There are lots of myths about the knuckleball. It is the most rare and least understood pitch in baseball. Despite its name, most fans are now well aware that the knuckleball is actually thrown with the pitcher’s finger nails. Another curiosity about the knuckleball are reports that it can throw off a batter’s timing and rhythm and mess up his swing for the next few games afterwards. Pete Rose, the all-time leading MLB hits leader, once said about the knuckleball pitcher Phil Niekro ‘I work for three weeks to get my swing down pat and Phil messes it up in one night’. The 1972 AL MVP Richie Allen said about the knuckleball ‘I never worry about it. I just take my three swings and go sit on the bench. I’m afraid if I ever think about hitting it, I’ll mess up my swing for life’. Despite hitter’s claims that the knuckleball messes up their swings, is it actually true? To see what the statistics say about this myth I took a look at knuckleball pitcher and 2012 NL Cy Young award winner R.A. Dickey’s 2013 season with the Toronto Blue Jays.
To determine if Dickey’s knuckleball was throwing off hitter’s swings allowing the pitcher following him in the rotation to have an added benefit I compared the starting pitcher’s stats in the games after Dickey’s 34 starts when they faced the same team Dickey pitched against, or a different team. If Dickey does not pitch the last game of a series, the next starter will face the same hitters (or most of the same hitters) as Dickey did, whereas if he pitches the last game of a series the next starter will face a new team. Presumably, if the knuckleball does mess up the batter’s ability to hit other pitchers, then when pitchers are playing the same team following Dickey their stats will be better than when they are pitching against a different team who did not just have to try to hit a knuckleball.
Last season there were 23 games after Dickey pitched that were against the same team that Dickey pitched against. In these 23 games the Blue Jays starting pitchers allowed 56 earned runs in 129.1 innings pitched for an ERA of 3.90. In the 11 games following a Dickey start where the starting pitcher was facing a different team they allowed 46 earned runs in 45 innings pitched for an ERA of 9.20. In the average start for a pitcher after Dickey and against the same team, a starting pitcher would allow 2.43 earned runs and pitch 5.62 innings pitched, but against a different team allow 4.18 earned runs and only pitch 4.09 innings, which are both statistically significantly different from one another. This strongly supports the theory that the knuckleball is messing up batter’s swings to allow the pitcher following Dickey in the rotation to be more successful when facing the same line-up.
Since the Blue Jays had so many injuries and complications in the 2013 season leaving the starting rotation in disarray at times, there were 9 different starting pitchers who started at least once in the game following an R.A. Dickey start. We can divide their starting pitching stats into two categories for comparison; the starts after an R.A. Dickey start and against the same team, or the starts after an R.A. Dickey start but against a different team and their starts not after a Dickey start. However, two of these pitchers, Esmil Rogers and Mark Buehrle, only had one start after R.A. Dickey which was against a different team from Dickey. So we cannot compare their stats since they don’t have starts in both categories. The other 7 pitchers we can compare to see if they perform better when they are pitching after Dickey and against the same team than when they are not, and the results are shown in Table 1. As you can see, 6 out of 7 pitchers had a better earned run average when pitching after Dickey and against the same team (which I will now call the ‘Dickey Boost’ starts) compared to their other starts, with the only exception being Chad Jenkins who had the same ERA in both categories and only had 3 games started (GS) all year. Only two pitchers had at least 4 starts in both categories, Brandon Morrow and J.A. Happ. Morrow’s ERA was 2.73 in 4 Dickey Boost starts last season compared to 8.36 in his 6 other starts, and Happ’s was 3.55 in 9 Dickey Boost starts compared to 5.79 in his 9 other starts. Clearly these pitchers are seeing an added benefit by being able to pitch against the same team that R.A. Dickey just pitched against.
Pitcher After Dickey Same Team Different Team or Not After Dickey Morrow 4GS, 2.73ERA 6GS, 8.36ERA Ortiz 2GS, 4.82ERA 2GS, 6.43ERA Jenkins 1GS, 3.60ERA 2GS, 3.60ERA Wang 4GS, 4.89ERA 2GS, 23.57ERA Redmond 1GS, 3.60ERA 13GS, 4.49ERA Johnson 2GS, 5.90ERA 14GS, 6.24ERA Happ 9GS, 3.55ERA 9GS, 5.79ERA
Table 1: Games started (GS) and earned run average (ERA) for starting pitchers for the game after an R.A. Dickey start and against the same team, or starts after a Dickey start and against a different team and starts not after an R.A. Dickey start in the 2013 MLB season.
We can also compare the 23 Dickey Boost starts from last year to the starting pitcher ERA for the rest of the Blue Jays 139 games. In the 139 non-Dickey Boost starts the starting pitching ERA from last season was 4.97. So the starting pitching ERA is more than 1.00 whole point lower in the games after Dickey pitched when facing the same team than all other starts. The starting pitchers should not be the only ones benefitting from Dickey’s knuckleball. If hitter’s swings are out of whack then they should also have trouble hitting the Blue Jays’ relief pitchers and so the team’s ERA as a whole should be better in games following a Dickey start that is against the same team. In the 23 Dickey Boost games last season the Blue Jays team ERA was 3.59 and their record was 12-11 which is a .523 winning %. In the 11 games after a Dickey start that was against a new team the team ERA was 5.72 with a record of 2-9 or a .182 winning %. Again, we can compare those Dickey Boost games against the stats for the 139 other games. The team’s overall ERA in the other 139 games last season was 4.36, and their record of 62-77 was good for a .446 winning %. The Blue Jays are clearly more likely to win a game when facing the same team that Dickey just pitched against.
This boosted winning % does not just last for one game. The hitters need more than one game to get their swing back to its normal rhythm. The second game the Blue Jays played after one of Dickey’s 34 starts last season was against the same team that Dickey faced 11 times, and against a different team 23 times. In those 11 games versus the same team Dickey pitched against the team ERA was 3.53 with a 7-4 record or a .636 winning %. In the 23 games when the Blue Jays were playing a different team their ERA was 4.29 and their record was 11-12 which is a winning % of .478. It is difficult to say if the effect lasts for 3 games or not. Theoretically, for every game after a Dickey start the batters should get a little back into the swing of things and the Dickey Boost may wear off by 3 games after his start. Dickey only pitched the first game of a 4-game series once last season. Three games after that start, or the final game of the 4-game series, the Blue Jays allowed 11 earned runs in an 11-1 loss to the Tigers, but it is difficult to make any conclusions from such a small sample size. However, the statistics clearly show that the Blue Jays overall pitching is improved in the Dickey Boost games (which I will now use to refer to any of the next two games after Dickey, and against the same team).
We should also expect the Blue Jays relief pitchers to have better stats when pitching in the same game as Dickey. We can split the Blue Jays relief pitching stats into 3 categories; pitching in the same game as R.A. Dickey, pitching in a Dickey Boost game, and pitching in all other games. Relief pitchers pitching in the same game as R.A. Dickey had 75.1 innings pitched with 28 earned runs for a 3.34 ERA. In the Dickey Boost games relief pitchers had 120 innings pitched with 38 earned runs for an ERA of 2.85, and in all other games relief pitchers had 357.1 innings pitched with 141 earned runs for an ERA of 3.55. These stats suggest that relief pitchers are seeing a benefit the next couple of games after the knuckleball, but not immediately afterwards, which would seem strange. I argue that there is one reason for this discrepancy: Dave Bush. Last season, Dave Bush played one game in the major leagues. One game. He pitched 3 innings and gave up 5 earned runs and hasn’t pitched in the majors since. It just so happened that this one terrible game came in relief of an R.A. Dickey start on April 7th. If we assume that Dave Bush would have been terrible no matter whom he pitched after and it was unfortunate for R.A. Dickey that it came in one of his starts, then we can remove his stats from the relief pitching in the same game as R.A. Dickey category. This would leave us with 72.1 innings pitched with 23 earned runs for an ERA of 2.86. This is almost identical to relief pitching ERA in the Dickey Boost games of 2.85. This suggests that relief pitching benefits for 3 games after an R.A. Dickey start when it is against the same team.
Now those of you who play fantasy baseball are probably already thinking ahead to try and steal whichever Blue Jays pitcher will follow Dickey in the rotation next year to get the Dickey Boost starts. But, the more important problem is how can the Blue Jays improve their chances of success next season? The perfect situation is where Dickey pitches the first game of every series, maximizing the number of Dickey Boost starts. Since the All-Star break allows teams to reorganize their rotation we can split the season into pre and post All-Star break to find the rotation spot in each section to maximize the number of Dickey Boost games. However, indications from last season show that this may be more complicated than it seems. Due to all of the injuries last season the Blue Jays struggled to find a quality 5th or even 4th starter in their rotation and often had fill-in pitchers scattered through their rotation. They took advantage of rained out games and off-days to skip a spot in the rotation 5 times just before the All-Star break alone. This makes it difficult to plan out Dickey’s starts based on a 5 man rotation if they might skip a starter from time to time, and in doing so move Dickey into a different spot in the rotation. As a side note the Blue Jays did a terrible job of maximizing Dickey Boost starts before the All-Star break last season. In the pre All-Star break section of the season they ultimately finished with 16 Dickey Boost games, when if he had been in the 3rd spot in the rotation, where Mark Buehrle was, they would have had 25 Dickey Boost games. They could have had an extra 9 games for hitter’s to whiff at baseballs, trying to forget about the knuckleballs they just faced. Just before the All-Star break Dickey’s last start was scheduled for the first game of a series against the Orioles, the last series before the break. Due to a day off earlier in the week they moved him up by one start to pitch the last game of a series against the Indians, missing out on two Dickey Boost starts against an opponent from their own division. They may have figured things out after the All-Star break, or they just got lucky. If they did not skip any spots in the rotation then pitching Dickey in the first game after the break would have resulted in 18 Dickey Boost games, the most of any of the 5 spots in the rotation. However, the Blue Jays pitched Dickey in the third spot in the rotation where he was on pace for 15 Dickey Boost games. By skipping a pitcher in the rotation twice on the August 29th and September 9th off days Dickey ended up with 18 Dickey Boost games. The same number as if he had started in the first spot and they didn’t skip any pitchers, but if you have injury problems and this allows you to skip a low quality pitcher twice then it may be for the best. I don’t know if the Blue Jays planned this out and plan to order their rotation accordingly for the upcoming season, or if they had their own reasons for holding Dickey back until the 3rd start after the break and skipping rotation spots.
So for the upcoming 2014 season, where is the best spot in the rotation for Dickey so the Blue Jays can maximize on their Dickey Boost starts? Unfortunately, the way the schedule works out next season most spots in the rotation are fairly similar and the Blue Jays can only improve their Dickey Boost starts by a handful of games. Before the All-Star break the second spot in the rotation results in the most Dickey Boost games of 23, which is only 2 ahead of the first and third spots in the rotation. After the All-Star break the first spot in the rotation results in the most Dickey Boost games of 16, which is again 2 more than the next highest. This season it may only influence a few games, but that may be the difference in the Blue Jays still playing baseball when the Maple Leafs start their season for once. The great news is that if the Blue Jays were to make the playoffs, in a 7-game series if Dickey pitched games 1 and 5 this would give 4 Dickey Boost games in addition to the 2 games that you already have your best pitcher, Dickey, pitching in. It would be very difficult for the same team to go back and forth from knuckleballs to other pitches over the course of 7 games. But again, these numbers for the season only hold if they don’t shuffle around their rotation. If the Blue Jays do decide they can skip spots in their rotation like they did last year, it is possible that it may actually be to their benefit. For example, after the All-Star break if you started Dickey in the 4th spot in the rotation and then skipped whoever’s turn it was in the rotation at each of the days off on August 14th, September 1st, and September 11th this would result in 20 Dickey Boost games. That is 4 more than the 16 from Dickey being in the 1st spot in the rotation, which was the most if you kept Dickey in the same spot the whole time. But this might throw off the rotation and result in skipping some quality starting pitchers. The important thing to note from all of this is that whenever the Blue Jays decide to move around the starting pitcher rotation they should determine how it will affect the Dickey Boost starts. If Dickey gets injured and is coming back from injury, they might want to hold him back a start or two from when he feels ready in order to maximize the Dickey Boost games. Or if they decide they want to skip the 5th spot in their rotation, they should look at if this will decrease their remaining Dickey Boost starts. The Blue Jays need to make sure that after the knuckleball has taken away the batter’s timing and rhythm, those confused hitters are still facing Blue Jays pitching and not helping out some other team get the Knuckleball Boost.
As stemac and I were going back and forth about this article another variation on the same theory was posted to fangraphs. You can find it here. Its conclusion is similar, namely that Dickey has value beyond his starts.