Using Video Analysis for Strategy

Video analysis can be used to evaluate sports performance and technique. We discuss many of the biomechanics aspects and tools used to do performance assessments right here on this website. Video analysis however can also be used to provide detailed game strategy and analysis of on field play.

We see these game strategy statistics all the time when we watch almost any sport on TV. How many shots did Kobe Bryant take and make. What percentage of first serves did Roger Federer get in and how many resulted in winning points. These statistics are fun for the viewer, but they are essential to the coach and player in refining their strategy.

Let's take an example from basketball, as we are currently watching the NBA playoffs. Any team coming up against the Orlando Magic this year will need to find a way to get past Dwight Howard, who was once again named defensive player of the year. The teams employ a coach or video analysts to watch every play that Howard makes. Every time he blocks a shot or an opponent manages to beat him and score. Using this information they can determine how best to play against him.

Below is video of some of Howard's best blocks of last season. Notice how often Dwight is alone and far away from an opposition player as he gets into position for the block. This may seem necessary in order to make the block, or as the Orlando Magic play a zone defense he may not have a specific man defensive assignment. Coaches or video analysts though would look at this video and may be able to create plays that take advantage of his position on the floor.

Of course a team also needs to play to its strength and similar video analysis of the coaches own team can be used to determine how they may take advantage of any opposition mistakes, close any gaps or change any patterns that may be obvious and strategies the best approach to take on an upcoming game or player.

To do Video Analysis of this sort properly you need to have some basic equipment.

• You will definitely need a video camera and to set that camera up so that you can see the whole game. This can be difficult in sports with large fields such as soccer or rugby.
• For larger fields you can use 2 video cameras, with each one focused on, one half of the field.
• A great advantage in well televised sports is that much of the video footage already exists. A recorded video of an NBA game will allow you to track most of the players on the court for most of the game.
• Some video analysis software is also recommended. This software can be the same as the software that you use for performance and technique analysis, but the software must allow you to tag events. Tagging events such as the first serve will allow you to go back and look at all the first serves in one place. This will enable some statistical analysis of the results of the serve and if you have software that allows it you can also analyze each serve for technique and performance.

The website http://www.thevideoanalyst.com/ has some great information on video analysis of this type and everything that you may need.

When working with a team or player combining this type of video analysis with biomechanics and technique analysis will provide excellent resources for performance improvements wherever they are needed.

Measuring Your Vertical Leap Using Video Analysis

Before Kobe Bryant or Dwight Howard was considered for a contract in the NBA, they likely had to undergo a series of physical tests and exercises. In fact, most if not all professional and college sports programs require prospective players to complete these "combines", so that they can have a quantifiable way to compare the athletes. Some of the commonly used tests include the 40 yard dash, bench press, and vertical leap. For players in the NBA, the vertical leap test is the most crucial.

The vertical leap test demonstrates how high a player can jump and reach. Ideally, the test shows how high the athlete's center of mass (COM) moves. But since a person's COM is often difficult to determine, the test usually estimates the results by measuring the difference between how high a subject can reach from a standing position, and how high the subject can reach when they jump (either from a stationary or running start). One variation of this test is to simply see how high the athlete can jump and reach. While this latter method measures the maximum height the athlete can touch, it doesn't provide us with how high he/she actually jumped. In simple terms a player with an 9' reach, jumping 1' will reach 10'. A player with a 7' reach, jumping 3' will also reach 10', although it is obvious that the 7' player jumped higher. Simply measuring how high the subject can reach from a standing position before they jump, gives us the additional information needed to calculate how high the athlete jumped.

Professional and college organizations often conduct this test by having the athlete stand next to a tall pole that has little plastic flags extending from it, as shown below. The athlete then jumps and swipes the highest flags he can, and the result is recorded.

If you want to measure your own vertical leap but don't have access to one of these testing devices, you can simply film yourself jumping from a standing position. Besides the basic principles of video analysis (see http://videosportsanalysis.blogspot.com/2009/03/video-analysis-of-sports-101.html), there are two important concepts to keep in mind:

1. In your video clip, you must be able to see an object of known height. This object should also be in the same vertical plane the subject is jumping in. If you're jumping next to a wall, marking a known height on the wall should be good enough. The reason you need to have this information is so the analysis software can scale the distance measurement appropriately (this concept will be shown on the video below).
2. The video analysis software you're using must be able to calculate the distance between two points and be able to use the known height of the object to calibrate the measured distance accordingly; otherwise, the distance between the two points will simply be in units of pixels, which isn't very useful. We need to be able to convert the distance to something like inches or centimeters. If you're interested in finding out more about software packages that offer this type of feature, just let us know and we'll be happy to point you in the right direction.

Below we have a video clip of a subject performing the vertical leap test:



As you can see, we have one point identifying the highest point the subject reached while standing, and one point identifying the highest point the subject reached during the jump. As mentioned in the two concepts above, we have also shown the known height of another object (in this case, the distance from the ground to a point marked on the wall is 37"); and the software automatically uses this information to display the distance between the two points the subject reached in units of inches (23.7").

Compare this subject's results with the likes of NBA superstars Kobe Bryant (38") and Dwight Howard (40"), set to meet in the 2009 NBA Finals. Bryant has been vitually unguardable, draining shots while having defenders' hands in his face, and zipping around opponents with his lightning-quick first step. Howard is a physical force to be reckoned with; players move out of his way when he elevates, to avoid getting "posterized" when he throws down one of his thunderous dunks. It's amazing what these athletes are capable of; and using tests such as the vertical leap allows us to have a more complete understanding of how incredible they really are.