Get better at DFS. This has been something that I constantly cite as a goal.
Over the All-Star Break, I’ll be posting some strategy articles for those interested in stretching their game and learning more.
Many of us DFS players, are in love with the game but regular guys. Don’t have to make thousands of dollars to love it, but wanted to continue being successful. This week I want to nerd out over process and improvement. In my mind, the road is just as fun as the results.
I’ve settled on an incredibly corny title for the series. “No Days Off”… Here she goes.
No Days Off 1: Player Selection Checklist
I recently read a book called the “Checklist Manifesto”. The core concept is that, even though they are seemingly simple, checklists are actually at the core of some of the most complicated processes in the world. Pilots, surgeons, structural engineers all use checklists in order to eliminate human error. As a result, we can fly across the ocean, have heart surgery and sit on the 95th floor of a high rise without dying.
Principally, they work to eliminate your own faults and errors. We could all use this in DFS. There’s nothing worse than talking yourself into a bad play, then watching it bust because you didn’t remain objective about downside.
So, what should we focus on in our player selection? I’m going to try and break down some simple principles that will help you vet your player selection beyond the scope of the robots (optimizers, projections etc.)
When players meet most of these reqs, they should be solid cash game selections. The list will be broken down into do’s an don’ts to check for upside and downside.
DO: Driven by measurable statistics, here’s what we do want in our plays…
1- Pace: Is my player in a pace up game, or going to see a positive pace differential?
Pace is actually one of the most important things in DFS. Extra possessions equal extra production. It is one of the primary things that Vegas uses to determine a game total. Sharps and algorithms will use it to drive projections.
As such, you should be targeting players who are going to see a high pace differential in their game.
For example, when the Wolves played the Nuggets last week, we saw Wiggins and Rubio at the top of the board in a lot of models. This is because these are high usage players on a slow team, who were going to see extra possessions.
Often, I will concentrate my player pool to teams who will play in pace-up games, or see a high differential in pace
You can determine this with the RotoGrinders ‘Defense v Offense ADV’ tool or simply by looking for the highest team totals on the night.
2- Price: Is your player’s giving you an advantage?
Price watching matters. We don’t overpay for stocks, cars or anything else we invest in, so why would we overpay for a PG? You may be attracted to a player because he’s ‘hot’ and ‘had 30 last game’ but you will be paying for that.
Price watching allows you to understand when a player is at his ceiling price or dropping because of a recent dud. DFS pricing is usually driven by a combination of recent performance and matchup. However, you want to maximize the points you get for the money spent, overpaying for an asset could be the difference between a player getting you 4x and 5x.
Knowing what a player usually scores or is capable of scoring should drive your decision making.
Most nights, a player will score near his average but if he is coming off some above average performances you will likely overpay for him. Tracking this has to do with regression, both positive and negative. What we call, looking for a ‘bounce-back’ or ‘struggling’.
Looking at a recent example, Gordon Hayward went on a tear before the All-Star break. Scored 40 FDpts on multiple nights, eventually his price went up to 8.1k. This was a ceiling price for him, given his body of work this year. The night he was priced that high, he regressed only scoring 13 pts. His price dropped back to 7.7k, which jives with his overall performance this year and is a fair price for his role recently. However, the ideal time to roster Hayward is when he is priced in the low 7k range. We know he can easily score 35 pts a night, but he is also capable of going for 40. That would maximize your pts/dollar.
You can find salary changes on the RotoGrinders site. Sort by declines in price and look for some names on that list who are due for positive regression, and do the reverse. See who’s price is rising and be cautious about rostering a player priced for his ceiling games recently.
3- Position: Is he in a position to succeed given minutes, usage and matchup?
I put this in third place for a reason. A player needs to be in a position to succeed given his matchup, role and minutes. Minutes and role are obvious. We want starters and closers. We want minutes, and touches/usage that allow a player opportunity to score. DVP is less obvious.
Defense vs Position is the metric that measures how successful teams are at defending a certain position.
DVP is a common starting place for a lot of players in player selection, but it’s important to combine that knowledge with current trends, past performance and Vegas.
For example, we were accustomed to attacking the Bucks at the SF position this season. But that was because of Jabari Paker’s terrible defense. Giannis is now guarding the 3 in many situations and that gaping hole will be corrected.
That said, DVP is an excellent indicator of a team’s weakness. Coaches and teams know this too and will often game-plan to exploit a team’s weakness. If you can find a well priced player in a pace up game, who also has a DVP advantage, that’s ideal.
For example, Nicola Vucevic in recent weeks has been hot/cold. However, his best games in the last 2 weeks came when Frank Vogel knew that he could exploit front-court matchups with poor defenders like Jahlil Okafor, Hassan Whiteside and Clint Capela. If you examine those games, you will also see an increased number of attempts- higher usage- for Vuc, it’s clear the team understood how to exploit the matchup.
There are ways to expand on DVP knowledge. Investigate a player’s history against a team, you can do this easily on stats.nba.com player pages. Look at ESPN’s Defensive Real Plus Minus in order to understand the weakness or strength of the individual who likely guard your player. And, finally, profiling similar players performance in that matchup. (If Russ is playing the Knicks, look at what he did against them in the past and how other stud PGs have done against them, recently.)
DON’Ts: These are little more opinion driven but I think they are helpful
1- Don’t Fall for your Bias: Did you select this player because you like him?
Don’t play a player because he helped you out last game. Don’t select a player because you watch a lot of his games and know him better than another. Your lineup doesn’t care who you root for.
I fall into this trap with Jimmy Butler. I watch him a lot as a Bulls’ fan. But I have to know when I shouldn’t play him and when I should.
2- Don’t Chase Points: Did you select this player because of ‘last game’?
Hot streaks are only spastically proven real inside of games. They do not extend outside to consecutive games, statistically. Now, basketball intuition and Dion Waiters would say otherwise, but it’s been investigated and proven.
Regression is real. Players will not hit their ceiling every night. Michael Beasley was the world’s favorite play last week but we were chasing that 40pt game. He only scored 19 against a brutal Brooklyn defense. It wasn’t rocket science, it was regression to his season average. Fading that is a huge asset to your lineup. Don’t get FOMO, just get smart.
3- Don’t Force it: Are you trying to fit a player and compromising other spots?
I find myself comprising to fit in multiple studs at times because I love their matchup or projections. But in all, forcing high priced players into your lineup is only viable if your value plays are airtight. You may reduce your ceiling by choosing Frank Kaminsky over Jokic, but if it allows you to balance your selections in other places, then it is wise. At the end of the night, you want to have a fighting chance to be in the cash. Avoid risk in your cash game lineups by locking in a single stud, single value, then build around them with players with strong and clear roles. Grout guys, players with solid mins, clear floors and usage cause a lot less stress.
Watching 2 punt plays bust in the last game of the night is heartbreaking way to watch that little marker slide from green to white. (Hint: Lakers)
I will publish a quick reference version of this list that is more design friendly later this weekend.
Feel free to ask questions @heymrkane
And don’t forget the pod https://soundcloud.com/lineuplogicdfsbasketball