Computing and The Super Bowl

Most people think of sports as the antithesis of computing.  Computing is all indoor activity, staring at a screen and sports is outdoors, running around on a field or track and usually involves the use of balls, nets, rackets and other equipment.  But modern sports involves a lot of computing power.  There’s a ton of CS in just a single game.

Let’s start with the football.  While there’s still some hand work involved in making a football, much of the process is done by machines that are calibrated and run by computers.  Most manufacturing of any kind today involves specialized robotics to put things together.  Designs for the templates and the machines themselves are often made using CAD software.  Check out this video and see if you can spot the computing.

Now let’s move on to the field.  Fields must be carefully groomed and accurately marked, which is often done by robots or other computerized methods, but what most of us watching at home see is that magic first down line.  Creating that line uses a variety of computing tools.  First, a 3D model of the field is created, because each field is slightly different.  Next, the color of the field is recorded as the shades of green also vary, and the first-down line relies on green-screen technology to work.  Camera computers send position information to another computer, where a person basically right-clicks in the correct location to generate the line.  Wikipedia has a decent explanation of the process:

Each set of camera encoders on a camera transmits position data to an aggregator box that translates the digital information into modulated audio where it is sent down to the corresponding camera computer in the truck. This data is synchronized with the video from that camera. At the camera computer the camera position data is demodulated back to digital data for use by the program that draws the “yellow line” over the video.

 

Separately, the chroma-keying computer is told what colors of the field are okay to draw over (basically grass) and that information is sent to the camera computers.

 

That’s a lot of computation for one line!

The broadcast itself also involves a ton of computing.  Nowadays, the SuperBowl is live streamed in addition to being broadcast on regular television.  And it’s being streamed for free, so the algorithm to handle that many simultaneous streams is going to be complex.  According to the senior VP of digital media for NBC, Rick Cordella, the live Super Bowl stream will be available at variable bit rates ranging between 500 kilobits per second up to 5 megabits per second, delivered as an HLS stream.

Fans in the stadium won’t be able to stream the game, but there is an app available for them, allowing them access to those famous commercials and to different camera angles (http://bit.ly/167zQCA)

For many footballs fans, besides watching the games on television, participating in fantasy football online is a huge part of their interaction with the game.  All of that involves some serious programming, using data collected from real games to create the outcomes for the fantasy ones.  The fact that there is data to be collected that’s available through an API is of course, another way the game uses computing power.  Many teams use that data to improve their performance and select players for the next season.

So while you’re watching the Super Bowl tonight, or watching any sports game, really, think about all the computing power that makes the game possible.