphasis is placed on walks in this formula, the resulting correlation between walks allowed and FIP is obviously significant. Let's take a step away from team totals and focus more on the individual pitcher.
There are 412 blue squares in that graph. Why can't you see all of them? Because most of them fall extremely close to that trend line, meaning there is an extreme correlation between walks allowed and FIP. An r-squared of 0.86 is extremely rare in these cases, and I'm convinced we've found our answer.
(All statistical information used in this article was taken from the Baseball-Databank.)