This week, we published a lot of content that will be applicable for a while. Within this FantasyLabs Friday Recap, you can find links to all of that content, for your weekend reading pleasure.
General
– Cowboy Correlations and Finding Truly Predictive Data, by Bryan Mears
A lot of reasons exist for why we might see correlation that really means nothing: Sample size, sample representativeness, randomness, etc. Our tools and data are very powerful and can help DFS players be profitable — but only if you use them correctly. Don’t blindly accept everything you see in your DFS research. Always dig deeper and find out why the data or trend says what it does. Only then will you be able to withstand the allure of Cowboy Correlations (trademark still pending).
PGA
– Video: The Players Championship Model Preview
Watch as Peter Jennings (CSURAM88) breaks down his personal model for this week’s The Players Championship.
– Video: PGA Lineup Review for the Wells Fargo Championship
Peter Jennings (CSURAM88) breaks down his PGA lineup for the Wells Fargo Championship.
– Video: Using FantasyLabs PGA Tools
Adam Levitan walks through how he uses the FantasyLabs PGA tools.
– How Predictive Is Specialist Data?, by Colin Davy
After all of that, my hunch is still that specialist metrics aren’t all that useful based on my experience trying to fit other individual tendencies to data. It’s hard enough to get good general rules on how to weight things. When you try to apply them at an individual level, there’s just an overwhelming amount of noise. Somewhat ironically, the only individualized metric I’ve found that works well is course history. But as far as the specialist approach, I was skeptical of its usefulness before doing a data dive, and I’m even more skeptical after.
MLB
– The Daily Fantasy Flex, MLB: 5/13/16 Main Slate
Jay Persson and Peter Jennings (CSURAM88) break down the 5/13/16 MLB DFS main slate.
– MLB Plays of the Day: 5/13/16, Main Slate
Toronto has the third-highest total on the board and it’s common knowledge that Josh Donaldson destroys southpaws. He’s so exquisitely chalky that I would expect to see his essence announcing the evening’s specials on the blackboard outside a small Parisian café.
– MLB DFS 5/13/16 Slate Breakdown, by John Daigle
As a knuckle-baller, Steven Wright has a (low) pitch speed that typically stays between 82.9 and 83.3 MPH. He may be projected to allow “only” four runs tonight, but note Houston’s +3.60 Plus/Minus when facing pitchers who have average velocities lower than 84 MPH. Additionally, Houston’s top-four batters will likely be rostered in a low percentage of lineups, given the Coors chalkiness of the slate.
– BvP — The Bigfoot of MLB DFS, by Mitchell Block
If the batter is routinely finding success against pitchers of a comparable makeup — and you could look at additional factors such as pitch selection, arm angle, etc. — there is likely something to the high BvP performance. If, though, a batter doesn’t do well against comparable pitchers, then the high BvP performance (such as Posey vs. Kennedy) might be an outlier, and perhaps over time we would be likely to see player regression.
– State of the Stacks, Vol. 3, by Mitchell Block
Completely independent of the pitching matchup, we can see that players who A) have a Distance Differential Score in the 85th percentile, B) are in the top half of a lineup, and C) play in a stadium with a minimum Park Factor of 32 collectively register an average Plus/Minus of +0.56. Again, this is without factoring in handedness or pitcher quality.
– Regressing to the Mean: The KBXE Review, Vol. 2, by Mitchell Block
Strikeouts per nine innings is a great stat, especially for tournaments, as it conveys a pitcher’s strikeout potential. Elite pitchers often have high K/9s. But if a starter is routinely pulled around the fifth or sixth innings, he’s never realizing his strikeout potential. Such had been the case with Jose Fernandez up until recently, routinely capping his ceiling. Keep this in mind, as it’s easy to overlook a pitcher who may have a slightly lower K/9 projection but a better chance of going seven or eight innings.
– On The Contrary: A Small Dog With Bite, by Brandon Hopper
The Brewers are a top-10 team at hitting the fastball, but that’s only Shields’ fourth-best pitch and not one that he relies on often. He has more success with his cut fastball, curveball, and changeup, which is why I’m not too worried that his velocity has dropped by 0.8 miles per hour over the last 15 days, per our advanced stats. Using our Trends tool, we can see that an MPH decrease between 0.5 and 1.0 hasn’t (historically) resulted in catastrophe when it’s coupled with a strikeout prediction between eight and 10. For a finesse pitcher with a good chance of getting strikeouts, losing a bit off a fastball isn’t as bad as it would be for a hard thrower.
– MLB Trend Testing: Ks and Umps, by Bill Monighetti
This week, I wanted to create a really strong cash game trend for pitchers. When I am selecting pitchers for my cash game teams, the two main things I’m looking at are strikeout potential and win probability. In this trend, I’m going to be using a relatively new feature we have at FantasyLabs, our K Predictor (which you can see in our Player Models). In addition to looking for pitchers whose K Prediction is high, I want an umpire who has historically benefitted pitchers.
– MLB Recent Form Report: 5/9/16, by Bill Monighetti
Felix Hernandez is the type of pitcher who can get by, and even excel, without an overpowering fastball. In 2015, his fastball was just his fourth-most frequently thrown pitch. What is troubling is that the velocity is down on each of Hernandez’s main pitches this season. Also, in 2015, Hernandez had three pitches that generated around a 15-percent Swinging Strike Rate. In 2016, only his changeup is over 10 percent (13.6). His BABIP on each pitch is BELOW career norms, meaning that, although King Felix’s surface numbers have not been great, they should probably be worse.
– More to the Robustness of Strikeout Pitchers, by Matthew Freedman
If it’s true that, in general, a high-K pitcher can be just as or almost as effective with reduced velocity as he is with his normal velocity, then I imagine a time will come — and we’re probably not there yet — when many DFSers will have access to advanced data and will fade high-K pitchers merely because they aren’t throwing as hard. And when that happens game theory will dictate that we should explicitly want to roster these pitchers in tournaments. And, to me, that will be exciting.
– What About The Bullpen?, by Matthew Freedman
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. 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.
– More to the Proper Valuation of Designated Hitters — P.S. Pinch Hitters Are Gray Swans, by Matthew Freedman
The NL pinch hitter is a known unknown. We know that he will likely have an impact on any given game, but the extent of that impact is unknown. Or, to borrow from Taleb, perhaps the NL pinch hitter is a Gray Swan. He belongs to the domain of the somewhat random. It could very well be that the pinch hitter gives NL batters a DFS edge, not that the DH takes away the DFS edge that we might expect AL batters to have. It’s possible that DFS platforms are simply not very sharp at evaluating the known unknowability of the Gray Swan that is the NL pinch hitter.
Trends
– MLB 5/9/16: Elite Pitchers With Different K Predictions, by Jonathan Cabezas
– MLB 5/10/16: The Offensive Impact of the Pinch Hitter, by Matthew Freedman
– MLB 5/11/16: Why You Should Pay Attention to Pitch Counts, by Jay Persson
– MLB 5/12/16: Whether You Should Worry About the Weather, by Bryan Mears
– MLB 5/13/16: High Strikeout Pitchers with a Recent Drop in Velocity, by Mitchell Block
