The Sweet Spot: MLB Fantasy Picks Today for DFS (5/13)

Every Monday, I try to turn the daily MLB slate writeup into more of an evergreen, DFS strategy/theory piece to which you can return at any time and still find value. The GPP Scores that have served us so well this MLB season were inspired by the GPP Scores fueling an incredible PGA season, so considering the fact that this week is the PGA Championship, it only makes sense to dive deeper into the crossover between these sports as it pertains to DFS.

More specifically, I want to find intuitive examples in one sport that make counter-intuitive ideas in the other easier to understand. We’ll talk projection models, how missed cuts in PGA can embolden us to fade chalky pitchers, and of course — GPP Scores.

Important Note: For those of you who don’t play PGA DFS and know nothing about the sport, fear not. Pitcher and stack recommendations can be found at the end of the article for today’s slate.

Projection Models

The first similarity between these sports is that they are extremely data-rich. It’s not just how much data exists, but that both can be modeled with very little manual input (unlike NFL and NBA, for example). The second similarity is that we at FTN have been able to build projection models for both that use unique inputs (like the course fit model and the data powering our new FILTH metric, which will be publicly available soon), providing a substantial edge over our competition. 

Defining Failure for Chalk

A few nights ago, I came close to a takedown:

I was let down by a 61% rostered Cole Ragans, which begs the question — should I have played him? It’s a tricky question to answer, especially because:

• We didn’t expect him to be anywhere near this popular
• I had plenty of salary, so paying down for pitching wouldn’t have made a ton of sense
• Ragans had a -4 GPP Score, but the two other expensive pitchers, Freddy Peralta and Nick Lodolo, had even worse GPP Scores

123.20 won the contest, so I was well within striking distance, but there were a handful of lineups in front of me that also had Ragans, so I do think I missed out on an opportunity to leverage the field just enough for the takedown. Which brings me to the overarching point — it can be really difficult to visualize a chalky pitcher failing because it’s less defined than it is in PGA DFS. 

A pitcher failing can mean an early injury. It can mean a number of sloppy innings without strikeouts, four strong innings and then a complete ejection in the fifth, or like Ragans, a little of both. Consequently, it’s extremely difficult to have an intuitive understanding of how likely a pitcher is to fail. On the contrary, if I were to ask you what failing looks like in PGA, it would require only three words: “A missed cut.”

We fade players for two reasons in DFS:

1. For the opportunity to play someone else
2. For the opportunity to leverage the field if they fail

Everyone talks about the first, I rarely see any discussion outside of FTN about the second. Should I have faded Ragans? If I had known he would be anywhere near as popular as he became, the answer would have been yes. Not because I had a great pivot in mind, but because doing so would have provided an opportunity to capitalize if he failed… and like we’ve learned through both PGA and MLB DFS, failure is always more likely than we’d like it to be.

The GPP Scores

Once again, this is where the GPP Scores come into play. As it feels like I’ve written hundreds of times now (sorry to those of you who have read this as many times as I’ve written it, it’s still new information to some), the GPP Scores do a tremendous job of quantifying what our intuition fails to understand — this idea that it’s not just about who we play, but also who we purposefully do not play. 

Let’s take a look at Monday’s GPP Scores to see if we can take this reminder into Monday’s DFS slate.

Pitchers

As usual, I have sorted the pitchers by projected value:

Yoshinobu Yamamoto is the perfect example. He is the top projected pitcher on the slate in our model yet has a GPP Score that screams to fade him. Why?

Well, he’s only ninth in projected value, and though he’s second in projected strikeouts, there’s little separation between his projected K’s and many of the pitchers in the $6k and $7k range. So, think about what happens if Yamamoto fails…

Something like 25% to 40% (depending on your contest) of your competition will not only have their pitcher fail, but it will be a pitcher whose price excludes a number of expensive offenses. That’s too good to pass up.

I will likely be using Spencer Arrighetti (we believe he’s a much better pitcher than his results so far have shown) and either Jose Soriano or, if I have a bunch of money leftover, and I’m not stacking Atlanta, Shōta Imanaga.

Hitters/Stacks

So far, only one offense really stands out to me — Atlanta (sorry, Shōta).

Atlanta Braves

I love this because it fits perfectly with the Yamamoto fade. Atlanta is one of the few offenses expensive enough to be a near-impossible fit with the Dodger righty.

Marcell Ozuna and Matt Olson both have the top GPP Score at their position and the outfield is full of great options. Do I wish the best GPP Scores on the team weren’t almost universally from outfielders? Of course, but we can find infielders elsewhere.

Boston Red Sox

The Red Sox were the first team that I noticed with positive GPP Scores at the positions we would still need to fill alongside a Braves stack:

I also like that you can turn this into a really expensive (Rafael Devers and Tyler O’Neill) or really cheap (Dominic Smith, David Hamilton, Vaughn Grissom, etc.) secondary stack. Hell, you could also turn into a balanced primary stack and have plenty of money leftover for Shōta.

Philadelphia Phillies

It’s very rare that I go against the model, but if ever there were a spot to do so, it would be this one. Yes, the Phillies are my favorite team, but that’s not what this would be about — I simply don’t think Sean Manaea is any good and no haircut will ever change that. I think I’ve made more money attacking Manaea than any other pitcher over the last few years, and it’s all about his propensity to just get absolutely shelled at times. It’s only happened once so far this year, and he’s done a tremendous job of limiting power (hence the poor GPP Scores for Philly other than Bryce Harper), but perhaps it happens again tonight. Importantly, I would never go this heavily against the model on a chalky team — the Phillies are super contrarian, especially if you pick the right pieces.

Final Thoughts

There are a couple of additional thoughts to take away from the comparison between MLB and PGA. For one thing, think about how likely a stack as a whole is to fail! Even Alek Manoah went seven innings Monday night with zero earned runs! Shoutout to the model for that one — he was first in GPP Score for Monday’s slate.

This is why the chalkiest stacks almost never pop in the GPP Scores — the industry is simply too confident in the success of the “best stack” on most slates.

Lastly, consider the differences in roster requirements for each DFS sport. MLB lineups require 10 players, PGA only six. Your first thought may be that since it would be nearly impossible to hit on all ten players if they’re all contrarian, we want more chalk in MLB than PGA. However, you can get takedowns without hitting on all 10 MLB plays, whereas you almost always need to hit on all six for a PGA takedown. Consequently, I don’t think it’s about playing X amount of contrarian guys and Y amount of chalky guys. Instead, I just think it’s about following the GPP Scores, and using the desire for a balance between contrarian upside and high probability plays as a tiebreaker when the GPP Scores don’t provide a crystal-clear route.