Hoops Lab: DFS Strategy Down the Stretch

Earlier this month, I went to the MIT Sloan Sports Analytics Conference in Boston, Massachusetts. This was my second time going to the conference, and Day 1 of this year's conference was a lot like what I experienced when I last went in 2014: lots of research talks on interesting subjects like algorithms to diagnose NBA play calls, myths about 3-pointers, or how the NBA uses analytics to set team schedules. Also, I got to see a cool interview of NBA commissioner Adam Silver by famous stats guy Nate Silver, which brought about fun tidbits like this one:

Adam Silver: Kevin Durant made more on endorsements in last year in OKC than entire Warriors team combined including @StephenCurry30#SSAC17

— Andre Snellings (@ProfessorDrz) March 3, 2017

That was when I attended a panel entitled Daily Fantasy Strategies, which was presented by DraftKings.

The panel was full of interesting characters with their own histories in DFS. If you're interested in reading more about them, be sure to follow that last link. But in the meantime, let's get to some of my takeaways that I've already started incorporating more into my DFS advice articles.

1. Know what type of DFS player you are

2. DFS tournaments are the type of game with the most opportunity for strategizing

3. In DFS tournaments, high-variance players and situations allow the most opportunity for success

4. Don't take DFS personally, and it's often better to swing hard and miss hard than to play it safe, especially in tournaments

5. Pay attention to the second-level trends, such as stylistic match-ups and teammate correlations

I threw those down as a list, but let's flesh them out a bit. There are different types of DFS games out there, such as 50/50s were roughly half of the entrants win a small amount; leagues with friends where the winner comes out of a small pool; or tournaments with a low entry, a low percentage of winners, but potentially huge upside for those that do place highly as the most common three. Each game takes a particular type of skill set and interest to thrive, so it behooves you to decide which type of game you like best/are best at and then concentrating there. I'm solid at all three, but I prefer to play in tournaments.

As mentioned, tournaments offer the most opportunity for strategizing. That's because they usually have giant entry pools, which means that there is a wide range of expected outcomes. But since most of the possible team outcomes will be represented, in order to win a tournament you have to hit a home run with your picks on a given day. Thus, just like the home-run swing means that you'll miss a lot, likewise playing to win in tournaments generally requires risky strategy. That's where the high-variance player comes in.

A good example can be found on Monday, when Wesley Matthews and Kentavious Caldwell-Pope are both playing. Matthews has been extremely low-variance of late, averaging 22.1 fantasy points in the FanDuel system with a min of 19.6 and a max of 25.5 over his last seven healthy games. Thus, his expected value in the low-20s is pretty well supported for Monday. KCP, on the other hand, has seven games in the teens of fantasy points scored out of his last 10 games. But when he hits, he hits big. He has a 37.6-fantasy-point effort in the last couple of weeks, and multiple instances of 40 to 50 fantasy points scored on the season. The safe play would be Matthews, as you pretty much know what you're going to get from him. But you also know that those low-20s of fantasy points aren't really enough to help you win a tournament. If you slot KCP in, odds are that you may come up short. But if it's a good night for him, he just might push you up the tournament standings by himself. For DFS tournaments, often it's better to swing for the fences than it is to go for the safe single.

In my DFS articles, there is a subset of information that is essentially a given as a first-pass: fantasy scoring average, median, min, max and variance over a recent subset of games (often 10 games). This group of stats gives a quick feel for what level a player is on at a given time, with a sense of whether their production is consistent or high-variance. I complement those player stats with an evaluation of the opposing defense, usually a position points allowed examination over either the season, the last 10- or the last five games. Put those two together, and it's enough to give basic odds on a given match-up.

As a second pass, though, I also like to go further in depth. Especially at this time of year, most players have already played their opponents at least once this season, if not more. And if those games are recent, it can give a good idea how a player matches up with a given team. Also, some sets of teammates tend to either synergize or antagonize each other in a given game. For example, if Ricky Rubio has a great game, it's almost a given that Karl-Anthony Towns had one as well. On the other hand, if Kevin Durant (when healthy) went nuts it was unlikely that Stephen Curry did as well. Point guards often synergize with scorers that act as finishers, but two score-first creators hardly ever blow up together. It's just the nature of the game, and the kind of thing to keep in mind when putting teams together.

All in all, these types of things aren't cutting edge strategy. I was already doing most of them before going to Sloan. But, it's always good to consciously put strategy ideas into words, especially if it helps codify your approach to the game. Two other things pointed out at the conference were: 1) no matter what your model or analysis style, it's really hard to be good at DFS if you don't play every day because we still get so much unconscious information just by being immersed in the game; 2) the prep of articles and models helps a lot, but in the last hour or so before a DFS game tips there's almost always information on player availability and injury that changes the equation. And it's the people who are able to take in that info and make the best snap decisions about what the new information means that generally do the best in DFS games.

So by all means, analyze your numbers, read my articles and even use your models to come up with a general template of good players on a given day. But ultimately, the success of your DFS teams will come down to how often you're playing, and how good you are at putting all the info into a team that makes sense for you. So with that said, best of luck in the last weeks of the season, and remember that DFS continues almost all the way through the playoffs so we still have months of NBA DFS still to go.

Around the NBA

Cousins' ankle: DeMarcus Cousins sat out on Sunday with a sprained ankle. The injury isn't considered overly serious, though with the Pelicans trying to make a late-season push for the last playoff spot out West, it had to be somewhat serious for him to sit. Keep an eye out for information on when he'll return, as his first opportunity comes on Monday against the Jazz.

Gasol's foot: Marc Gasol has a left foot strain, and the team hasn't said much about the severity of the injury. Zach Randolph replaced Gasol in the starting lineup and presumably will continue to do so until his return. The Grizzlies play against the lowly Kings on Monday, so my own personal (completely unverified) opinion is that Gasol will sit on Monday as well. But watch for updates on his status.

LeBron's eye: LeBron James suffered a scratched cornea on Friday, and on Saturday started the game off wearing protective goggles. That didn't last long, though, before LeBron tired of them and took them off. He played the majority of the game without them, and performed at his usual level. This doesn't seem to be an issue moving forward, but keep an eye on the situation just in case. The Cavaliers are in a battle with the Celtics for home court advantage in the East, so LeBron likely plays as much as he physically can down the stretch.

Hayward's quad: Gordon Hayward suffered a bruised thigh on Saturday, but X-Rays were negative. He is listed as questionable to play on Monday, but this doesn't sound like a serious issue moving forward.

Melo's knee: Carmelo Anthony has sat out the last two games with left knee soreness. Another completely unverified opinion of mine, but I don't see any reason for Melo to play another minute for the Knicks this season. The team is headed for the lottery, Melo is in his 30s, he's had knee issues the last few years, and the Knicks really want to see what they have in Kristaps Porzingis anyway. I feel like Melo should and likely will be shut down soon, but the team hasn't said anything like that, so keep an eye out for his availability.

Durant's progress and future minutes restriction: Kevin Durant will be evaluated on Thursday, and we may learn about his potential timeline to return after that. The team still hopes to have him back playing by the end of the regular season, but it's been announced that when he returns he'll be on a minutes limit. His fantasy status for the regular season is obviously nil, but if he returns he'll have DFS value in the postseason so it's worth it to follow his progress.

New Additions and DFS value

Malcolm Brogdon, G (49% owned in Yahoo! leagues): Brogdon has averaged 12.3 points, 4.6 assists, 3.5 rebounds, 1.1 steals, 1.1 treys, shot 48% from the field and 100% (on very limited numbers) from the line in the last eight games. The numbers aren't sexy, but they're useful.

Willy Hernangomez, C (42% owned): Hernangomez has been one of the few bright spots for the Knicks of late, averaging 11.2 points, 10.3 rebounds, 1.1 steals and 0.7 blocks in only 23.7 minutes over his last nine games. He's coming off a 24-and-13 effort against the Spurs on Saturday. If I'm right, and the Knicks do shut down Carmelo Anthony soon, Hernangomez is one of the main likely recipients of added time and production.

Terrence Ross, GF (36% owned): Ross looked great in his first few games in a Magic uniform, then went through a multi-week drought where he didn't contribute much of anything. But he maintained his big minutes even through the cool period, and he seems to have re-found his groove in the last week. He has averaged 17.3 points, 3.7 rebounds, 3.3 treys, 1.7 assists, 1.3 steals and 0.7 blocks in his last three outings.

P.J. Tucker, GF (22% owned): Tucker seems to be reprising his role with the Suns of late, as the utility forward that puts up starting caliber garbageman minutes in solid minutes. In the last three games he has gotten 33 minutes or more in each, helped by injuries to DeMarre Carroll and the suspension of Serge Ibaka, and he has averaged 8.7 points with 9.3 rebounds, 2.3 assists, 2.0 steals and 2.0 treys in those outings.

Ish Smith, G (21% owned): Smith is still battling Reggie Jackson for minutes, but at least for a short period Smith has/had taken over the starting point guard job. He didn't do much in his first couple of games as the starter, and he could lose the job again at any time, but a borderline starting point guard that has shown in the past that he could be productive is worth a roster spot in deeper leagues and/or nightly DFS consideration.

Keeping up with the Professor

The Hoops Lab is now a blog with daily content, so check out/follow for the freshest NBA content every day. Follow me on Twitter @ProfessorDrz. Also, don't forget that you can catch me on the radio on Rotowire Fantasy Sports Today with Chris Liss and Jeff Erickson on XM 87, Sirius 210 on Tuesdays at 1:30 PM EST. Plus, I'm doing DFS articles just about every day, writing analytics pieces for Nylon Calculus on the Fansided Network, and I co-host the TYTSports basketball show on the weekends.