2018 Research Projects: Known and Unknown
Fantasy baseball slowly evolves and with the start of 2018, I’m going to focus on some topics in which savvy and unconventional owners could utilize against their competition. Some of these topics I’ve been contemplating and researching for a while, others are still just dust in the air. The following contains some ideas I hope to have a better understanding of by the year’s end and frame it so owners and utilize it. I’m not the only source of ideas. I’m 100% positive I’ve missed some simple useful topics and would love to research those in presented in the comments.
StatCast Batted Ball Data
As a baseball community, we’ve been slow to adapt and utilize StatCast batted ball information. Besides major league baseball employees, no one is leading the change. Part of the problem is having an easy to access database for everyone to use. That barrier is being removed as tools like Bill Petti’s baseball-r package can be used to help scrape the needed information from Baseball Savant.
With the data now becoming available to some of the public, some headway is being made like the work by Jim Albert and Anthony Shattell. Personally, I’ve made some major advances but my lack of programming knowledge has limited me to processing just one player at a time. I need to step up and be comfortable programming in R.
Additionally, my understanding changes. Just yesterday, I had to recalibrate my understanding of batted balls. The exit spin rate needs to be part of the equation and the data is not publicly available. The spin generated can make a difference of 30 feet of distance traveled which helps describe some the discrepancies researchers are seeing.
Bias
I expect readers will eventually tire of reading about biases but I find them to be a frontier for owners to gain a major advantage. Currently, a ton of material has been written on behavioral economics and I have been consuming everything I can find.
Most of the written material is directly related to stocks. At the root core of fantasy baseball, owners are trying to get as much return as possible with the lowest cost, just like traders are with the stock market. For every known bias in the real world financial sector, there is a direct parallel to fantasy baseball.
For example, I’ve started to see and determine how much recency bias is baked into ADP and player rankings. Some initial values point to the most recent season getting double the weight as it should have historically. For example, if the historical weights of 4, 3, and 2 create the best projection, owners seem to be valuing players at 8, 3, 2. Now the most recent season’s weight may not be the only factor pushing up some players. I’d likely to additionally see how much positional and shiny-new-toy bias inflates a player’s value.
Biases are everywhere and I will try my hardest to point them out and how they can be exploited.
Other Studies
The two above topics are ones I want to continue to dive into at a regular basis. I expect to create some hard fast rules from the work to help owners gain an edge.
Besides the two main, overarching topics, here are some more detailed studies I plan on completing.
- When do prospects get called up (likely to be published at BaseballHQ)?
- How to approach FAAB and correctly value free agents? I feel the FAAB process is one of the least understood processes and hope to find a system to take advantage of its inefficiencies.
- I want to find the ideal time and players for trades. Ownership bias is tough to overcome but it can be weakened and deals found.
- Quality pitch finder. A basic quick test to get an idea of a new pitcher’s repertoire after one major league appearance using Pitchf/x data.
- Perfect game theory and how to apply it to fantasy baseball.
- I’m trying to get some outside voices (e.g. business executives) to help point obvious flaws with our game.
Outside ideas
I know topic ideas will present themselves as the season progress. Even so, I’m sure there are some great ideas I can check and verify provided from our readers. What fantasy questions do owners want answering? What industry norm just doesn’t quite smell right? How can I make you a better fantasy owner? Let me have them.
While I believe some fantasy formats are stale, overall, there are many advantages which haven’t been exploited. My goal for the next 12 months is to find as many advantages as possible. I’m not sure these findings will radically change player valuations or the overall game. People are still stuck in their old ways. It’s time to use that stubbornness to an advantage.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
I need to see batted ball spin rates and I need to see them now. No idea why they aren’t public. All season I was saying Castellanos underperforms compared to his statcast data because his extreme uppercut appears to generate a lot of top spin fly balls that end up being routine fly outs – strict eyeball opinion only. Some could even be registering as barrels.
What I have long wanted to know is whether there is such a thing as throwing too many strikes? Getting batters out is not just about throwing strikes, but it’s important to know when to throw outside of the strike zone.
I have an experimental theory that some pitchers who have a constantly worse era than fip are beeing punished by throwing too many strikes. What do you think? Is this even a possibility?
Will try. I can start by using Zone% as a proxy for strikes and go from there. Thanks.
Jeff, what valuation system is best for H2H leagues? I’ve seen that SGP shouldn’t be utilized. Thanks!
You can use the auction calculator to get some general ideas of rankings but I generally just use roto values with an emphasis on OBP and PA. I need guys getting on base as much as possible and hope the other stats come.
I usually don’t like to intentionally drop categories in roto (may as the season goes on and my team takes shape) but have in H2H. I dropped SB (did when some with a Villar pickup) and Wins (used K/9) in the Tout Wars H2H league and squeaked out 1st place (thank you Yelich).
Jeff, I have many years of experience with R. If you ever want to post some sample code on github/pastebin/etc, I’d be happy to offer some suggestions on streamlining it. I’m sure many readers would offer the same.
I’m at the point of not even knowing which code to put out. I’ve tried to start learning it several times but writing deadlines and real life keep getting in the way. I need to find a point to move to it 100% and level my SQL interface behind.
Hey Jeff, here are some ideas I have
*I feel like we still do a poor job of predicting stole bases. I think we could better utilize the new sprint speed data and speed scores to predict SBs. Taking it one step further would be to try and quantify each managers propensity for SB attempts.
*Deeper dive on launch angles and exit velocities. Lots of people now quote a players average LA and EV, but this can be deceiving. A player could have a great average LA, but if too much of their contact comes from each extreme of the spectrum then that is actually a poor outcome. I’d love to see the frequencies of contact at the optimum launch angles and the corresponding EV on that contact.
*On the pitching side, I’ve always really enjoyed the arsenal scores that Fangraphs has published. Not sure exactly how exactly to arrive at it, but feel like there is lots of fantasy value in these scores since they do a great job of quantifying things that FIP doesn’t measure (pure stuff).
* Agree on the steals. I always wonder if losing teams are more or less likely to steal. Why no run.
** As I stated in several paragraphs, LA and EV improvements are coming.
*** For the arsenal scores, they morphed for me into my pERA metric. All the computations for the data can be found in this article or all the linked articles: https://www.fangraphs.com/fantasy/2017-pera-update-with-exit-velocity-grades/
I had missed your pERA article but that’s really good stuff. Thanks Jeff!
Not exactly fantasy related. But I’ve always wondered what the correlation between foul balls and a pitchers success is. Is there a skill in hitting it foul or is it just happenstance and does a foul ball in a given at bat give a hitter an advantage or the pitcher?
This is pretty park-dependent, one of the reasons that Fenway is generally high-offense beyond balls hit to the Monster is that there’s almost no foul territory to speak of.
Oakland’s the opposite, O.co’s deep foul territory leaves a lot of room for fielders to run down foul popups.
I think being able to foul off a pitch that wasn’t what you guessed/cheated on is a skill, but I’m not exactly sure what the best way to measure that is. Two-strike fouls?