Author Archive

Curveballs Are Underutilized Early in the Count

I got the idea for this article thinking about pitching strategy. It makes sense to me that getting to two strikes for a pitcher is an important strategy for good performance. With two strikes, pitchers can get a hitter to swing out of the zone and either make bad contact or miss completely, two of the best possible results for a pitcher. The problem is, how does a pitcher get there without getting knocked around? If a pitcher throws a meatball down the middle in order to get early strikes, good hitters may take advantage and hit the ball hard before the pitcher can get to that good situation. So if a pitcher can throw a strike early, and maximize the chance a hitter chooses not to swing, that seems like the most effective strategy to get to this situation. The research below suggests that if this is the case, throwing a curveball high in the zone early might be a great strategy that almost no one uses.

I initially looked at first pitches going back from the beginning of 2016. I wanted to see which pitches had the highest swing rates on 0-0 counts. I was fairly certain that we would see fastballs with the highest swing rate. To my surprise, changeups have the highest swing rate, despite the lowest zone rate. Curveballs had the lowest swing rate. Below is the breakdown.

Changeup: 34%

Fastball: 29%

Slider: 29%

Curveball: 18%

The changeup swing rate suggests a well-placed changeup on the edges or out of the zone can be a good pitch to throw on the first pitch on occasion. However, with a curveball, you can throw it in the zone and not get a swing a large amount of the time. Given a pitcher’s goal to get to two strikes, the most advantageous count state for him, throwing first-pitch curveballs seems like a smart idea. However, this is not the strategy we generally see from pitchers. Below is percent of pitches thrown on first strikes.

Fastball: 60%

Slider: 14%

Curveball: 9%

Changeups: 7%

These frequencies suggest why changeups are so effective at getting swings out of the zone on 0-0 counts. Pitchers overwhelmingly throw fastballs early in counts, so when the changeup comes, it is very hard to distinguish it from the fastball, which a hitter will expect most of the time.

There are some practical reasons why pitchers throw mostly fastballs on 0-0 counts. First off, they are much easier to command, and as stated earlier, throwing in the zone and getting to two strikes is the main goal for a pitcher early in the count. Offspeed pitches, on the other hand, tend to have much more movement and can be harder to locate. Second, swings and misses aren’t a big deal without two strikes. Fastballs tend to have higher contact rates than offspeed pitches, but contact rates are much more relevant when whiffs lead to strikeouts.

But there are a few reasons why it makes sense for curveballs to be a go-to pitch early in the count. Some pitchers do locate the curveballs very well. Rich Hill is a great example. He famously throws his curveball about 50% of the time, throwing in the zone about 55% of the time the past three seasons. Throwing his curveball so often is probably why hitters swing so little against Hill despite his incredibly high rate of throwing the ball in the zone. Throwing his curveball, especially early in the count, may be a big reason behind Hill’s resurgence.

My next piece of research was looking at pitches high in the zone. I hypothesized that when pitches are located in the part of the zone that moves opposite to the pitch’s movement, hitters would swing less. For example, curveball breaks sharply downward, so a curveball high in the zone will look out of the zone to the hitter, therefore garnering less swings. I think this is logical and probably a well-known concept, but it was something I had never looked into.

I looked at all pitches thrown in the upper third of the zone on non-two-strike counts. Separating out curveballs and non-curveballs, the swing rates were vastly different.

Curveball swing rate: 26%

Non-Curveball swing rate : 65%

The results were overwhelming. There is nearly a 40% difference in swing rate between curveballs and non curveballs high in the zone. Hitters swing a lot high in the zone in general, but with curveballs they barely swing at all.

Very few pitchers utilize high curveballs without two strikes. The ones that do are a mix of bad and good pitchers. Of all pitchers who threw more than 200 curveballs on non-two-strike counts, Carlos Martinez had the highest percentage in the upper third of the zone, 15.3%. Hill is up there as well at 12.7%. But so is Paul Clemens at 14.6%, one of the worst pitchers in baseball. Jake Arrieta was the lowest at 3%, and he’s one of the best.

Early in the count, changeups and fastballs tend to have high swing rates, while curveballs tend to have low ones, especially high in the zone. Pitchers mostly use fastballs early in the count, but sparsely curveballs. While it makes sense to throw curveballs low with two strikes in the count to get swings and misses, this research suggests that a high curveball is an underutilized pitch early in the count.


Why Dylan Bundy Will Succeed as a Starter

(Originally written before last Sunday)

It was announced that Orioles pitcher Dylan Bundy will start this Sunday on the road against the Rays. The move makes sense — the Orioles need good starting pitching and Bundy could become a good starter. I think Bundy will do very well as a starter, and in this article I’ll talk about why.

Dylan Bundy’s career started with incredible promise. Drafted fourth overall, his first eight starts in the minors were punctuated by a 0.00 ERA and a 20/1 K/BB ratio. By the end of the year, he was considered the top prospect in all of baseball. The next few years were rife with injuries — first Tommy John surgery in 2013, followed by complications in his shoulder which caused him to miss almost the entire year in 2014. Bundy hasn’t looked like the same pitcher since. His fastball velocity this season started at 92 MPH, much lower than the high 90s we saw in the minors. But since the beginning of June, Bundy has made a remarkable turnaround. Since June 9, the numbers are beyond outstanding, with 14.1 IP, 19 SO, only 4 BBs, and 0 earned runs. But the peripheral stats are even better.

I am currently in the process of writing an article about how I think the most important skill of a starting pitcher is getting to two strikes quickly. Since June 9, Bundy has done this better than any pitcher in baseball. In the top 10: Clayton Kershaw, Max Scherzer, and Stephen Strasburg, arguably three of the best pitchers in baseball. This obviously is not to say that Bundy is one of the best pitchers in baseball; his track record is far, far too short to proclaim that. But it bodes well for Bundy that over the past month he is controlling the ball as well as baseball’s top pitchers.

Bundy’s fastball velocity is also encouraging. Bundy throws a rising four-seam fastball, which bodes well for his ability to miss bats. But at the low 90s, he wasn’t able to generate a lot of swings and misses, and as a fly-ball pitcher was susceptible to home runs. Last appearance, Bundy threw his fastball harder than he’s ever thrown it.

12917_P_FA_20160706

The chart may not look like much, but there’s a clear trend here: Up. His fastball velocity has increased over 4 MPH since the beginning of the season, which is a gigantic leap.

The Orioles desperately need starting pitching, and Bundy could be that answer. The Orioles do not have the worst starting pitching in the league. In terms of WAR, that is currently the Reds. But the Orioles’ staff is really bad, even if they look worse pitching in a hitter’s park. Chris Tillman is their only competent starter, while the rest of their rotation contain some of the worst pitchers in the league. So stretching Bundy into a starter seems appealing.

There is a risk that Bundy will be much worse as a starter. Pitchers are notorious for throwing harder in the bullpen than they would as a starter, and given that the majority of Bundy’s success has come at a higher velocity, it would be reasonable to assume Bundy will not be nearly as effective as a starter as he is as a long reliever. I think this is correct thinking; we should not expect Bundy to start and still average 11 K/9. But his numbers as a reliever have been elite, so there is a lot of room for Bundy to come down and still be a quality starting pitcher. Starting pitcher is where Bundy has the most upside, and the sooner he gains experience, the sooner we can expect him to improve.

Bundy will probably do well this Sunday, especially against a Rays team that strikes out the second-most in the league. Don’t think this is a mirage. Bundy has the stuff and command to succeed, and I think we will see that as a starter.


Are the Rays Swinging Harder?

The Rays are known as one of the most sophisticated organizations in the MLB, mostly thanks to an advanced analytics department. They have been first adopters of some of the now prevalent advanced baseball strategies today. They perennially are winners with annually low payrolls.

The Rays sometimes blow me away with the strategies discovered and implemented by their analytics departments. One of the most fascinating strategies they implemented at the beginning of last year was getting their pitchers to throw fastballs with more rise high in the zone, causing pitchers like Drew Smyly and Matt Moore to make drastic improvements in their results. Now I believe they are having their hitters implement a new strategy.

Swinging hard.

I cannot be positive they are telling their hitters to swing harder, but there is some evidence to lead me to believe this is true.

The Rays strike out a lot. Almost all their players have strikeout rates above their ZiPS and Steamer projections and they currently have one of the highest strikeout rates in baseball. A strikeout is the worst outcome possible for a hitter, so at first glance the Rays appear to have a lot of hitters who have gotten a whole lot worse. It’s clear looking at the data this isn’t just variance. Across the board for the Rays, the contact rates of their hitters have been much worse this season than the previous, an average decline of about 5%.

There are a few possibilities that come to mind that could explain the decreased contact rates. The first one is luck. It is possible most of the Rays contact rates have decreased because of chance alone. This is certainly possible, but also unlikely. The 5% decrease in team contact rate is by far the highest margin in the league.

Because of the degree of the contact rate change, it’s unlikely that the Rays’ worse contact rates are happening purely by chance. That leaves two possibilities in my mind. One possibility is the Rays have advised some or all of their hitters to take more of an uppercut swing. A steeper or uppercut attack angle of the bat theoretically should lead to less contact, so this is a possible explanation. If this were true, we would expect the Rays to have more fly balls from their hitters. And they do. Their fly ball rate is up about 3% from last year

But the increase in fly ball rate is only the sixth highest in the league, and can mostly be accounted for by the addition of extreme fly ball hitter Steve Pearce and the loss of extreme groundball hitter John Jaso. I’m also skeptical that a team would try to drastically change all their players’ swing planes. I can’t rule out this possibility though.

That leaves us with the explanation I believe to be true: The Rays have adopted a grip it and rip it mindset. The Rays currently have the highest ISO in baseball, meaning they hit for power better than every other MLB club, a 30% improvement year to year. They also have a large increase in hard contact percentage across the board, an average of about 5% per player, by far the highest increase in the league.

Hitting for power and hitting the ball hard are not unrelated. With MLB Statcast data, we can now see their is a clear and strong relationship between hitting for the ball well and hitting the ball hard. The harder you swing, the harder the ball will be hit. That is if you make contact at all.

2015_exit_velo_woba_qlqbqaxy_l47u9kzb

If contact wasn’t an issue, swinging hard would be a no-brainer. But there is a trade-off here. While a home run is the best outcome for a hitter, the strikeout is the worst. If you hit the ball in play, you can advance runners and get on base. With a strikeout, neither of those things will happen.

But is the increased power really worth increased strikeouts? The Royals would beg to differ. They won the World Series last year with historically good contact and strikeout rates. However, no one would argue that hitting was the biggest reason for the Royals success. On the contrary, it was really their bullpen and defense that carried them to a championship.

I can only imagine that the Rays have done the math and have decided: Yes, it’s worth the trade-off. Hitting the ball high and hard is good, and the Rays are doing that better than practically everyone else in the majors. Yes they are getting less contact, but the Rays do not have an abundance of talent in the batting department, so given their results I would have to say this change in approach has been a success.