For a while, I’ve been thinking about doing some pieces in which I respond to questions posed by subscribers. And this is one of those pieces. (And this is the final sentence to the first paragraph of one of those pieces: Aren’t you glad that you read it?)
This is the 50th installment of The Labyrinthian, a series dedicated to exploring random fields of knowledge in order to give you unordinary theoretical, philosophical, strategic, and/or often rambling guidance on daily fantasy sports. Consult the introductory piece to the series for further explanation.
“It’s Pronounced ‘Mill-e-wah-que,’ Which is Algonquin for ‘The Good Land’”
Very recently, I noted that the earlier a pitcher leaves a game the likelier the batters on his team are to score more fantasy points (probably because Gray Swan-esque pinch hitters enter the game sooner). Yesterday, I highlighted a trend suggesting that the Milwaukee Brewers should be stacked in tournaments, on account of the likelihood that A) they would be rostered in a low percentage of lineups and B) their pitcher (per the opposing team’s implied Vegas run total) wouldn’t play far into the game. I also noted that Brewers shortstop Jonathan Villar had a good chance of benefiting from that trend and could be a strong contrarian contributor.
By the way, this is what it looks like when a guy points at the scoreboard in the most mundane fashion possible.
The Brewers and leadoff hitter Villar were great yesterday, and their performance wasn’t entirely due to the fact that their pitcher lasted only five innings and batted just twice — but it definitely doesn’t hurt that Kirk Nieuwenhuis, who replaced starting pitcher Zach Davies in the lineup, got on base.
I might be indulging in confirmation bias, but the Brewers’ performance seemingly affirms the ultimate point I made yesterday in the Trend of the Day:
If you want to leverage exploitable pitching performances in a contrarian way, don’t stack batters who will be facing those pitchers. Instead, stack the batters on their own teams.
It pays to pay attention to the impacts that pitchers have on their teammates.
But What About the Bullpen?
Yesterday, a FantasyLabs subscriber asked me on Twitter if there was a way to use our tools to identify bad bullpens and to quantify the extent to which they are bad.
That’s a great question.
Here’s the short answer: No. Right now there’s no direct way to use our tools to identify (and quantify) bad bullpens.
Great article, everybody. Thanks for reading. Lots of fun. See you next time.
But What About the Bullpen? — The Longer Answer
There are a number of reasons why it doesn’t make a ton of sense for us to have data on bullpens in the Trends tool and Player Models. It’s not that the data would be entirely useless, but it might be misleading. (Also, if it sounds as if I’m channeling PGA Director Colin Davy, that’s because I am.)
For one, what constitutes “bullpen”? It’s not uncommon for a team to have a pitcher (especially in the middle of the season) who moves back and forth between the bullpen and the starting rotation as needed — which gets to a larger point: The makeup of bullpens is fairly fluid. Pitchers go on the Disabled List and new pitchers replace them, some pitchers get traded or sent down to the minor leagues and minor leaguers get called up, etc. The composition of bullpens isn’t especially stable.
Furthermore, even if the composition of the bullpen were stable, we would still have no real idea of knowing which pitchers in a bullpen are likely to pitch in any given slate. We have an idea that the closer could pitch. And maybe the set-up man. But it’s not certain that they will pitch, and it’s a fool’s task to attempt to forecast usage for relief pitchers other than closers and set-up men. Essentially, there’s no firm way to predict how a manager will deploy his bullpen.
Finally, there are questions about the sample. When we look at a starting pitcher’s advanced data over the last 15 days, we have a small sample, but at least it’s fairly consistent. When the pitcher started his contests, both teams had zero runs. For relief pitchers, the sample is much smaller, and it’s much more uncertain. There isn’t the same consistency in circumstance. As problematic as the issues of sample size and representativeness might be for starting pitchers, they are amplified for pitchers who come out of the bullpen.
And, of course, there’s one more matter.
Viva, Las Vegas
At the beginning of the season, Bryan Mears and Jonathan Bales did an introductory MLB podcast that was fantastic. In it, Bales highlighted the importance of Vegas totals to MLB DFS. In fact, Bales values Vegas so highly that in his Bales 2016 Model the factor with the highest weighting is Vegas Score.
Despite Vegas’ relative inaccuracy at the beginning of the season, Vegas is hugely important to MLB DFS. Knowing how to leverage what Vegas says about sporting events basically is DFS.
Anyway, here’s where the topics of bullpens and Vegas intertwine: I doubt that all of Vegas’ implied run totals are based solely on starting pitchers. Vegas probably takes bullpens into account.
So if you want some indirect sense of the quality of a bullpen, maybe just look at what Vegas says. It won’t be precise, but it will probably be good enough.
And if you wanted to be complicated then you could probably correlate pitch counts, WHIP, etc., to get a sense of how many innings a starting pitcher might play, and then you could apportion some of the Vegas total to him and the rest of it to the bullpen . . . but what’s the point?
It’s Not That Bullpens Don’t Matter
The bullpen almost certainly matters. It’s just that our ability to isolate the signal and separate it from the noise is limited . . . as is the degree to which the bullpen actually matters. In terms of what we can quantify and what we should care about, the bullpen is fairly low on the list.
Look at it this way: In the eyes of the average manager, the perfect game is one in which he doesn’t need to rely on his bullpen. He is explicitly hoping not to use his bullpen. That doesn’t mean that he won’t. It does, however, mean that the task of projecting bullpen usage is even harder than it otherwise might be. Why should we spend a lot of time trying to quantify the small contributions of players whose own managers want them not to contribute?
In life and DFS, I try to focus most on the factors that matter the most. There’s a reason that guys are in the bullpen to begin with: Out of all MLB pitchers . . . they matter the least.
By the way, thanks for making it through 50 of these pieces. You, the reader, matter the most.
———
The Labyrinthian: 2016, 50
Thanks to @ahosier2 for asking the questions that prompted this piece. Previous installments of The Labyrinthian can be accessed via my author page. If you have suggestions on material I should know about or even write about in a future Labyrinthian, please contact me via email, [email protected], or Twitter @MattFtheOracle.
