The Knuckleball as a strikeout pitch

by
June 19, 2012

image
R.A. Dickey (US Presswire)

The knuckleball pitch is known for a few things, but control and strikeouts aren’t one of them. Of the best in history at throwing the knuckleball none has ever even approached striking out a batter an inning. The closest to that was Hoyt Wilhelm who threw 90 strikeouts in 93 innings of work in 1962. None of the rest even came close including Jesse Haines, Phil Niekro, Tim Wakefield or Tom Candiotti.

This season, however, after 103 career starts and 826 innings pitched R.A. Dickey has thrown a strikeout per inning through 13 games and 90 innings. Before this season Dickey struck out less than 6 batters every nine innings and had a season high of 134 strikeouts in 208.2 innings pitched last season.

Wakefield might be the only solid comparison during the Pitch F/x era and he threw what would be considered a batting practice knuckleball compared to Dickey. Wakefield had a knuckleball that averaged 66 mph in his career and would top out at 75 in his younger years. Dickey was around 70 mph a few years ago, but has been averaging 76 the past two years and has a “blazing” 84 mph fastball compared to Wakefield.

In the past two years Dickey has also weaned off his change up and is almost exclusively a two pitch pitcher. He throws 85 percent knuckleballs, which is very similar to Wakefield in his career. It’s tough to say if taking what Wakefield did and adding 10 mph to his knuckleball would have made him a better strikeout pitcher, but it could help explain his success so far. Dickey has only been throwing the knuckleball since 2005 and varied his usage from season to season.

Trying to look into the “movement” of the knuckleball is pretty much a lost cause. His average knuckleball has essentially no vertical or horizontal movement. Of course, the movement of each pitch varies wildly and it’s tough to say if he’s doing anything differently. It does appear to be hitting the zone more as his ball percentage is down from 34 percent in 2011 to 31 percent this season.

His knuckleball also has a 28 percent whiff rate, which is up from 21 percent in his career. It’s tough to compare him to anyone else as he is the only pitcher in baseball using the knuckleball right now. And other than Wakefield there hasn’t been another pure knuckleball thrown in baseball during the pitch f/x era according to Fangraphs.

It’s hard to see Dickey keeping up this level of strikeouts as not only has he never done this before, but no leading knuckleball pitcher has done this in baseball. Batters are swinging at pitches more out of the zone and making less contact on strikes. The easy assumption is that those numbers will return to normal, but since the knuckleball is so hard to break down it’s just possible that Dickey has it figured out. If he does and wins the Cy Young he would be the first pure knuckleballer in history to win the award.


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

Can you perform an analysis that shows how far off of center he is on average? Radial distance from the center of the strike zone? That might be a good way to compare knuckleballers, since they are also trying to hit the edges of the strike zone.

Is there data on the path the pitches are taking from release point to glove? Again, variation from the curve a similar MPH fastball that travelled from the release point to the glove point, wouild be another good way to judge relative movement of a knuckleball pitcher.

0
Alan Nathan
11 years ago

In thinking further about your article, perhaps I have more than a semantic disagreement.  If you look at the typical movement scatterplots for knuckleball pitchers, you see an amorphous blob.  The movement is all over the place.  You seem to argue that, therefore, there is very little information in such a plot.  I would argue that the essence of the knuckleball is contained in that plot.  The movement is essentially random from pitch to pitch and therefore not predictable by anyone. 

I have not looked at swing-and-miss statistics for Dickey or Wakefield.  I suspect someone else has done that.  An interesting question is the effect of the higher speed of Dickey relative to Wakefield.  Wind tunnel experiments suggest that the movement on a knuckleball is roughly independent of the speed.  If such is the case, then the advantage goes to Dickey, since the batter has less time to respond to the movement with the faster pitch.  So my expectation is that the faster knuckleball will induce more missed swings than the slower one.  But this is just “informed speculation” on my part.

0
Sylvan
11 years ago

It’s worth noting that the 76 MPH average understates what Dickey is doing, velocity-wise. He actually throws three distinct speeds of knuckleball: 78-81 MPH, 72-75 MPH, and 63-66 MPH. The fast knuckler is his main strikeout pitch.

0
garik16
11 years ago

Alan, your guesses seem to be somewhat backed up by Dickey—> he has gotten better whiff rates on the faster knuckler he’s thrown throughout the last 3 years than his slower ones.

But this is of course one example.

0
Fatbot
11 years ago

Two possible conclusions: Dickey is the greatest knuckleballer ever, or there’s something different with today’s MLB hitters. With due apologies to the Dickey fanbois, I will choose the latter.

These hitters grew up under the steroid era where HR are everything and if they strike out we get the Mark Reynolds response, “so what”. They have no fundamental skill of how to shorten a swing with two strikes to simply put the ball in play when necessary.

Watch the head/eyes of every batter in http://mlb.mlb.com/video/play.jsp?c_id=mlb&content_id=22397883&topic_id=28033182

Not one is looking to simply make contact, every one is falling down swinging out of their shoes, pulling their head not even looking at the ball. So we get Dickey looking more like Cy Young than Tim Wakefield.

And if you haven’t noticed, 2008-2012 have been increasing the league-wide K% every single year. Batters strike out more today than ever in the entire history of baseball: http://www.fangraphs.com/leaders.aspx?pos=all&stats=bat&lg=all&qual=0&type=8&season=2012&month=0&season1=1871&ind=0&team=0,ss&rost=0&age=0&players=0&sort=9,d

0
Mark Himmelstein
11 years ago

To an extent I think you have to account for Fatbot’s point here, but there’s one other thing Dickey’s doing that’s getting overlooked with his strikeout and walk dominance that’s also quite unique among knuckleballers. It’s actually something he’s been doing ever since converting to throwing the knuckler, and it further suggests he’s legitimately doing something unique even among his own species: he’s a ground ball machine.

Obviously there’s no discreet GB% data for most knuckleballers prior to Wakefield (and even then just the latter part of his career), but B-R does have GO/AO, for which guys like Niekro and Candiotti were in the 1.10-1.15 range for the most part, essentially neutral. Wakefield was an extreme fly ball pitcher, around 0.80. Dickey’s career rate is at 1.34, and since becoming a knuckleballer is actually higher than it’s ever been. Since joining the Mets, the lowest mark he’s had was 1.45 in 2011. In terms of discreet percentage, he’s been over a 50% ground ball rate since joining the Mets as well. It really seems like he’s found a way to throw the seemingly oxymoronic “heavy” knuckleball.

We don’t expect knuckleballers to dominante any of the three true outcomes, but Dickey is legitimately managing all three right now.

0
John C
11 years ago

Comparing Dickey to Wakefield is silly. Dickey was a real pitcher who actually made the major leagues as a conventional pitcher. Wakefield was a minor-league first baseman who couldn’t hit, whose career was salvaged when one of his managers noticed him throwing an effective knuckleball while horsing around in practice. He threw nothing but 65 mph knuckleballs because that’s all he could do.

Dickey throws the knuckler most of the time these days, but he throws it much harder than Wakefield, and he does have other pitches that are legitimate, if below-average. He’s not Tim Wakefield, he’s Tom Candiotti, another Quad-A caliber pitcher who took a huge step forward when he mastered the knuckler.

0
Alan Nathan
11 years ago

To say that his average knuckleball has no movement is very misleading.  The direction of the movement is essentially random, so any given pitch is as likely to have +x as -x movement.  If you average the movement over all his pitches, you get something close to zero, despite the fact that any given pitch has lots of movement.

0
Troy Patterson
11 years ago

That was what I was trying to say, but guess not very clearly.  Averages or pitch f/x graphs can’t tell much because of the randomness of each pitch, but yes their is tons of movement.

I was just stating that the average or collection of points on a graph can’t tell us much from year to year.

0
Alan Nathan
11 years ago

OK.  I just wanted to help clarify that point.  I think the RA phenomenon is the most interesting story in MLB so far this season.  Couldn’t come at a better time for me as my own research into lnuckleball trajectories is kicking into high gear.

0
Simon
11 years ago

Presumably fatbot can explain the difference in results for Wakefield and Dickey then. And the lack of success from other would-be knuckleballers like Haeger and Zink. Or did the hitters just forget how to hit in the last couple of years?

0