Saturday, August 9, 2008

Ground Ball Grafix: The F/X of the GB

When my dad called to tell me his thoughts on the website last week, we had a conversation that resulted in the creation of the TPC Reference Guide I posted last night. After we covered some of the basics, he wanted to know why I was so obsessed with ground ball rates and why I gave ground ball pitchers so much credit. We went back and forth on the benefits and shortcomings of ground balls from a defensive standpoint, but I'm not sure I explained my reasoning to him all that well.

One of my dad's primary objections to ground balls was that fly balls are so much easier to turn into outs. I couldn't really argue with him on that point, since fielding a grounder involves several steps and multiple players before an out is recorded. I remembered that somewhere between 70-80% of ground balls resulted in an out, but I had no clue what percentage of fly balls turned into an out. My dad guessed about 90% or more, and I guessed somewhere just south of that. After a bit of research, here's what I found.

I only used numbers from the American League (all stats from 2007), but the National League is pretty much the same. On average, a major league team will turn 74% of ground balls into outs, and 83% of fly balls into outs. So, in a way, my dad was right. Ground balls don't create as many outs as fly balls. Here's a graphical representation of the above stats for those of you who like to see instead of read things.

Now that we have the basic numbers nailed down, we still have to figure out why ground balls are better for a pitcher than fly balls, despite the fact that fly balls are easier outs. In order to answer that problem, we have to look at this from a different angle.

In a general sense, a pitcher's job is to create outs. But in more specific terms, his job is to eliminate the creation of runs by the other team. That means that anything that could produce a run becomes the common enemy of all pitchers, including hits, walks, errors, home runs, and so forth. However, it is quite obvious that just as a strikeout will produce an out 100% of the time, a home run will also produce a run 100% of the time. And just like a ground ball doesn't always produce an out, a single or walk doesn't always produce a run. So, there are essentially a variety of levels of enemies a pitcher must deal with, and they are not all created the same.

What does this all mean? It essentially means that fly balls and ground balls don't generally result in the same number of runs. In fact, there's quite a difference between the 2, as you can see in the following chart.


Once again, the graphical representation of the above chart.

On average, a fly ball results in almost 4 times the number of runs as a ground ball. And if you think about it, it makes sense. Most extra base hits (like home runs, doubles, etc.) are the result of fly balls, while ground balls rarely create extra base hits.

If we translate the above numbers into real life, we can see more clearly why fly balls are such a bad thing for pitchers. Let's say we have a group of 7 pitchers, each exactly like the other in every way. They have the same pitching repertoire, the same velocity and movement on every pitch, the same strikeout rates, the same walk rates, the same defense behind them, and so forth. However, they all have a different ground ball rate. The best of them has a 65% GB%, while the worst has a 35% GB%. Per 100 batted balls they allow, how many runs will each of them give up if we go by the MLB average of runs created per ground ball and fly ball?

With a 30% drop in ground ball rate, the a difference of almost 4 extra runs is allowed by the 35% GB% pitcher as compared to the 65% GB% pitcher per 100 batted balls. It may not sound like a lot, but it's a difference of 40.8% more runs allowed by the worst pitcher than the best pitcher on our list. In a season of 162 games, the difference in runs allowed becomes a major issue.

Using a rough estimate, if a pitcher gives up 500 batted balls over the course of a season, that could be a difference of 20 extra runs over 20 to 30 starts. That means our 35% GB% pitcher is giving up an average of close to 1 extra run per start compared to the 65% GB% pitcher. The Braves, who were 4-21 in 1 one run games through July 1st, could certainly have benefited this season from a Chuck James with a 65% GB% instead of the ghastly 28% GB% they've seen from him.

In conclusion, ground balls don't always create the same number of immediate outs as fly balls, but they certainly reduce the numbers of runs an opposing offense can create. Not to mention the fact that by giving up fewer runs a ground ball pitcher can stay in games longer, allowing relief pitchers to carry a lighter load throughout the season, while also giving their team a greater chance to win day in and day out. And that's why if given a choice between 2 seemingly identical pitchers, I always go with the pitcher with the higher ground ball rate.
...Read more

5 comments:

Mike said...

If GB's are turned to outs 74% of the time and FB's are turned to outs 83%, how is the normal BABIP at .300? Are line drives not included in either FB or GB?

Adam G said...

Line drives are considered separate from ground balls and fly balls. I don't have the numbers in front of me, but I believe line drives create the most runs and create the highest BABIP of the 3 types of batted balls.

Sam Baker said...

could you please do a top 5 pitching and hitting prospects for the nationals?

Skewtosis said...

Excellent post! It's interesting to consider how GB% affects a pitcher's ERA.

One way is to assume that a decent pitcher gives up just over 1 hit per inning, so that 100 batted balls corresponds to about 90 innings (say). A difference of 4 runs over 90 innings then translates to a change in ERA of about 0.5.

A (slightly) better way is to take QERA (a function K%, BB%, and GB%) and fix the K% and BB% rates to their league-averages, then plot the functional dependence of QERA on GB%. I find a drop in ERA of about 1 between the GB values of 35% to 60%.

Joel L said...

I would definitely take a higher GB% type of pitcher over a high FB% guy.

As an A's fan I had the pleasure of seeing Tim Hudson and Barry Zito during the their prime years. Both of these guys were studs, however, Hudson always had a much higher GB% than Zito. Although Zito did have one great year (2002), Hudson has always been a more consistent pitcher due to his high GB%. We all know what's happened to Zito.

Looking at their careers since leaving Oakland, Hudson is by far the better of the 2. I just hope he is able to bounce back after his elbow injury. I hate seeing former A's stars struggling, especially a class act like Hudson.