• Winning Air Guitar Strategy? Better do the Math!

    Posted by Cary MALLON on 9/15/2008 6:45:00 PM


    So you say that air guitar is all about putting together a fun and entertaining performance with your classmates and having the bravery to do some crazy stuff in front of the whole school but has nothing to do with math?  I say that’s only half the story.  See that? I already used a math term.  I’ve noticed some trends over the years and decided to take a close look at the 2007 performances to see what turned up.  Sure enough, there is some math that can explain how the judges received the performances.

     Faux Jimi Hendrix

    First let me explain the metrics I applied to this study.  I dissected the video with the following categories in mind:  Number of segments in each act, the median number of people on stage during the act, the number of appearances of the Eagle Mascot, the length of the act, and the amount of time devoted to ‘dialog.’  There are a couple of other sub categories that could have been considered, but I think I found enough revealing detail.  After taking data on these categories I found that another measure would be useful and wouldn’t require any more data collection.  That was the percentage of the act that was ‘dialog.’  This turned out to be quite revealing. 




    # On stage (ave)

    # On stage (med)




    % dialog
















































    The first category of data considered is the number of segments in the performance.   A segment is simply a part of a song that is lip synced  or danced to as part of the performance.  It turns out the each class and the staff chose similar numbers of segments for their performances.  Three performances had 11 or 12 segments and the other two had 6 and 8.  No ‘winning’ trend was observed in this category. 


    The next category measured the number of performers ‘on the stage’ or in the focus area.  This number changed each segment, so I chose the median number for the collection of each set of numbers for each class.   But, since it is common for all the performers for all the segments to appear in the ‘finale’ I decided to throw out the largest number in each performance.   There was a little bit of a trend here.  In fact, the judging criteria does reward participation.  What seems to have happened in 2007 is that one class that had few performers ended up in last place and among the other four competitors there wasn’t much of a trend. 


    Before moving on to the next category of data I want to comment on another aspect of the number of performers on stage at any given time.  It seems intuitive that the number of performers would be inversely proportional to the ‘tightness’ of the performance.  What does that mean?  The argument goes like this:  One person is more likely to be ‘in step’ than two people, two people are more likely to be in step than three people, etc.  The bottom line- the more people that join an act, the more likely it is that the act will look sloppy. 


    Further, it could also be argued that the number of segments a person is in is proportional to their probability of being out of step.  What does that mean?  It means that if Suzy Q is in two dances she has a greater chance of making a mistake than if she’s in only one dance, and she if she is in three dances she has a greater chance of making a mistake than if she’s only in two dances, and so on.  I don’t have any data on this and Suzy Q might be really talented and could do ten dances perfectly but the notion does have merit.


    Another statistic I measured was the number of times the Eagle mascot appeared in a performance.  I decided to include it because the Eagle seemed to be gaining popularity.  I don’t remember another time when it appeared in every performance except the staff’s.  The data doesn’t show a strong correlation between final rank in the contest and appearances of the Eagle.  However, staff and the class of 2011 were the highest finishers of those that will compete again in 2008, and the Eagle didn’t appear in the staff performance at all and only twice in the class of 2011’s act. 


    The length of each performance seemed to have little effect on the outcome of the contest.  There seems to be a standard being followed so most of the performances had similar lengths.   Nevertheless, one measure of time did seem to prove to be the undoing of some of the performances. 


    The percent of time devoted to ‘dialog’ turned out to be a make or break statistic.  The staff and the class of 2011 had the highest rank of returning performers, and had the lowest percentage of dialog.  The lowest rankings were given to the two classes had the highest percentage of dialog in their performances.   So, less dialog meant a better rank.  It only makes sense.  If you are spending too much time explaining the performance with dialog, you are not spending enough time entertaining. 


    One more statistic that performers would be wise to consider is this: The median age of the judging panel.  I don’t have specific personal data on the judges but I can say with confidence that this number was more than 50.  Student performers should be mindful that some of their culture references may go unnoticed by the judges.  That’s not to suggest that a performance should pander to the judges by including only references to 1960s or 1970s culture.  Although it could mean that the performance should be understandable to someone two or three times your age. 


    By now you’re thinking that all this statistical talk makes air guitar sound like some clinical and contrived exercise.  Perhaps you’re convinced that the staff has a formula for selecting our music and organizing the performance.  Nothing could be further from the truth.  I know some teachers promote the idea that we practice all summer or all winter but we probably rehearse the least of anyone.  Often, our performance only takes shape at the last minute and it’s never an exercise, instead a creative awakening.


    I will suggest that any class would do well to study our performances searching for pointers.  After all, we have won the contest frequently and consistently been near the top when we haven’t won.  That’s not to say students should try to do cover versions of what we have done.  But the classes who have beaten us have learned from what we’ve done and taken it to a new level. 


    Over the years air guitar has seen a lot of changes and the staff is responsible for many of them.  Many years ago we were the first to do a medley of songs instead of a single song.   We also introduced the ‘theme’ concept to our medleys.  We were the first to use computer graphics for background to enhance our act.  Some ideas we lifted from students but took to a new level are the use of recorded dialog and interludes between segments of the medley.  The student performances that judeges liked the best used ideas from staff performances and new creations by students.


    As we enter a new air guitar season it’s interesting to wonder what the new  shape of air guitar will be.  It needs something.  The last few years have seen the performances become more similar to each other rather than different.  I wonder what the new direction will be and who will come up with the groundbreaking performance.  Will it be up to the staff to show the way?  Or will it be the students who pull something new out of the hat?  Two things are certain.  You don’t have to have boys removing clothing or mention Mr. Kadell in your act to win.  I’m on the edge of my seat.




    Comments (1)


By Month