Run the Sims: Week 17 DraftKings Player Pool

Full Send Edition

Welcome to the Week 17 Player Pool! I'll be using the following sources of data to create a pool of players to use in both cash games and tournaments on Draftkings. If you play on Yahoo or Fanduel and want advice feel free to hit me up on the Gridiron AI Slack Channel.

To create the player pool, I run this super fancy (not really) program I wrote to aggregate projections from the following sources:

• Establish the Run Projections and Ownership Predictions
• Gridiron AI Projections
• FantasyLabs Projections
• FFAnalytics Expert Consensus Projections
• Rotoviz Game Level Similarity Projections

If any of this information is helpful to you, feel free to Buy Me a Coffee for my troubles.

You can find the projections here under the "Sims" tab.

The "DFS Cash Pool" was created using FantasyLabs optimizer tool to generate 150 lineups by median projection while limiting the max player exposure to 30% so I would have a more diverse portfolio of players in my lineups. The tab includes the aggregated median, floor, and ceiling projections. It also includes Ownership% and Exposure percentage. Ownership% is Establish the runs current prediction for how popular that player will be in the Million Dollar Maker Tournament. For cash, we can ignore this, for tournaments this becomes very important. (More on that later.) So for cash games, I tend to gravitate toward more popular plays as the "Wisdom of the Crowd" is usually a decent indicator of the "best plays".

The "DFS Tournament Pool" was generated in a similar way. Using the FantasyLabs optimizer tool I generated 150 lineups, this time by Ceiling Projection as you need your players to hit the higher end of their range of outcomes in order to win a tournament. Here are the rules I used to develop the tournament player pool:

• Stack QB with at least 1 WR or TE from the same team
• Stack WR with at least 1 WR or TE from the opposing team
• Include at least 3 players under 15% ownership
• Include at most 2 players over 25% ownership
• Limit entire lineup cumulative ownership to 125%

Note: these cutoffs are somewhat arbitrary, but are rooted in studies done on the most successful builds in top 1% tournament lineups.

Notice that unlike the cash pool, I am not forcing the optimizer to limit player exposure to 30% for the sake of diversification. I want to loosen the restrictions on the optimizer so that it is free to create lineups that fall within the upper 1% of potential outcomes. Another way to put this is that I am allowing the model to go "full send" on the plays it likes the most.

Tournament Fade of the Week

This week, I tweaked the process slightly by completely fading a single player that projects well and is going to be incredibly highly-owned(44%+): Antonio Brown. Excluding Brown from the player pool is a game-theory based play that will help us gain ground on approximately half of the field if we are correct about one thing. This is the type of leverage play that the best tournament players utilize in their leinup construction.

Why do I think the projections and field are wrong on Brown?

The beat out of Tampa Bay is neutral at best on Antonio Brown's status. It would not surprise me if he sat out.

On top of that, the Buccaneers leaned much more run heavy in their game without Evans, Godwin, and Fournette last week against the lowly Panthers. I think this is a similar spot against the Jets.

Update:Since writing this, the projections have been updated for points and ownership. In these update projections, Antonio Brown is no longer appearing in the optimal tournament lineups.

For the tournament pool, I added a column called "Leverage". This column is simply the exposure of the player in the 150 lineups - the ownership percentage. This number can tell us how much leverage we gain on the field if this player outperforms the more popular plays at the position. Again, allowing for a maximum exposure of 100% can show us the true leverage spots as recommended by the model.

Using leverage is a good way to determine what the model thinks are "good chalk" plays. If your total exposure to a player is significantly higher than the players projected ownership, then that is a likely a good play now matter what the ownership percentage is.

When choosing players from the pool I would prioritize ceiling first and foremost. Next, I would make sure that I am including only 1-2 higher owned players and filling in the rest of the lineup with high leverage players. Again, keep an eye on ceiling while doing this as a low owned players is great, but if they don't have a high ceiling, they may not help you gain a ton of ground on the field.

Lets Dive in!

Quarterback Cash Pool

For cash games this week I think it makes a lot of sense to play one of Dak Prescott or Justin Herbert as they both offer really solid projections at their current cost. Josh Allen is likely the safest play with the highest floor, but you'll have to pay all the way up for him which could put pressure on you elsewhere in terms of salary. For those that like to live dangerously (me), consider Trey Lance at $4800 against the Texans.

Quarterback Tournaments Pool

For Tournaments, the model recommends a full send on Trey Lance It is not often that we get a Quarterback with a 30+ point ceiling projection at $4800. Actually, it happened 2 weeks ago with Tyler Huntley and won someone a million dollars. This is a spot where Trey Lance should be massively owned, but as of right now he is not. As you can see from the leverage score (66.6), his ownership percentage can come up a lot before he is not considered a good play.

Update: Sadly, the model is no longer full sending on Trey Lance due to his increased ownership projection. Instead it is most into Dak stacks with some Josh Allen, Tyler Huntley, and Trey lance sprinkled in.

Running Back Cash Pool

At Running back, Jonathan Taylor projects on a completely different level than the rest of the backs, but is also priced that way. It is likely that the optimal cash build will include two of David Montgomery, Sony Michel, and Ronald Jones as they project for lead back volume at modest prices.

Running Back Tournament Pool

In tournaments, the model seems to like eating chalk at the position recommending a heavy exposure to David Montgomery and Sony Michel. Jonathan Taylor shows up here as well as the only play with with a ceiling projection over 30 points. There aren't a ton of exciting plays at the position this week. Because of this, I may prioritize paying up for Jonathan Taylor more than the model recommends as I think that the context of this slate could set up to be a spot where Jonathan Taylor outscores the next closest back by 10+ points.

Wide Receiver Cash Pool

Again, for cash, I am going to target players where the model and the field agree. As stated before, I will be fading Antonio Brown fully in tournaments, but taking on that risk is not necessary in cash. He is a really solid play as long as he is healthy. If we get any updates about AB's status before Sunday Kickoff I'll let you via live stream or Twitter.

Otherwise, I think an interesting decision can be made as to play Cooper Kupp or Jonathan Taylor. You can't play both. I think the mid-ranged plays are solid enough in cash that I am going to prioritize Kupp over Taylor. Braxton Berrios has flashed in his opportunities over the last few weeks. He is solid discount over another popular WR on a bad team in Amon-Ra St. Brown.

Wide Receiver Tournament Pool

In tournaments, the model really likes three contrarian options in Christian Kirk, Deebo Samuel, and Michael Gallup. Gallup is seeing plenty of volume on a team that will need to air it out against an Arizona Cardinals offense that should rebound after a few down weeks.

Tight End Cash Pool

I hate the tight end position. I will try to pay down as much as I can. My model absolutely loves Brevin Jordan and CJ Uzomah. Again, I hate tight end.

Tight End Tournament Pool

Again, the model absolutely loves Brevin Jordan as a bring back with Trey Lance stacks. He's an athletic pass catching tight end that is emerging as a target leader, albeit on a bad team. I am happy to go back to the well on George Kittle after a let down week, especially since we are fully send on Trey Lance.

Defense/Special/Teams Cash/Tournament Pool

New England Defense is a pretty clear smash play this week against the dumpster fire Jaguars. Bill Belichick typically dominates rookie QBs. A couple tips: Pay down, fade high-owned defenses in tournaments, and try to pair your defense with an RB thats on a team that is heavily favored at home.

If you made it this far thank you for reading. Let me know what you think of the article/plays on twitter @run_the_sims. Good luck everyone and don't forget to Buy Me a Coffee if you hit it big out there this week!