Wednesday, June 3, 2009

On Beer and Bowling

The consumption of alcohol is frowned upon in most sports, but tolerated in others. Golf carts, for example, are commonly used for the transport of six packs. In even fewer sports is drinking mandatory, bowling being the clearest example. Given the traditional relationship between the consumption of fermented hops and barley and the rolling of 11 to 15 lbs balls, it may be surprising that the scientific community has yet to carefully examine the relationship between alcohol consumption and athlete performance on the hardwood. This is the intent of this study.

Based on a nonsystematic survey of bowling alley patrons, there are at least three common models concerning the effects of increased blood alcohol concentration (BAC) on bowling scores. We can refer to them as: 1) Bad, 2) Good, and 3) the Carla postulate.

Beer is bad- The first model is perhaps the most obvious and states that as you drink more, your skills as a bowler will decline. It is well known that alcohol impairs motor function and reaction time, both of which should be detriments when bowling. Thus, from the first to the last drop, we might expect to see progressively declining bowling scores.

Beer is good- Although the reasons why are unclear, some bowlers will attest that beer is a performance enhancing substance. The most obvious explanation relates to mental aspects of the game. In the fierce competition of the bowling alley, knocking a few back might relax the bowler, thus eliminating or mitigating the problem known as “over-thinking your throw”.

The Carla postulate- The most widely held belief about drinking and bowling is that beer consumption is both good and bad for bowlers. In this framework, beer consumed in moderation actually improves the performance of the bowler, but when one approaches the stage of downright sloppiness, bowling scores drop dramatically. This model is often expounded in the game of barroom billiards as the well known “somewhere between drunk and buzzed” hypothesis. In this paradigm, it is widely believed that performance is maximized at an easy to attain but difficult to maintain state between drunk and buzzed, or a moderate BAC. I call this idea “the Carla postulate” because it is best exemplified by a model, reproduced in Figure 1, which was constructed by a Ms. Carla Young and popularized by the blog “Graph Jam”.1

In pondering this critical frontier of scientific knowledge, it became clear to me that an opportunity was at hand on Monday nights during the Bernaski Memorial Bowling League at the Laramie Lanes bowling alley in Laramie, Wyoming. The weekly competition of our four man bowling team, the Bowl Movements, provided an opportunity for actualistic research of this nature. Over a 16 week period, I recorded the number of pitchers of beer consumed by the team,, frame by frame bowling scores, and who was the designated driver for the team that week.

These data permitted three independent means of assessing the effect of beer on performance in the sport of bowling:

Analysis 1- Do average bowling scores increase, decrease, increase then decrease, or stay the same from Game 1 through Game 3? This analysis is based on the assumption that BAC progressively increases through the night from Game 1 to 3. In other words, BAC is at its minimum at the start Game 1, and it is maximized at the end of Game 3. This assumption is strengthened by personal experience and observation.

Analysis 2- What is the relationship between the absolute amount of beer consumed and team average for the night? Beer consumption for the team ranged between four and seven pitchers depending on several factors, most prominently weather, tides, and mood. In contrast to Analysis 1, which uses a relative measure of beer consumption, Analysis 2 is based upon the absolute amount of beer consumed.

Analysis 3- Do designated drivers bowl better, worse, or the same as non-designated drivers? The performance of the designated driver can serve as a control. Designated drivers did not completely abstain from beer consumption, as that would be a rule violation, but they kept it to a responsible level.

In Figure 2, I show the average number of pins recorded by the team for Games 1 through 3 for the 16 weeks of the study. Average game scores do decrease progressively from game to game, but these differences are not statistically significant (One-way ANOVA, df=47, f=0.041, p=0.96). From Game 1 through 3, team average pins drops ten pins from 147 to 137, but the change is too small to be considered meaningful.

Likewise, there is no apparent relationship between the number of pitchers consumed and the average game score for the team. As shown in Figure 3, bowling performance appears to remain steady no matter how much alcohol is consumed per bowler. Again, any differences that are evident are not significant (Pearson’s r = 0.167, df = 14, p>0.1).

Does relative abstention from beer consumption improve one’s game? Given the prior results, it should be no surprise that volunteering to be the designated driver will not be advantageous. The team was evenly split between bowlers who bowled better while drinking and those who bowled better as designated drivers. For the team as a whole, bowling scores of designated drivers were approximately three pins higher than designated drinkers (147.6 vs. 144.8), but again this difference was not significant (Unpaired t-test, t=0.671, df=244, one-tailed p = 0.25).

Which of the three common models of beer consumption and bowling performance is correct? In short, none of them are correct. Beer apparently neither helps nor harms bowling scores, but one caveat must be made. It seems readily apparent that given enough beer, bowling scores must eventually decline. It is difficult to imagine a bowler rolling a 200 game if he or she has imbibed themselves to oblivion. At some point, scores must come down, but in this study, the participants apparently never reached that point. This final observation points to clear avenues for future research.

References cited