“I like sports, I could do something in sports. . . . Maybe I could be like an announcer, like a color man. You know how I always make those interesting comments during the game?”
— George Costanza, “The Revenge,” Seinfeld
At the beginning of March, FantasyLabs Co-Founder Peter Jennings (CSURAM88) attended the 2017 MIT Sloan Sports Analytics Conference and served on a panel about daily fantasy strategies (presented by DraftKings).
There were also lots of other panels, almost none of which had anything to do with fantasy sports but all of which are full of invaluable sports knowledge and available to watch online. If you want to be a positive expected value DFS player, you should probably quit your job immediately and watch these videos.
Even if you don’t watch all of them, be sure to watch the one with CSURAM88.
Also, here are the notes I took when watching that video. You know, because I always make those interesting comments.
Daily Fantasy Strategies Presented by DraftKings
For your convenience, here’s the video of the 2017 DFS panel at Sloan:
And here are the Oracle’s annotations. They’re the worst ever. You’re welcome.
The Oracle’s Annotations
0:00 – On a scale of zero to 10, this intro music is #bad.
0:10 – Who is the guy introducing the panelists? Someone needs to introduce the introducer.
0:30 – Strong panel. All the guys got the memo: They’re wearing sports jackets. It looks like CSURAM88 has a black V-neck T-shirt under his buttoned gray jacket. Bold. Also, I think I see a light purple splash on his socks. He looks so powerful. It’s hard to imagine that just a couple weeks before this he injured his face with safety glasses while playing racquetball. I wonder how long Pete leaves his jacket buttoned. I’m setting the over/under at 5:00 and taking the under.
0:55 – Neil Paine isn’t just the senior sportswriter at FiveThirtyEight. He’s also someone who wears ties. He classes up the panel.
1:20 – Renee Miller is one of my favorite people in the industry. She’s smart, humble, and humorous. A neuroscientist at the University of Rochester, Renee is a long-time friend who has appeared on The DFS Roundtable. That someone of her acumen actually thinks seriously about DFS as an intellectual enterprise is a massive credit to the industry. Poor Pete — following Renee is not easy.
2:05 – The jacket is still buttoned. “In 2012 I was fortunate enough to qualify for the FanDuel Football Championship, and . . . fortunately I won that tournament.” Two main thoughts:
- If Pete’s subtly trying to remind everyone that he’s won a fortune playing DFS, his use of the word “fortunate” is A-plus work.
- Peter was fortunate to qualify and win that tournament — but he’s also one of the best DFS players in the world. Don’t confuse “lucky” and “fortunate.”
3:00 – Brandon Adams wears glasses, went to Harvard, and looks smart. D*mmit, Pete. Why didn’t you wear glasses at Sloan? Or have you given up unnecessary eyewear after being attacked by your racquetball accessories?
4:08 – As the camera pans right, we see that Pete has unbuttoned the jacket. #NailedIt
4:20 – I object, your honor: He’s leading the witness.
5:05 – Brandon is totally correct. We’ve talked about this multiple times in articles and podcasts. Cash games tend to be more efficient than guaranteed prize pools, which is why we place such an emphasis on contrarianism. You don’t always need to be contrarian as a DFS player, but you at least need to have an understanding of what it means to be contrarian.
Also, Brandon’s practice of studying the lineups of the best players in the industry is sharp. CSURAM88 and Labs Co-Founder Jonathan Bales periodically post in-depth lineup reviews, which are available to Pro subscribers in our Premium Content Portal.
5:52 – Brandon: “Always your strategy is a function of other people’s strategy.” Yep. DFS is a marketplace.
6:15 – Preach.
6:53 – Neil: “The game theoretical aspects . . .” Oh, Neil. You’re wonderful.
7:35 – Renee: “You’re not necessarily trying to get the highest score every night. You’re trying to beat your opponent. And in order to beat your opponent, you need to have some idea of what your opponent is doing and why.” This might be the most useful statement in the whole video.
8:15 – Actually, that might be pink on Pete’s sock, not purple. Hard to know for sure.
8:40 – Pete: “So the ownership angle is a huge thing, especially in some of these tournaments where it’s very top-heavy.” Pete’s both good-looking and correct. Ownership is the name of the game in GPPs. Rostering low-owned guys with high upside is what fantasy contrarianism looks like. To review ownership across tournaments of all stakes, see our DFS Ownership Dashboard.
8:55 – Brandon and Pete have paid each other compliments within the first 10 minutes of the panel. Good job, guys. #IndustrySolidarity
9:15 – Good thought process by Brandon. When Pro subscribers are setting percentages and building stacks with the Lineup Builder, they should always be thinking in the leveraged way that Brandon describes.
11:35 – Pete hits the nail on the head in talking about short-term variance and how that changes the approach people should have to sport-specific game theory. Also, props to Pistol Pete for the smooth pass to Renee.
12:07 – Polonius: “To thine own self be true.” Renee’s right. The people who are most successful at DFS play to their strengths.
12:53 – Brandon: “In many ways baseball is the best DFS tournament sport.” Bales is nodding his head in agreement . . . wait a minute, no, I was wrong. He was just saying that he wanted another breakfast sandwich. Nevertheless, he still agrees.
13:55 – Pete audibly giggles when Brandon mentions Justin Bour. It’s such a gentle giggle.
14:35 – Neil: “It also sounds like a really complicated data-modeling problem at the core. What do each of you do to kind of manage the complexity of those issues in terms of getting a model that sort of can tell you the risk that you’re taking and the potential for a probable lineup that you put out there?” . . .
14:58 – Pete’s hypothetical answer to Neil’s question: “I use the CSURAM88 Model — ever heard of it?” Paraphrase of Pete’s answer: “Everyone should subscribe to FantasyLabs, use the Labs Tools to backtest our massive database, and fall in love with our industry-leading Plus/Minus metric.” Remember, I’m paraphrasing.
18:55 – Renee: “If you’re aware that people are making biased decisions, they’re people that you’re playing against, and I think that’s where you can kind of apply some of that to game theory. The one that we probably see most often is recency bias, box-score chasing. Nothing drives a player’s ownership up higher than him scoring 40 points the night before.” BINGO!
20:00 – Renee: “It’s easy to be distracted by a lot of data points that don’t really correlate well with fantasy performance.” This dovetails with what Peter said earlier. In order not to be distracted by irrelevant data, we need to backtest data so we know what’s important.
20:20 – I just noticed that Neil’s New Balance sneakers perfectly match his tie.
- That’s the man I want building my data models. His attention to detail is epic.
- He’s basically wearing the sartorial version of the mullet: Business up top, party down below.
- Is that gum on the bottom of Neil’s right foot?
21:20 – I love this back-and-forth exchange between Renee and Pete. OMG, she just mentioned primacy bias. If our marketing guy hadn’t forbidden me from using “Jizz in My Pants” GIFs in articles, there’d by one right here.
Ah, f*ck it . . .
23:00 – Right now, Renee is basically LeBron-ing this panel. I wonder if her favorite song is “Crazy on You,” because I #Heart her.
24:00 – Pete bangs the opportunity drum the way that Gene Frenkle hits the cowbell. “I got a fever . . .”
26:00 – Brandon: “Fantasy players focus on these high over/under games, but they should focus on them even more.”
- This is intriguing.
- It suggests that, in terms of game theory, it’s not a horrible ideal to roster the underdogs in high-scoring games, since they (in comparison to the favorites) will likely have lower ownership.
27:00 – Pete nails it. Target . . .
- High-scoring games.
- Pass-heavy games.
- Players on both sides.
28:08 – Neil: “I wanted to talk about the sports that you guys choose to play in and also the ones you choose not to play in.” [Pete’s brain at this moment: “Dude . . . I grind NASCAR. There is no sport I don’t play.”]
29:20 – Pete: “Yeah, I try to play pretty much every sport.” Pete’s thoughts on engagement are right: Millennial engagement in golf has gone up 20 percent. Fantasy sports are good for real sports. Also, dogs is up 30 percent. I heard that somewhere. I now care about sports that I never would’ve cared about if not for fantasy sports.
30:30 – We’ve reached the halfway point.
32:30 – Brandon: “[Watching games] is a necessary component for DFS.” I respectfully disagree on this point. Some people are great at DFS because they know the underlying sports — but some people are great at DFS even though (or maybe because) they don’t follow or watch the games. I think this varies person to person and sport to sport. I don’t want to harp on this too much because Brandon is making this statement in passing — but I do think people shouldn’t feel as if they can’t succeed at a DFS sport simply because they don’t watch it or aren’t experts in it. To the extent that DFS is a marketplace, someone who’s good at investing and/or trading can have DFS success. You don’t need to know much about how oil is extracted and processed in order to know how to trade oil futures. You just need to know enough about how the oil market works.
34:43 – Renee: “Most of us are very good at emphasizing the positive and de-emphasizing the negative. . . . It’s a trick of our visual systems and our cognitive processing to isolate these positives” — and that’s why strong inference is valuable. Renee and I think similarly when it comes to facts and data.
35:50 – I agree with Renee when it comes to the connection between wealth of data and overconfidence. It doesn’t matter if you have more data. What matters is if . . .
- You have the right data.
- You know what to do with the right data.
38:00 – Pistol Pete fakes a pass to Brandon.
38:53 – Pete: “Baseball is the king correlation sport.” Baseball is governed by the Law of Plenitude whereas football, basketball, and hockey are ruled by the Law of Scarcity. There’s so much I could say about this, but (in general) MLB DFS benefits from correlation plays and stacks because baseball is untimed.
39:00 – I just noticed that Pete’s socks and Renee’s shoes match. These panelists are so in sync with each other.
41:28 – Renee and Pete talk about multiple lineups and using a portfolio approach. As part of your DFS process, it’s usually better to have 10 $1 lineups than one $10 lineup.
43:29 – Brandon: “Possibly the main skill in DFS is basically marshaling your attention in close to an optimal way as you approach that deadline [of lineup lock].” I agree. Knowing how to balance old information with new information in a Bayesian way is crucial to DFS success — especially when that new information emerges shortly before a slate starts.
45:50 – I love how Pete talks about the mental importance of physically standing up . . . while he’s sitting down with his legs crossed.
46:20 – I’m still not sure. Is that gum on the bottom of Neil’s right shoe? Or is it a hole???
49:03 – Pete follows up on Brandon’s point about the importance of a daily process. If you stop tracking a sport for a period, not only do you need to refresh your research on players and teams when you return, but you also must refamiliarize yourself with the game theory and ownership trends that surround the sport. That second type of refamiliarization is harder and more important than the first. It’s also usually more overlooked.
50:20 – Peter: “You can actually gauge what ownership’s going to be based on which content providers are recommending different things.” I might write a Labyrinthian that touches on this in more detail. Basically, it’s the Law of Gravity. DFS is its own solar system.
53:00 – Seven minutes left in the panel, and Pete brings up #NarrativeStreet and Steve Smith. I’m disappointed I didn’t see this coming. Renee literally (and I think inadvertently) raises her eyebrows whenever Pete mentions Smith’s first game against the Carolina Panthers. If you look closely at her face, you can actually see her thinking about whether the Smith narrative was useful or just bullsh*t.
I’ve literally rewatched the Renee eyebrow raise about 17 times. To me, that’s the best moment of the entire panel so far.
54:10 – OK, Brandon’s bit about fading players who party hard might be better than Renee’s eyebrow raise. After Brandon finishes talking, Pete says with a self-satisfied smirk, “Yeah, I like that.” This is perhaps the most non-Sloan two-minute segment of the entire conference. This is amazing.
If it weren’t 5 am right now, I’d be drinking.
Ah, f*ck it . . .
56:55 – Peter talks about stacking. I should say something like, “At Labs we have a Stacking tool,” butt im to drunkk too due that.
58:07 – Renee, Pete, and Brandon give their final thoughts on the new trends and challenges arising in daily fantasy sports. What Renee and Pete said in particular resonates with me. The ability of players in the industry continues to improve. That’s both a challenge and an opportunity. I love it. On top of that, the industry needs to safeguard the experience of the casual fan. Without that, the upside of the industry will be capped.
Unlike this Coors Light.
The Labyrinthian: 2017.37, 132
This is the 132nd 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. Previous installments can be accessed via my author page or the series archive.
