Sig Mejdal, Senior Quantitative Analyst for the Cards, was kind enough to answer a few questions that I sent him via email that deal with his daily responsibilities / activities in his role with the Cardinals. For a little more background on Sig here’s a short bio as posted for his appearance at the JMP Innovators’ Summit.
Sig Mejdal has been working as a Senior Quantitative Analyst for the St. Louis Cardinals since opening day of 2005. He provides analysis, player projections and data-driven decision making for the General Manager’s office. Mejdal’s work is used for the amateur draft, and both the minor and major leagues. While his baseball playing career ended in Little League, he has had an almost unhealthy interest in baseball research ever since. This led him on a quest to become one of the few “quants” within baseball’s front offices. Mejdal has degrees in Mechanical and Aeronautical Engineering, and Masters in Industrial Engineering and Human Factors Engineering, and has worked both as a Travel Writer and a NASA sleep researcher.
Basically he has the job that most of us amateur saberists would love to have a chance to do.
Anyhow, on to the exchange, my questions and comments in bold.
How is the analysis time / effort broken out between the Major League side of the house (Free Agent valuations/projections, defensive metrics etc.) versus the player development side of the house (prospect (both draft and internal) valuation, etc.)?
The entire quantitative team here does their best to have all that you ask about covered. Of course, as the season progresses, we throttle up our attention or analysis on different areas. But what you describe are all areas that we analyze.
How much do you leverage the freely available information (I’m thinking specifically of stats sites like fangraphs, and analytical studies like MGLs aging study, the work that Jeremy Greenhouse is doing, Victor Wangs research on prospect valuation, etc.)?
We are on those sites daily. Moreover, I’d say in general we are pretty aggressive about trying to build on research and ideas generated by people outside our club and even outside our industry, so our reading extends beyond the sabermetric community. We don’t leverage their results specifically, but instead are taking a look at their processes, their ideas, their conclusions in order to occasionally spark work of our own. Many of those sites that you mention do very good work, however, you can imagine that the decision makers are going to have to know each and every step of the process that led to the statistical evaluation.
Somewhat related, how much time are you able to spend keeping up with the ”amateur” sabermetric community?
Not as much as I like – others in the group spend more time. I have my google reader directed at many of those sites and I try to take a look at that daily. To some degree, those in the front office rely on each other to forward that which we think others will find useful or at least interesting. We would be foolish not to. There is a tremendous amount of brainpower and energy out there thinking of all sorts of things. To not expose ourselves to that would be missing a very good opportunity to brainstorm and/or learn. To me the answers to those two questions are fairly important and clearly Sig gave favorable answers. There’s a whole community of subject matter experts out there doing quality work. It’d be foolish to ignore it.
As more and more sabermetric related news/viewpoints are entering the mainstream have you experienced an increase in people looking for work/internships?
We have experienced an increase in those looking for work. Not sure if it is because more of the research has entered the mainstream as you mention, but yes, we get more letters than ever from very qualified
persons looking to get their foot in the door.
How much do you interact with the “traditional scouting” people and try and use their information to inform/improve the quantitative things you do?
As much as possible. John Mozeliak, our GM, has always been a strong advocate for breaking down the artificial walls between traditional scouting and analytical analysis. So blending those two approaches is something we take very seriously. In the analytic group, we talk to our scouts frequently, often on a daily basis. The conversations with the scouts or the opportunities to go to games with them has led to many hypothesis about how best to combine the on-field performance and the expert ratings. And yes, you are exactly right, we do use their information to improve upon our models. This is unquestionably the right answer. It only makes sense to put ALL of the data you have to use, even if some of it is subjective in nature.
Many thanks to Sig for taking the time to answer a few question. The Cards appear to be in good hands with him running the analytical shop.