DFS Roundtable: Week 7

A peek behind the curtain at a staff discussion pertaining to this week's DFS topics

This week, we'll discuss the following topics:

The staffers we talked to this week are Phil Alexander, Chris Feery, Will Grant, Devin Knotts, John Mamula, and Alex Miglio, with special “cameo” appearances from David Dodds and Jeff Pasquino.

GPP Roster Construction

Hester: Tell us how you construct your rosters for GPPs. Do you pick a stack you like first and build around it? If so, do you prefer the stack be contrarian and you build more common players around it? Or do you start with two-to-three players in every GPP lineup and then build variations around that core? How do your strategies change (if at all) for single-entry contests vs. large field multi-entry games? Discuss how contrarian you feel you need to be for specific contest types.

Alexander: I devote the majority of my bankroll to five entry max GPPs ranging from $2 to $12 each week. Since these types of tournaments have fewer entries, the scores needed to make a deep run usually aren't quite as high as the larger field games. For that reason, it's OK to play chalkier lineups in these contests, provided you sneak at least one or two fairly low-owned players onto your roster.

For a long time, I began each of my lineups with a specific quarterback-pass catcher stack, added in my core players, and filled in around them until I had a roster I liked. Then I would save that lineup, import it into my next entry, and I'd swap out the "filler" players around the stack/core combo to get a different variation. I would repeat this process until I had about six to eight variations on each stack. It should have taken me a shorter time to realize this isn't the best way to go about it if the goal is to spread your player exposure out properly over 40-60 unique lineups each week. I was ending up with my exposure "clustered" into certain stacks and roster constructions, which resulted in one or two bad plays sinking nearly all of my otherwise great stacks.

Now, I try to begin each lineup by paying up at different positions. It allows you to build more varied rosters around your core players, without having too many lineups with near-identical construction. Building multiple lineups this way forces you to find the "value pockets" at each position, while also constructing many rosters in a way the rest of the field will never think of.

Mamula: Roster construction for GPPs should depend on how many entries that you plan to enter. The first step is to figure your bankroll allotment for the week and enter your contests with a default lineup. If you are building 10 or fewer GPP lineups, I would roster a core of three or four players across all of your teams. These core players do not necessarily have to be chalk plays, just players that you believe have a strong workload and will return value for their perceived matchup. For every lineup that you build, enter these players first, then alternate stacks and corresponding players from the top two or three games that you are targeting. Try to differentiate the lineups and have one or two players that you estimate will carry less than 10% ownership percentage.

On a typical NFL week, I build 50-100 unique GPP lineups. I play a variety of contests including single entries, three-max entries, five-max entries, 100-person leagues, and large field GPPs. I prefer to have as many unique GPP rosters as possible to cover my entries (min cash/high floor) with potential for a big payday (top-three finish/high ceiling.) If you are playing cash games for the majority of your weekly bankroll allotment, you are limiting your high ceiling opportunities.

When building a significant amount of lineups, I alter the core three to four player strategy. With mass entry GPP, you need to know how your individual player exposure shakes out. I rarely go over 30% player exposure for NFL GPPs unless I am willing to accept the risk involved. This past week, my player exposure was 42% for LeSean McCoy over 81 lineups. In hindsight, it would have been great to have 100% to McCoy in GPP. But, had he gone down with an injury early (which nearly happened) or struggled, it would have sunk a majority of my teams. I was willing to accept 42% exposure with McCoy because the perceived matchup outweighed the risk. The key is to estimate what you expect ownership percentage to be for each of the key players and then adjust your player exposure whether you are bullish or bearish for their individual matchups.

Feery: If you're playing in a large field event where you have to finish very high (such as a qualifier for a live final), you’ll definitely need to apply some contrarian thinking to your lineup construction. There are two different approaches you can take in that respect, and it really depends on what the week as a whole looks like from a DFS perspective to determine which makes more sense. Let’s walk through both approaches, which can be used for other GPPs as well.

For the first approach, I look towards my two favorite spots to mix things up: quarterback and defense. Nailing the under the radar selection that happens to go off at these positions can prove to be quite fruitful if the rest of your lineup produces as well. That requires some digging, and you can start by answering a few simple questions.

Which players are the chalk for the week at their respective positions?
Of the remaining players, which ones simply do not have the upside or ceiling to make some serious noise on the leaderboards?

You can eliminate the players that turn up as answers from both categories, and then focus your contrarian research efforts on the remaining positions. Question number two can eliminate some players that end up having fantastic weeks, but tough decisions are necessary to better focus your research. If you’re planning on playing a boatload of lineups, you can bypass question two if you feel it’s absolutely necessary to have an enormous player pool to choose from.

Once you have the base of your quarterback and defense, add in the signal caller's top target – i.e. target most likely to blow up that week – and start building out the remainder of your lineup. There’s plenty of ways to skin the cat there, but I’m perfectly content to add in the chalk and value plays I’m most confident in for the remaining positions, as my differentiation is built in at QB and DEF. For a little extra uniqueness, I may sprinkle in an outside the box RB, WR, or TE that I’m extremely high on for the week.

For approach two, it’s a similar process, but the contrarian position targets are switched up. I’ll look for a top contrarian selection at RB, WR, and TE, and then feel free to add in a more highly-owned QB and DEF.

The process can be expanded for multiple lineups as well. Approach one can be done with a few different quarterbacks and defenses around the same core, and approach two can look towards a chalkier quarterback and defense with multiple combinations of contrarian selections in the middle.

Grant: I really look at GPPs as 'intelligent' lottery tickets. I'm not a big stakes player like a lot of guys on the staff and have no problem playing in GPPs with $5 or fewer entry fees. However, I will also play in Qualifier, Satellite, and Survivor style lineups as well. In doing this, I've been able to earn tickets to play in the bigger contests, including the "million dollar" contests on the bigger sites.

When constructing my lineups, I look at the game script. I want stacks of players from the games that I expect where one team (or both) will score a lot of points. For example, this week I'll target the San Diego at Atlanta and New Orleans at Kansas City games, both of which have projected totals above 50 points. From there I'll pair a quarterback and one or two receivers from one team (like Matt Ryan and Julio Jones) and then build the rest of the lineup around them. I build my cash lineups first, so by the time I get to GPP, I have several strong value plays that I can plug into these lineups to build that strong base.

I'll target single-entry GPPs whenever possible, hoping that the lineup I enter will be the one that hits. I normally don't have more than six or seven total GPP lineups, so jumping into the big GPP contests doesn't make a lot of sense for me. By entering these lineups, if my value picks are good, then even if my stack doesn't hit completely, I will still do well enough to cash.

Knotts: I am a big fan of what I call a core GPP strategy. Far too often, I have found that I have all of the right players on a GPP team, but they aren't on the same team. So what I have done is I have gone out and created a strategy that puts a core group of players on all of my GPP rosters, with the primary theory that if this core of two to four players hits, then I will have built a few different combinations of rosters around this core group of players that one or several of my GPPs should end up very well. It is a high-risk strategy, as if that core group of players does not succeed, it will end up being a very difficult to cash no matter what combination that you have.

However, GPPs are hard by their very nature, and they are built so that the top 1-5% of rosters are what succeed. So instead of trying to build the perfect combination of all players, I focus on building the perfect combination of two to four players and then fill in the rest of the rosters with players that I like but don't want 100% ownership to.

[return to top]


Allocation of Funds

Hester: Let's say you only have time to create one lineup this week and you have $200 to put in play with it. How would you allocate that money to different contest types?

Knotts: I'm going to do some self-promoting here, as I have a DFS blog where I do this on a weekly basis, evaluating contests and how to enter GPPs. For me, I am always a 90/10 player in terms of cash games and GPPs. This means that in this $200 scenario, I would have $180 in cash games and $20 in GPPs. For cash games, it is all about risk tolerance.

If you want the most upside, I would play $150 in 50/50 games and $30 in 3-player and 5-player contests. If you're looking to build your bankroll (which is the premise of the blog), I recommend doing a high number of head-to-head contests, as this gets you more variety within opponents. From a GPP standpoint, since you're only building one roster, look at the single entry GPP contests that are at FanDuel. This is very scalable across all levels of bankrolls, so if your bankroll is $20, do $15 in 50/50 contests, $3 in 3-player and 5-player contests, and a $2 GPP.

Dodds: I like to build cash lineups by stacking my QB with a WR or TE. I also sometimes add an opponent's WR (as that team tries to play catch up). What I am looking for is the high-game of the week (the shootout that combines for a lot of scoring). This type of roster can do well in cash and GPPs if constructed right. I build lineups like this because I want to earn more than 2x my money when it goes off. And I do that by spreading my money something like this: 50% in large-field, low-cost double-ups or 50/50s, 20% in Triple Up contests, 10% in Quintuple Up contests, 10% in 100-man leagues that pay the top 12, 10% in large field $1-$5 GPP.

So with the $200 example, I would have $100 in true cash, $40 in Triple Ups, $20 in Quintuple Ups, $20 in 100-man, and $20 in GPP. An example of this kind of lineup would be to start with Matt Ryan-Julio Jones stack (highest implied total based on Vegas odds) and then choose normal cash plays around said stack. If Ryan and Jones connect for a big game, you could very well be on your way to a nice score with this lineup across multiple contest types.

Mamula: If I only had time to create one lineup for the week, I would target a QB-WR stack that provides both a high floor and a high ceiling. Last week, Drew Brees and Brandin Cooks would have been the perfect stack to meet these criteria due to their perceived matchup and game total. With only one lineup, you should concentrate on single-entry cash games due to the risk involved. With a $200 bankroll allotment, I would spread 70% ($140) across as many single-entry 50/50 and double-up contests as possible.

I would start with $1 games and select contests with at least 100 players or more. After you select all the $1 contests that meet your criteria, I would then move up to $2 level and continue until you spread the 70% across a variety of contests. For the remaining 30% ($60), I would allocate 10% ($20) in 100-man leagues, 10% ($20) in large field GPPs, and 10% ($20) in Triple Ups and Quintuple Ups. These types of contests provide your lineup with a high ceiling if your stack and complementary players both excel.

Feery: When constructing just one lineup, I’m looking for a combination of safety and upside, and that translates into how I’ll allocate my funds in play as well. For a single lineup, I’ll go with a split of 80%-10%-10%, which is broken down as follows: 80% $1 and $2 50/50s and Double Ups, 10% in 100-man leagues, 10% GPPs.

Assuming my research for the week is on point (and no devastating in-game injuries occur), I’m confident I’ll be in the running to cash in all of the 50/50s and Double Ups. If the lineup happens to perform even better than that, I’ll be in the running to cash in the leagues and tournaments as well. If I happen to knock it out of the park, the sky's the limit on overall return for the week.

As the others have mentioned, I’ll look towards the quarterback that I’m highest on for the week, and then add in his preferred passing game target. Next, I’ll narrow the defenses down to the ones I feel most secure about, and then choose the one with the most upside for the week. The process is similar for the remaining positions, but I’ll be more mindful of looking for intriguing and safe value plays to ensure that the salary cap isn’t an issue.

Grant: I'm a bit more conservative, so if I only had one lineup to enter, I'd look for a similar approach to what Devin describes – 90% in cash games with 10% in GPPs. I'd look for running backs with high floors, who will get a lot of touches and then try to stack a quarterback with a receiver that's in one of the higher projected scoring opportunities of that week (like San Diego-Atlanta this week). I'd approach the games like John describes above, though – targeting 50/50s and Double Ups that have the most possible entries – targeting the $1, $2, and $5 lineups where I think I can face as many people as possible. Then I'd throw a couple small GPP lottery tickets out there, but not expecting a lot unless I got really get lucky and hit on everything.

David has a good approach with stacking an opposing WR against your QB-WR, but with only one lineup, I don't think I'd put all my eggs in that basket because a one-sided win can leave you falling short. The San Diego-Atlanta game might be that exception, though, given the super high point total and potential shoot-out.

Miglio: The guys have hit the nail on the head – cash, cash, cash. You have a much, much better chance at a good ROI placing one cash-friendly lineup into cash games than to spend your money entering a bunch of different GPPs. As Will said, that lineup should also be entered into some GPPs because few things are more gut-wrenching in DFS than having a terrific cash lineup that isn't entered into any GPPs.

[return to top]


Chalk Talk

Hester: Last week, we were given the gift of a high-usage running back playing in a home game against a team that had given up four straight 100-yards rushers. Playing LeSean McCoy was an easy call that worked out beautifully. Which player this week is the most obvious cash game play?

Alexander: I'll have a hard time not clicking on Julio Jones this week. Jones looked dominant in roasting the Seahawks vaunted secondary for a 7-139-1 receiving line in Seattle last week. Now he returns home, where he's averaged over 11% more fantasy points per game since 2014, to face a San Diego team that's been burned by opposing WR1s. T.Y. Hilton lit up the Chargers for an 8-174-1 line in Week 3, and Amari Cooper finished with 6-138-1 in Week 5 (a game in which Cooper probably should have caught three more touchdowns). Football Outsiders ranks San Diego 29th at defending the oppostion's top receiver, a ranking that may not have fully bottomed out yet, with Jason Verrett out for the season. To top it off, Atlanta's 29.5 point implied team total is the highest on this week's slate. Fade at your own risk in any format.

Mamula: DeMarco Murray is in a prime spot to bounce back strong this week. I have targeted the Indianapolis rush defense every week so far this season. The Colts have allowed 139.7 rushing yards per game over their past three games. Last week, Lamar Miller gashed them for 178 total yards and two touchdowns. The week prior, Jordan Howard had a career-day with 163 total yards and one touchdown. Expect similar results from Murray this week. Murray is one of the top bell cow backs in the league with at least 18 touches in all six games this season. He meets the criteria of opportunity plus matchup for this week.

Grant: I like Giovani Bernard this week, especially in full PPR formats. Vegas expects the Bengals to win by 10 points. That seems like a ton given how they have been playing as of late, but Cleveland coming to town cures a lot of pain and suffering. Bernard and Jeremy Hill will split touches pretty evenly, but Bernard is significantly more productive in the passing game. The Browns have given up six touchdowns to running backs over the last four weeks, so I think there's plenty of room for Bernard and Hill to have a decent game.

Feery: I’m leaning towards Mike Evans as one of the week’s more popular plays that should have little trouble producing. Tampa Bay receives the privilege of facing off against a 49ers team that has allowed 10 touchdowns to opposing wideouts thus far, and it’s not hard to envision Evans doing his part to increase that total. Add in the fact that the Buccaneers have question marks at running back, and they just may approach the game with a plan that’s more heavily tilted towards the pass. Evans receives a healthy amount of targets and is in a prime position to score. He’ll make his way into a good portion of my cash game lineups this week.

Knotts: This is Rob Gronkowski week, everybody. There are a number of value plays on the board this week that can allow you to pay up for Gronkowski. He has shown that he is getting back to 100% healthy as he has gone over 100 yards in each of his last two games including 162 yards and a touchdown last week against Cincinnati. Pittsburgh is coming off a week where they allowed over 100 yards to Miami tight ends, including 81 yards to MarQueis Gray. There is enough value on the board this week to go up to Gronkowski and separate yourself compared to other tight ends on the board.

Miglio: Quarterback is tough this week, which might be why the guys have stayed away from the position thus far. As far as chalk goes, how about Andy Dalton? The Cincinnati offense hasn't been great, but they go up against a defense that has given up the second-most fantasy points per game to opposing quarterbacks this season. The game script calls for more running back scoring in Cincinnati, given that the Bengals are 10-point favorites (hence Will's love for Bernard), but it seems like Dalton will have a pretty good floor against a bad pass defense.

[return to top]


The Forgotten Roster Spot

Hester: Generally, when we look for defenses to play, we want teams that are big home favorites as that game script promotes an inferior opponent passing frequently and generating chances for sacks and interceptions.

This week, however, on the main slate, we have just three home favorites of more than six points. Kansas City and Atlanta play against potent offenses in New Orleans and San Diego, respectively. The third, Cincinnati, has been a bad defense this year and faces a Cleveland team that hasn't been the dream matchup most would have thought. Where are you going for defenses in cash games and GPPs this week?

Knotts: For cash, I am all in on the New England Patriots. With Ben Roethlisberger out with an injured meniscus, Landry Jones is expected to make the start. Jones has thrown four interceptions in just 56 career passes. Jones is going to be forced to throw in this game as Tom Brady goes up against one of the worst passing defenses in the NFL and will be able to put up points at will. This creates a great opportunity for the Patriots as they won't give up a lot of points and should be able to force a few turnovers.

For GPP, I am going with the Minnesota Vikings on the road against Carson Wentz. Wentz looked like a rookie for the first time last week where he was just 11-22 for 179 yards. This has one of the lowest totals on the board, and the Vikings are third in the NFL in team sacks. Wentz has only thrown one interception so far this year, but he has been sacked 12 times, so I expect that the Vikings will be able to get to Wentz in this one.

Alexander: I'm with Devin. New England and Minnesota look like the most attractive options this week. I'd lean Vikings in both cash and GPP, since all I'm ever looking for in a team defense, regardless of format, is to avoid a bust. Through the first five weeks, Minnesota has ranked as a top-five unit four times. They've finished with single-digit fantasy points only once and reached 20 fantasy points twice. With two weeks to prepare for a rookie quarterback, it seems the rest of the league is beginning to figure out, the Vikings once again have both the highest floor and highest ceiling of any team defense.

Pasquino: Minnesota is a good call with Lane Johnson suspended for the Eagles. That loss is a big one, and Wentz was under a lot of heat last week from Washington (WASHINGTON!!!) without him. Just imagine how ugly that could look against Minnesota with two weeks to prepare.

Mamula: I agree with Devin and Phil. Minnesota and New England are the clear cut top two defenses for the week. When building my GPP lineups, I try to have exposure to three to six defenses so I don't have all of my eggs in one basket. For DFS, defense is often a high-variance position that can determine a top GPP score, if multiple touchdowns are involved. Another defense that I am targeting this week is Buffalo. The Bills defense has demonstrated success with generating pressure on opposing quarterbacks and forcing turnovers. They have at least three sacks in five out of six games and at least one turnover in every game. The Bills also have three defensive touchdowns on the season. Miami excelled in the run game last week because Pittsburgh couldn't tackle or generate pressure on Ryan Tannehill. Expect the Dolphins to come crashing back to earth this week as the Bills will dominate in the trenches. Tannehill had seven interceptions in the four games prior to last week.

Grant: Count me in with New England as the defense to take, simply because of the price. For a defense, you want to set it and forget it, and the Patriots make the most sense this week.

But there are couple other options to consider. If you're playing the Monday night slate, you might think about taking Denver at home against Houston. The Texans looked awful against Indianapolis last week, and now they travel to Denver, where the Broncos have been thinking about their loss to San Diego since last Thursday. #NarrativeStreet says they welcome Brock Osweiler back to the stadium by blitzing him as he gets off the bus, in the hotel room, during warm-ups and even after the game.

They are not minimum-priced, but Baltimore travels to New York to take on the Jets this week. And given the Jets' total lack of offensive production on Monday Night Football, you might consider the Ravens as an alternative. Ryan Fitzpatrick was benched in favor of Geno Smith last week against the Cardinals, and reports suggest Smith will be the starter this week against the Ravens. Regardless of who starts, a bad quarterback is starting for the Jets. Baltimore has played okay against weaker quarterbacks this season like Kirk Cousins, Blake Bortles, and Tyrod Taylor. They wouldn't be my first choice, but if you're looking for lineup differentiation, you could do a lot worse than the Ravens.

Feery: The guys have nailed the top choices at defense this week, and I’m also high on both Minnesota and New England. For Minnesota, the club is simply playing at another level defensively. I’m more than happy to keep riding it while it lasts. The Patriots are facing off against Landry Jones, and what was once a highly-anticipated matchup has turned into a potentially lopsided affair.

I’ll echo the point on the Broncos defense for those playing the Monday Night slate as well. Coming in off of a disappointing loss - and with some extra time to think about it to boot - does not bode too well for the prospects of Brock Osweiler and the Texans offense this week.

One other name to mention is Cincinnati, who will have the pleasure of facing off against Cleveland. Teams facing off against the Browns quickly find their way onto my short list for the week, and the Bengals offer up some nice upside due to the decent potential for turnovers from the Browns.

Miglio: I have been looking at New England as well. But it seems that is passé at this point, so why not take a true GPP shot – the New York Jets. That defense has been bad this season, so I expect ownership percentage to be down. But the over-under on that game is at just 40.5 now, meaning it should be a low-scoring affair. The Baltimore offense isn't particularly good, and the Ravens are on the road. The price and likely ownership percentage are going to be awfully tempting in GPP lineups.

[return to top]


Follow the contributors of this Staff Roundtable on Twitter using the buttons below!

Phil Alexander

Chris Feery

Will Grant

Devin Knotts

John Mamula

Alex Miglio

Ryan Hester - Moderator