DFS Strategy[h=1]DFS Strategy: Basic principles of roster construction for cash games[/h]By Colin Drew on <time class="entry-time" itemprop="datePublished" datetime="2014-10-09T14:33:13+00:00">October 9, 2014</time> at <time class="entry-time" itemprop="datePublished" datetime="2014-10-09T14:33:13+00:00">2:33 pm</time>
Last week, we touched a bit on large-field tournament strategy. If you took notice across the industry, you saw many of the comments come to fruition.
DraftKings crowned a millionaire on just a $27 entry, and that winner employed multi-entry tactics as well as stacking Denver’s Peyton Manning with both Demaryius Thomas and Emmanuel Sanders. After anchoring his roster with a stack from a high-powered offense, he went contrarian at running back with Arian Foster (0.8-percent owned) and Ben Tate (0.7 percent), two guys returning from injury to feature-back roles. He rounded out his roster with Golden Tate, Antonio Gates, Terrence Williams and the Chargers defense.
You have to love the days when you can become a millionaire playing daily fantasy sports.
While tournaments can be great if you strike it big, they’re so high variance that most serious players invest the majority of their money in “cash games,” also known as 50/50 or head-to-head contests. I personally invest 75 percent of my weekly entries into this format and prefer to create three lineups at 25 percent each.
The next benchmark is understanding what score you’re looking for to cash an entry. While scoring varies week to week in the NFL, a safe target to shoot for is 120 points, which equates to two points per $1,000 of salary on FanDuel and 2.4 points per $1,000 of salary on DraftKings.
Do you expect the roster you constructed to produce 120 points? If not, it’s time to retool.
[h=3]CASH GAME STRATEGY[/h]There’s no playbook, and it kills me to hear blanket advice like “always pay for a stud QB.” Everything is salary and situation dependent.
Based on 2013 performance and pricing data from both DraftKings and FanDuel, QBs hit H2H value at a significantly higher rate than RBs, WRs or TEs. That was different in large-field GPPs, where QBs hit at a similar rate to other positions. What does that mean? Basically, it’s a lot easier to find value at QB in cash games than it is at other positions. So if it costs you a little extra to pay for a “high floor” at those other positions, do it, and then find your value pick at QB.
At its simplest level, cash games are about minimizing variance and reducing risk, while hopefully maintaining some semblance of upside. It doesn’t matter if you finish first or in the 49th percentile as long as you’re cashing. If you get a zero out of a key spot, it’s going to be tough to make up. Don’t target boom-or-bust players, and put a premium on consistency and high floors.
My favorite example of this from 2013 is Cam Newton and Ryan Tannehill. Newton ranked fifth in FanDuel scoring over the course of the season at about 20 points per game. Tannehill plugged along at about 16 ppg, all the way down at 23rd. But from a cash-game value perspective Tannehill was excellent, exceeding H2H value in 13 of his 16 games. Newton, on the other hand, was more of a roller-coaster. He only exceeded value in eight of 16 games.
In 2013, you were much better off slotting Tannehill into your cash-game lineups instead of Newton, given their difference in prices.
[h=3]OPPONENT SELECTION[/h]If you’re newbie, this might be under the radar. If you’re experienced, surely you’ve figured out that an absolutely critical element of the H2H environment is opponent selection.
Would you fancy your chances more against a 21-year-old college student with mom’s credit card, or a data-driven DFS professional with more than 120,000 games under his belt? It’s a no brainer, yet each week people donk off cash to players making a full-time living playing daily fantasy sports.
When selecting head-to-head opponents, do your homework and check their records. Avoid big names and guys posting hundreds of contests; it just isn’t worth it. While he is not god, playing Condia is not a +EV strategy.
[h=3]RESEARCH AND ROSTER CONSTRUCTION[/h]Matchups matter.
I typically begin my research with weekly projections from a third party, and then layer in snap count, overall targets, red-zone targets and pricing for both DraftKings and FanDuel. I use some consistency metrics and then calculate projected points per $1,000. Typically, a few players will jump out. If I trust the projections, I’ll pick a couple players of those players to anchor my lineups and build around.
Vegas lines are also an important element of my research, as it is an easy way to pinpoint projected high-scoring games and—given the importance of touchdowns—narrow down the player field. Similarly, it can help to project game flow (i.e., RBs will likely see more touches playing for a 10-point favorite than a 4-point underdog) and bake it into your opinions.
Last, we touched on seeking positive correlation by “stacking” in GPPs. While I would never advise pairing a QB with a traditional RB in a large GPP, it can be a great way to secure safe points in a H2H game. There were several times last year where I trotted out Manning and Knowshon Moreno, feeling pretty comfortable that I’d grab 4 TDs from those two players, elevating my floor.
Another way to reduce risk is to make sure to have highly targeted WRs and TEs and high-snap count RBs. In general, I avoid RB timeshares and guys who could be eliminated from the plan if the game flow doesn’t go their way.
In summary, look for high-floor players with low risk, target bigger favorites and games with high totals, and don’t feel like you need to get stuck paying for a top QB. And, whatever you do, there’s no reason to go after professional DFS players when these sites are spending millions of dollars to bring new players in each week.
[h=3]TWO-MINUTE RUNDOWN[/h]Here’s a quick look at some guys on my cash-game radar at DraftKings this weekend.
Last week, we touched a bit on large-field tournament strategy. If you took notice across the industry, you saw many of the comments come to fruition.
DraftKings crowned a millionaire on just a $27 entry, and that winner employed multi-entry tactics as well as stacking Denver’s Peyton Manning with both Demaryius Thomas and Emmanuel Sanders. After anchoring his roster with a stack from a high-powered offense, he went contrarian at running back with Arian Foster (0.8-percent owned) and Ben Tate (0.7 percent), two guys returning from injury to feature-back roles. He rounded out his roster with Golden Tate, Antonio Gates, Terrence Williams and the Chargers defense.
You have to love the days when you can become a millionaire playing daily fantasy sports.
While tournaments can be great if you strike it big, they’re so high variance that most serious players invest the majority of their money in “cash games,” also known as 50/50 or head-to-head contests. I personally invest 75 percent of my weekly entries into this format and prefer to create three lineups at 25 percent each.
The next benchmark is understanding what score you’re looking for to cash an entry. While scoring varies week to week in the NFL, a safe target to shoot for is 120 points, which equates to two points per $1,000 of salary on FanDuel and 2.4 points per $1,000 of salary on DraftKings.
Do you expect the roster you constructed to produce 120 points? If not, it’s time to retool.
[h=3]CASH GAME STRATEGY[/h]There’s no playbook, and it kills me to hear blanket advice like “always pay for a stud QB.” Everything is salary and situation dependent.
Based on 2013 performance and pricing data from both DraftKings and FanDuel, QBs hit H2H value at a significantly higher rate than RBs, WRs or TEs. That was different in large-field GPPs, where QBs hit at a similar rate to other positions. What does that mean? Basically, it’s a lot easier to find value at QB in cash games than it is at other positions. So if it costs you a little extra to pay for a “high floor” at those other positions, do it, and then find your value pick at QB.
At its simplest level, cash games are about minimizing variance and reducing risk, while hopefully maintaining some semblance of upside. It doesn’t matter if you finish first or in the 49th percentile as long as you’re cashing. If you get a zero out of a key spot, it’s going to be tough to make up. Don’t target boom-or-bust players, and put a premium on consistency and high floors.
My favorite example of this from 2013 is Cam Newton and Ryan Tannehill. Newton ranked fifth in FanDuel scoring over the course of the season at about 20 points per game. Tannehill plugged along at about 16 ppg, all the way down at 23rd. But from a cash-game value perspective Tannehill was excellent, exceeding H2H value in 13 of his 16 games. Newton, on the other hand, was more of a roller-coaster. He only exceeded value in eight of 16 games.
In 2013, you were much better off slotting Tannehill into your cash-game lineups instead of Newton, given their difference in prices.
[h=3]OPPONENT SELECTION[/h]If you’re newbie, this might be under the radar. If you’re experienced, surely you’ve figured out that an absolutely critical element of the H2H environment is opponent selection.
Would you fancy your chances more against a 21-year-old college student with mom’s credit card, or a data-driven DFS professional with more than 120,000 games under his belt? It’s a no brainer, yet each week people donk off cash to players making a full-time living playing daily fantasy sports.
When selecting head-to-head opponents, do your homework and check their records. Avoid big names and guys posting hundreds of contests; it just isn’t worth it. While he is not god, playing Condia is not a +EV strategy.
[h=3]RESEARCH AND ROSTER CONSTRUCTION[/h]Matchups matter.
I typically begin my research with weekly projections from a third party, and then layer in snap count, overall targets, red-zone targets and pricing for both DraftKings and FanDuel. I use some consistency metrics and then calculate projected points per $1,000. Typically, a few players will jump out. If I trust the projections, I’ll pick a couple players of those players to anchor my lineups and build around.
Vegas lines are also an important element of my research, as it is an easy way to pinpoint projected high-scoring games and—given the importance of touchdowns—narrow down the player field. Similarly, it can help to project game flow (i.e., RBs will likely see more touches playing for a 10-point favorite than a 4-point underdog) and bake it into your opinions.
Last, we touched on seeking positive correlation by “stacking” in GPPs. While I would never advise pairing a QB with a traditional RB in a large GPP, it can be a great way to secure safe points in a H2H game. There were several times last year where I trotted out Manning and Knowshon Moreno, feeling pretty comfortable that I’d grab 4 TDs from those two players, elevating my floor.
Another way to reduce risk is to make sure to have highly targeted WRs and TEs and high-snap count RBs. In general, I avoid RB timeshares and guys who could be eliminated from the plan if the game flow doesn’t go their way.
In summary, look for high-floor players with low risk, target bigger favorites and games with high totals, and don’t feel like you need to get stuck paying for a top QB. And, whatever you do, there’s no reason to go after professional DFS players when these sites are spending millions of dollars to bring new players in each week.
[h=3]TWO-MINUTE RUNDOWN[/h]Here’s a quick look at some guys on my cash-game radar at DraftKings this weekend.
- QBs: Matt Ryan, Russell Wilson, Nick Foles, Ben Roethlisburger, Ryan Tannehill
- RBs: Matt Forte, Le’Veon Bell, LeSean McCoy, Andre Ellington, Andre Williams, Lions RB (if either Joique Bell or Reggie Bush sits)
- WRs: Julio Jones, Antonio Brown, Brandon Marshall, Alshon Jeffrey, Julian Edelman, Pierre Garcon, Larry Fitzgerald, Andrew Hawkins
- TEs: Julius Thomas, Martellius Bennett, Larry Donnell
