A New Way of Looking at Depth

One of the great perks of working at FanGraphs is that I get to discuss baseball with my equally obsessed coworkers. Obviously, this is the kind of job you don’t get into unless you love the sport. A lot of the time, that means we just end up nerding out over how much we enjoy some minor but cool thing, or perhaps discussing our favorite of the game’s idiosyncrasies. Sometimes, though, we come up with new ideas together, or one person’s passing fancy turns into another person’s brainstorm, and before you know it, something nifty and novel is happening.
That’s why I’m writing this article today. At the December Winter Meetings, a subset of us sat down for our annual let’s-talk-about-fun-baseball-problems technical meeting. David Appelman and Sean Dolinar ran things. Folks like Jeff Zimmermann and Dan Szymborski popped in at various points. Jason Martinez and Keaton Arneson had big plans for how to improve the site’s functionality. Those guys are great at building models, running websites — advancing the state of how FanGraphs (and ZiPS) works, basically. I like to make jokes and write articles about bunts, so as far as I can tell, I got invited because I’m good at coming up with bad but interesting ideas.
That said, this year one thing was on a lot of our minds: depth. I’ve written a lot about how well our playoff odds reflect reality. They’re pretty good! But there’s always been an obvious problem with them. They use static rosters, which means they don’t account for the fact that some teams are more vulnerable to injury or underperformance than others. Read the rest of this entry »