Visual io’s Baseball Visualization Tool
April 29th, 2007 | by Neal Levene |
Managers of all kinds are frequently faced with a decision to continue on the same path or choose a new direction. When you start breaking the decision into all of the factors that come into play, it becomes apparent how complex the human decision process is.
Visual io, a company that develops interactive visual analytics software, developed a visualiation tool that would help a baseball manager determine whether to go to the bullpen or continue with the existing pitcher.
The Visualization Tool is built around the basic idea that a pie chart can represent a simple yes a no decision. In this case, the decision is one often faced by managers in a tight baseball game. Should the pitcher be pulled from the game?
But this is a decision in which many factors must be considered. For instance, how many pitches has the current pitcher thrown? What is his history with the current batter? What is the gametime situation? In the Baseball Visualization Tool, these individual factors are represented by slices of a pie chart. In becoming the manager, you can input the gametime situation, choose from pitchers in your bullpen, and gauge the confidence level of the pitcher you have out on the mound. As you do this, each slice of the pie will either start to fill up, or empty out. The more full the overall pie chart is, the more you should pull the pitcher. Conversely, if the pie chart is relatively empty, then you should probably stick with the guy you’ve got.
This is certainly an interesting application of visualization. That said, pie charts are fairly ineffective visualization tools unless the goal is just to give a general “gestalt” feel for the data. People are pretty poor at estimating the actual volumes within the slices. I might have chosen an alternate visualization technique. It might also be interesting to know over time whether the data is suggesting a change.
It could be the case that the increased data makes the decision harder. What do you do when one factor recommends one action and another a different one. You end up at an experienced-based decision, perhaps with a better understanding of the current situation.
The biggest issue for me is would the tool’s interface improve decision making? The baseball decision used here has a relatively short decision frame, the time between one pitch and the next. As a manager, I would want an extremely simple interface that would give a recommendation and then give me a simple readout of the critical reasons that suggest the decision.
I struggle a lot with the fact that many decisions are made via intuition, some type of applying pattern recognition to a scenario. So how do I as a technologist clarify the current situation for the decision maker? It is pretty hard at the end of the day. It relates more to human factors than technology.
This is certainly a fun simulation, and one to which many people can relate. Click here to see the full demonstration. I had not previously heard of Visual io, and their tool looks interesting.
Popularity: 8% [?]
