Thursday, Sep 5th
Monday, Sep 9th

End Around :: Week 8

Hilow’s End Around: Game Theory Training for DFS Play

Mark “Hilow” Garcia



Man, oh man, there are a lot of obvious places for ownership to congregate this week, and majority of it has a larger-than-ownership chance of failing at respective salaries (not providing the requisite score players are looking for; this week is far different than the last few). We also have three to four games with extreme weather in the forecast (the perception amongst a vast majority of DFS players is that weather (any weather) affects games in the same way, but the reality is that the biggest hindrance to fantasy production (and overall real-life scoring in the NFL) is wind, with rain and snow analytically having little to no affect on the game itself (or where fantasy points and box score numbers come from)). That last point was way longer than I intended, but I feel it’s a valuable point to understand, especially this week, with heavy winds and rain in the forecast for multiple games. As I allude to below, teams are not going to stop passing because it’s windy, they’re going to pass to different areas of the field than they otherwise would because of the wind. Wind (and my personal threshold is 25 mph+, anything below that is negligible to me; coming from an aviation background I know what 25mph wind looks like and what it can do to an object moving through the air) affects downfield passing (15+ intended air yards) and kicking the most, with these low-to-mid range aDOT pass-catchers seeing a slight uptick to their expected range of outcomes. What is most interesting to me is that the top four running backs in my expected ownership to start the week are all in games that are going to be affected by weather, and the perception will be that they all get a boost because of it, which in turn is likely to increase their ownership further this week (but we know, being the sharp gamers we are, that isn’t necessarily the case!). Let’s dig in!



In one of the best game environments of the week, with the heaviest winds forecast of the week, with Odell Beckham, Jr. out for the season, look for DFSers to get their exposure to this game through the home-favorite running back. Hunt brings one of the highest floors to the table this week, but as we’ve seen over the previous three weeks, his efficiency suffers the more touches he gets (paltry 3.67 yards per carry over that time). Wyatt Teller (PFF’s highest graded run-blocking guard in 2020) will also miss his second game in a row, further denting the ceiling for Hunt. The role is there, the matchup is there, but the per-touch efficiency leaves me wanting more when considering his price and expected ownership.

(Borderline BAD CHALK for large field GPPs)


There’s not much left to say regarding Henry. Always high ceiling, always low price-considered floor. This week I’d expect his ownership to creep up because of the weather and the dearth of pay-up options available on the slate, but without the pass game role and the possibility (likelihood?) of Cincinnati being able to dedicate an additional defender to the box or closer to the line of scrimmage, there chances of him failing are greater than the chances of his smashing. At ownership, I likely won’t have much.

(Borderline BAD CHALK for large field GPPs)


Gio is not Henry, but he does profile well to actually reap the benefits of the inclement weather (short passing) and actually carries a higher raw floor than henry at a steep discount. Ceiling will be highly dependent on touchdowns (same as Henry?) and we saw the Bengals call back-to-back quarterback sneaks at the goal line last week, so the chances of hitting said ceiling are lower than Henry, but cost-considered he is the preferred play.

(GOOD CHALK in all formats)


Locked into 18-22 high value “Green Bay Packers running back” touches against a defense that can be beat anywhere. Similar weather concerns as in Cincinnati, with the same logic being used to find the better play (he is a much stronger cost-considered play than Dalvin Cook); likely beneficiary of the inclement weather.

(GOOD CHALK in all formats)


Adams is averaging over 14 targets per game in his three healthy games. That is not a typo. Insane usage. His profile sets up well to not be affected by the weather as much as just about any wide receiver in the NFL, as he can beat both man (and press) and zone coverage equally. Locked in volume on a team projected to score points. Of the pay-up options available on the slate, he is the likeliest to return value.

(GOOD CHALK in all formats)


Brown is actually not simply a deep threat, as public perception would dictate. He comes into this game with a below average 10.0 aDOT and solid-yet-unspectacular 80.3 average air yards per game, and has increased his route tree when compared to last season. That said, he is likeliest of the Titans pass-catchers to see a hit to his ceiling when factoring in the weather, as he is the primary deep option on this heavy play-action offense.

(Neither GOOD CHALK or BAD CHALK with clear leverage to be covered shortly)


Let’s start the “chalk build” discussion by looking at another excerpt from my course behind the pay wall:

“Remember, we go into each slate with the same common knowledge as every other player, but DFS is a simultaneous game, meaning we do not know how other players are going to attack each slate. With this information, however, we can begin to formulate “chalk builds,” or how a majority of entries will choose to allocate their salary. We do this utilizing Non-Cooperative Game Theory, where we can begin to formulate conclusions that predict players’ actions through analyzing Nash Equilibria (a proposed solution to a Non-Cooperative game).”

The practice of analyzing where we expect ownership to settle, and thusly, where we expect people to allocate salary on a given slate, is driven by Game Theory Psychology. Many factors go into achieving this goal successfully, but if we can begin to understand how people are going to spend their salary, what positions we can expect them to do so at, and how different roster pieces come together, we can begin to formulate a plan to attack these “chalk builds” (make our rosters look different than others’ rosters without utilizing non-optimal plays).

This week, we can expect heavy three-running back teams, as we have four clear places for chalk to develop at the position, which becomes even more pronounced when you factor in expected weather. Similar to last week when I played Jamaal Williams and Davante Adams together to gain leverage (that is counterintuitive to generally accepted DFS practices), we look for ways to make our rosters as a whole look different from others’ as a means of gaining leverage. Most people are not going to want to play opposing running backs on the same lineup, so expect a good chunk of lineups to run Jamaal Williams (likely without Adams!), Kareem Hunt, and one of Giovani Bernard or Derrick Henry. Furthermore, those rosters that choose to utilize Gio, it is likely a good chunk include AJ Brown. So, without having to fade chalk completely, one way to handle the situation this week is to play Adams on the same lineup with Williams and/or play Henry on the same lineup with Gio (I’m not saying these players are must-plays, simply introducing more ideas on how to handle chalk!). Another easy way to utilize leverage is to play a wide receiver in the flex this week, as this week screams running back in the flex (so expect that to be a more popular roster construction!). 



Jeff Wilson, Jr. joins running backs Tevin Coleman (possibility he makes it back from IR for this game, but no guarantees) and Raheem Mostert on IR and Jerick McKinnon is “battling tired legs” (which has been obvious if you’ve watched the past three Niners games following McKinnon’s career high 92% snap rate in Week 4). With the emphasis amongst DFS players placed on the passing games from each team, Hasty brings potential lead back duties, in a Shanahan offense, at likely massive leverage off the San Francisco passing game to the table for Week 8 (Hasty himself has averaged 5.15 yards per carry on 20 totes this year, and is a capable pass-catcher out of the backfield). If Tevin does make it back for this game, it increases the leverage for Hasty but lowers his floor a good deal, and Tevin would carry high leverage in his own right. 


Those that fear uncertainty, please avert your eyes. Those looking for one of the highest leverage plays on the slate, read on. First off, let’s discuss some coaching tendencies out of Baltimore. In 2019, with a quarterback getting his first look at the NFL (and opposing teams getting their first look at the quarterback), John Harbaugh maintained his same “foot on the throat when they’re down” mentality, routinely wiping the opposition out of the stadium and running up scores. To start this season, we’ve seen a far more mundane Ravens team, appearing to side with the Andy Reid mentality, pulling starters off the field in blowouts, keeping Lamar mostly under wraps (lowering his per-game designed rushing attempts), and toning down the offense overall. In my opinion, we haven’t seen the full capability of this Ravens offense up to this point in the season. Enter the Steelers, setting up a battle for the lead in the AFC (Pittsburgh sit alone on top currently at 6-0, with the Ravens sniffing their heels at 5-1).

I broke down how strongly I feel about a floor plus ceiling game for Hollywood Brown, and he is likely to come at extremely low ownership in his own right (albeit with not much pure leverage as there isn’t expected to be overwhelming ownership from this game to begin with), but in a week in which quarterback is not as clear-cut as we’ve seen in the past few weeks, the pairing of Lamar Jackson and Hollywood Brown immediately differentiates your roster as a whole based on where you will be allocating salary.

This play requires a lot of trust in what I’ll call “speculative coaching tendencies,” in that there is a possibility this is the first week of 2020 where we see “post-season Ravens.” Here’s what I mean by that: did John Harbaugh shift his overall coaching tendencies this offseason, is he attempting to keep the playbook during the year different than what we can expect when games begin taking on a little more meaning, or have they simply attempted to keep their best chances of winning a championship under a thin layer of bubble wrap (Lamar Jackson)? This is a “bet on game environment being different than public perception” play (and I don’t fully believe this is a total tributary play, I think it is more likely the public perception of how this game is likely to play out is what is wrong), so the floor is lower than what we’d typically be looking for at their combined salary, but this is the type of play that wins people a million dollars in a large field GPP, as the ceiling does not match both public perception and chance at smashing when considering ownership.

When making a play like this, it is best to bet on the overall game environment overwhelming, instead of one side or one piece. Since that is the case, the highest Game Theory and DFS Theory leverage you can create is to bring back a Lamar Jackson/Hollywood Brown mini-stack with a piece of correlation from the opposition. Since the Ravens are so difficult to run on with the return of nose tackle Brandon Williams, the likely optimal correlation become a Steelers pass-catcher. Expect Marlon Humphrey to be tasked with primary coverage on Juju Smith-Schuster, leaving Diontae Johnson and his ridiculous 12.67 average targets per game across his three fully healthy games as the optimal correlation. With the amount of leverage achieved through this game stack, you could literally fill in the rest of the lineup with chalk and be differentiated, and you now need only one situation to go your way (a fully correlated game stack of the likeliest players to see elevated usage and box scores) instead of three (if you played three individual low-owned options paired with chalk).


Now that you’re done vomiting, allow me to explain. The weather forecast has bumped the winds projection slightly for the Titans/Bengals game. Corey Davis comes into Week 8 with a below-average-for-a-wide-receiver Average Depth of Target (aDOT) of 9.8 and has seen a team target market share of 21.7% in the four games he has been healthy (8, 5, 6, 10), paired with an above average 70.8 air yards per game. Yes, he is fantastic leverage off the likely chalk of Derrick Henry and AJ Brown, but the biggest plus (outside of the leverage) is the mix of where he is being used on the field and the expected weather during the game (I will update the weather on the four games with legitimate weather concerns on Saturday evening, so follow me on Twitter or check the SE/3-Max channel in discord for an update on the aviation forecasts). As it is currently forecast, we’re expecting 20-30 mph winds with swirling action based on the direction of the wind and how the stadium is oriented. Wind does not mean teams will not pass, it simply means they will have more difficulty hitting the deep areas of the field, and it should allow a smart defensive coordinator the opportunity to creep an additional man towards the line of scrimmage (so, as counter intuitive as it sounds at first, heavy wind is not an immediate boost to run games). What wind typically does boost is passes to the short-intermediate areas of the field. Enter Mr. Davis (this is also a boost to Jonnu Smith, but with the majority of his fantasy production coming in the way of touchdowns, and the fact that winds do drive down expected scoring in a game, he’s unlikely the player to most benefit here). With double-digit fantasy points in every game played, he brings a solid mix of floor and ceiling to the table priced at only $5,100. To gain even more leverage, pair Giovani Bernard on lineups that play Corey Davis, as his high pass game involvement falls under the same thoughts as Davis.


With how the chalk shakes out this week, and where we expect players to allocate salary, expect high ownership on the extremes of tight end pricing (high ownership on the pay-up pieces, and high ownership on the value pieces). This should leave very little ownership on the middle tier tight ends. Playing one of Noah Fant, Hunter Henry, Jonnu Smith, or Eric Ebron immediately differentiates your roster as it forces you to allocate salary differently than a vast majority of other rosters in GPPs, all through one high-variance position!

Want More Hilow?

You can catch Hilow (and Sonic, and Xandamere, and JM) on the OWS Discord server…

You can get inside Hilow’s DFS mind here.

You can also grab the first five lessons of Hilow’s Game Theory course for FREE.

You can listen to Hilow (and Lex, and Jordan Cooper) on Run To Daylight (hosted by TodFromPA || presented by OWS!), live at 3 PM Eastern this Saturday. (Note: the podcast runs live, but it will be archived shortly after it finishes.)

Collective Contest!

Drop Your Leverage Angles!

We’re looking for the best game theory // leverage angles on the slate! Drop your thoughts below, and let’s see where we end up!!!

Edge Points ::
Most upvotes below will get 25 Edge Points

Second most upvotes below will get 10 Edge Points

Third most upvotes below will get 5 Edge Points

Edge Points will be awarded based on upvote standings!