After a great deal of research, thought and consideration, I feel as though my project idea is beginning to gain clarity, at least with myself. This week I read a short article discussing the use of GIS technology for the purposes of studying sports. The article primarily focused on peripherals of sport, such as broadcasting patters, which fall somewhat outside of the intent of my project. My project is looking to concern itself far more with the actual sport aspect of hockey. That being, player statistics, game schedules, etc. However, as I read through the article I had somewhat of a eureka moment.
The article discusses using GIS as a tool for examining travel distances in game schedules, specifically how this dictates how broadcasters will determine rights and territories. I have other plans for this however. What I would like to do is map a season of NHL hockey for a given team. I want to document the journey of the season, with each stop identifying a game number and game statistics. My hope is that will create a catalogue of hockey information which can be referenced and examines for patterns and tendencies.
For example, I have been considering taking the Toronto Maple Leafs. In order for this to produce any practical application, its likely that I will have to do multiple seasons; 10 seems like a nice round figure. The idea is that players have individual tendencies. I would like to use travel distance as a barometer for performance, and potentially a gauge for forecasting how a certain individual will likely perform in similar conditions. The primary factors I will consider will be the age of the player and other players, as well as statistical performance on a trend and total distance of the journey to each point.
Some of the findings I hope to see are things like declination of performance based on distance travelled and age. My theory is that players who are over 30 years old will show a less moderate deterioration in their effectiveness on the ice as say a 26 year old.
The biggest challenge I encountered this week as researched and pondered was redundancy. I do not want to simply make a project, or reformat information that’s ubiquitous online. I want my project to serve a practical application. Obviously, forecasting the performance of a player in an absolutely correct fashion is impossible. My hope is simply that my project can serve as a tool to illustrate some patterns, and from those make a reasonable prediction about how the games will go and who will perform best.
I have been thinking all week about how to create this, and it seems that creating some kind of website, which incorporates GIS is the best avenue. Preferably this would be of the sort which people could contribute too, however this is simply another hurdle I must overcome. I have considered the amount of data I have saddled myself with, and although it is vast, I do not believe it will be overwhelming. I have kept in mind things such as shutouts, which will require zero data inputting depending on who was shutout, which should help me.
I feel as though, with this project I have confronted my biggest fear which was redundancy. Differentiation is paramount on my checklist of project requirements, and that has been difficult to address considering the plethora of information, statistical databases and web sites covering and analyzing sports online.
Steve Boyles Digital Humanities Blog 