A few columns ago, I touched on the idea that some categories are valued more highly than others. On offense, I theorize that the stat that gets the shaft is runs. In fact, perhaps this theory is an outgrowth of my own bias. I’m having a rough fantasy season this year and may not earn a single league title for the first time in at least four years. This weekend, I spent a bit of time looking into whether there were trends across my teams that could provide insight into strategic flaws in my preseason approach. Besides my penchant for drafting injured players, the other trend that became rather clear was that I didn’t draft enough top-of-the-order hitters and thus have experienced an omnipresent lack of runs scored.
I can peg two reasons why I dug this hole for myself. First and foremost, I am guilty of not paying too much attention to runs, for their own sake, during my drafts. I’m a sabermetrically inclined guy, as I’m sure many of you are as well, so I have a preference for higher OBP-type players, even in non-OBP leagues. I think there’s value to be had in these types of players because their walks bump their ability to score runs and keep the AB total a bit lower, helping to ease the impact of subpar batting averages. So, my way of addressing the need for runs scored has largely been to draft heavy on offense with my top picks and lean toward high OBP players, under the assumption rest will take care of itself. Maybe this autopilot method isn’t sufficient on its own.
The second contributing factor to my dearth of runs scored may have to do with my strategy regarding middle infielders. I’m known to employ something of a stars and scrubs approach to my middle-infield slots. I have no problem spending for a Chase Utley or Troy Tulowitzki, but I often feel that the second- and third-tier middle infielders are poor values. Many of the higher-caliber middle infielders lean more toward being run scorers as opposed to RBI-men, so perhaps that’s been skewing the skillset balance on my rosters.
But the question at hand here is really, how I do rectify this strategic blind spot? How much attention is it advisable to pay toward run scoring as a “skill” when drafting, and in what sort of metrics should I use as a proxy when drafting to ensure I’m getting enough of what I need? Intuitively, it still feels like folly, or at least not the greatest value proposition, to spend all that much effort tracking my runs. It’s hard for me to get past how much of a team stat runs scored really is. I also feel like once you get past the stars, runs scored totals are difficult to predict. And then there’s the question of volume. Runs scored are the most plentiful stat out there, which leads me to consider small differences in expectation all but meaningless at draft time.
Perhaps, filling this void would best be accomplished by proxy. One strategic modification that might help could be to scale back my willingness to take on players with poor batting averages. This might lead me to draft fewer lumbering slugger types and more top-of-the-order hitters. Perhaps, I’ve been all too willing to draft the Carlos Penas, and all too hesitant to draft the Michael Youngs.
Another way to address this might be just to track the projected line-up slot of my players as I’m assembling my team. You can never have too many third- and fourth-place hitters, but a quick overview of my rosters reveals that on most of my teams, I have more fifth- and sixth-place hitters than first and second.
As a columnist here, I don’t have all the answers. But sometimes it’s a good start to just come up with the questions. What are your thoughts about drafting runs scored?; do you pay that category equal attention to every other one or do you just use sort of proxy? I presume the full-on quants will say that one should pay each category equal attention. So, to them, I will ask whether their models have shown to be equally accurate in predicting performance in each category. And, if not, which categories are the most volatile, and do you attempt to address that volatility when drafting or just use the model and hope it’s having a good year in terms of accuracy?