As you might have noticed, I have not written much analytics content over the past month to five weeks. The reason is that I’ve been preparing for a talk that will be presented at the 2015 OptaPro Analytics Forum in London on Thursday. It’s been a lot of work, but presentation day is almost here.
Unfortunately, I am not able to appear in London due to budget and scheduling constraints, but I am very grateful that my friend Simon Harriyott has agreed to present the paper in my absence. Preparing a talk that someone else will give is always a challenge because what you might write could make perfect sense coming out of your head because you’re aware of the underlying context yet appear incomprehensible to someone else. Simon has done a really good job with helping me prepare this talk and I look forward to a very good presentation from him.
The topic of the talk is “Framework for a player career forecast model between multiple leagues”, which is a statistical forecasting model that projects performance output of a football player as he transitions between multiple leagues in a career. The objective is to create a soccer equivalent of projection systems such as PECOTA in baseball and SCHOENE in basketball while incorporating machine learning techniques as much as possible. Work on this model began at the beginning of the year, so do not expect a lot of results to be presented. The goal of this talk is to present at a high level the objectives and methodology of the model, obtain feedback from the soccer analytics community, and gauge interest from the broader football industry.
The presentation is scheduled for 2:15pm UK time on Thursday 5 February, which will be 9:15am Atlanta time. I know that the talk will be recorded, but I do not believe that it will be streamed live. I will be online at the Soccermetrics Twitter feed to answer any questions that attendees might have, and you can follow the conversation for the rest of the forum using the hashtag #OptaProForum. The slide deck will be made available at the Soccermetrics page on SlideShare within 30 minutes following the presentation.