More

    Beer Money Bets | Creating a Process for NFL DFS Pt. 3

    The worst part of creating a process is you’re going to fail. It’s even worse when money is on the line and the difference between winning and losing is drinking a terrible light beer versus drinking a fancy craft beer while watching football. Last week was one of those light beer weeks. With so many significant injuries to key fantasy players, even the most perfectly made lineup wasn’t safe. Luckily, my first experiment with DraftKings Classic contest last week was only slightly derailed by the injuries. I was able to finish in the top 48% of entries in my contest pool. While I am happy that I beat over half of all lineups, only the top 1,500 lineups cashed, and my 1645th ranked lineup missed out by 3.36 points. Let’s dive in and see what went right and what went wrong in last week’s process.

    Reviewing the Process: (Last weeks process)

    1. I played only the early games and looked at selecting players only from the teams with the highest implied totals, this included DAL, TB, GB, TEN, IND, and ATL. The highest scoring teams turned out to be GB, DAL, ATL, LAR, TEN, and then TB, BUF, and SF all scored 31. After Vegas got 5 out of 6 correct, I am incredibly impressed at how well this worked. I am sure every week will not be this good, but I have no reason to stop using this as the first step of my process. 
    2. Nyheim Hines… Oh wow did I screw this up. This was a miss that is definitely something that can derail a week. I identified the Colts as a team with a high total, knew they were favored, and then played their pass catching running back; what was I thinking? I fell in love with the previous weeks results and ignored the red flag signs for the optimistic outlook. I really should have looked at playing either Fournette or Ronald Jones who were both priced lower than Hines and were going to see a friendlier game script for their role in the offense. To correct for this, I need to dive a little deeper and find better ways to select players on the teams I am targeting instead of blindly picking players that fit under salary. Also I will add that Davante Adams busted, but considering it was injury related, I do not consider it a bad process.  
    3. Corey Davis and the Rams DST performed well as my cheap options and my only mistake was Chris Herndon in my TE position. Going forward, the process might be to either go dirt cheap with TE and hope for a touchdown or spend up for a reliable player. Looking back, I don’t think Herndon was overall a bad play. He was dirt cheap for his predicted production and I am not going to let one bad data point change this process although I will be more careful choosing these cheap players going forward.

    The goal of this article is to find a quick and easy way to become profitable in cash games in DFS. The process I have been making so far has taken me less than 15 minutes to create a lineup and I want to continue to try and keep it that way. I think the only way I will be able to continue to do this is to look to other experts in this field for data. I have found the data tables that Peter Howard (@pahowdy) has on his Patreon (link here) to be extremely helpful, especially his DFS specific data.

    New Process for Classic Double-Up:

    1. Look at implied totals, identify the teams with the highest totals and select players from those teams. How many teams depends on the number of teams on the slate, but I am going to try and keep it to only the top 25% to 30%. 
    2. Sort players from these teams by projected points per $100 (Peter Howard’s data makes this really easy to do). Use this as a lens to look for value with a focus on running backs that will benefit from a positive game script. Select Wide Receivers and Tight Ends with high target shares in their offense. Select Quarterback based on a player that scores highly in projected points per $100 salary. Do not be afraid to go for high priced and “chalky” players. 
    3. To solidify the last few players in my lineup, look for players that could come at a discount. This includes players that benefit from in-week injuries as well as pass catchers on teams that are heavy underdogs, mostly focusing on pass catching running backs. 
    4. Choose a DST that is going against a team with a low implied total and preferably at home. Do not spend a lot of salary on the DST.

    Next, I want to go back to the first process I made and refined in this series. There are plenty of showdown games to play each week and I think getting good at these will overall be the most beneficial. Because there are more of these contests to play, I can get a larger sample size to test my processes and evaluate their outcomes. Finding ways to increase my data sets to see what works and what doesn’t, will allow me to feel more confident about what I am doing, and in turn how I can better guide you as well.

    My biggest issue so far with my process has been predicting game script. If the game goes the way Vegas and I expect, I usually cash, if it doesn’t, I lose money. I am only trying to finish in the top-half of lineups, which makes me wonder if my process is fundamentally flawed since I am looking at charts that create optimal lineups, when really that isn’t the exact goal. To adjust for this, I am going to incorporate projected points more into the process.  

    Revised Process for Showdown Double Up:

    1. Choose a game and sort the players by projected points. For my projections I use the data from Peter Howard’s Patreon that I plugged earlier in this article. Use players with the top 10 most projected points as a possible player pool to initially focus on.
    2. Use Vegas lines to see what type of contest the game applies to. We have 3 defined special categories: 
      1. Shootout – more than 50 total points.
      2. Slugfest – less than 40 total points.
      3. Blowout – Team win by 15+ points. (Vegas lines almost never approach 15 points so I will apply this to games with spread greater than or equal to 7.5).
      4. If none of these apply, look at “All” category for guidance.
    3. Use the following charts from Kevin Cole (link and link to articles) to create a lineup.
      1. Choose a Captain first based on the chart in the resources section of the article. I have found a lot of success in avoiding QBs in the Captain spot to save salary. The only exception to not playing a QB as a Captain would be in Slugfest games.
      2. Create a value for projected points per $100 for the player pool. Compare these values with the chart for roster construction to fill out the lineup. If the salary makes it hard to find a final player at a cheap price, look for touchdown upside on the team with the higher implied total. 
    4. Make sure to stay up to date with game time news in case of injuries.

    The New Process in Action:

    1. I am going to look at the Week 3- New England versus Las Vegas game for this example. The 10 players with the most projected points are Newton, Jacobs, Edelman, Carr, Waller, Harry, Michel, Edwards, Renfrow, and Burkhead. 
    2. This game currently has a 47.5 O/U with the Patriots -5.5 so it does not fall into any of our special categories. 
    3.  
      1. I want to use Newton as my captain but to save some salary I am going to use someone cheaper. New England is one of the few teams that doesn’t have a solidified RB1, so I am going to use Edelman as my Captain since he is the WR1 on the favored team. 
      2. I am going to try and do something creative here and multiply the projected points per $100 by the percent of times that type of player is found on optimal lineups(on the chart in the resources section below). I will use those values to fill my lineup by selecting the players with the highest values. This gives me Newton, Michel, Edelman, Jacobs, Carr, and Burkhead. These six players cost more than the $50,000 salary limit so I am going to cut Carr considering the Patriots secondary is elite and plug in the next player which is Harry. 

    Example Lineup:

    PosPlayerSalary
    CPTJulian Edelman$13,800
    FLEXCam Newton$12,200
    FLEXSony Michel$3,600
    FLEXJosh Jacobs$10,600
    FLEXRex Burkhead$2,400
    FLEXN’Keal Harry$6,800
    Total:$49,400

     

    I am feeling more and more confident in these processes as we progress into the season. Check back next week and hopefully we will be able to enjoy some craft beer as we begin to cash more and more. Also, if you’re looking for other DFS help, check out some of the other resources we have here at the Undroppables including Paulie’s DFS Sleepers and Michael Reedy’s Money Makers and Heart Breakers.

     

    Resources:

    Captain Chart:DFS Captain Chart

    Percent on Optimal Lineup Chart:

    Stay in the Loop

    Get the daily email from CryptoNews that makes reading the news actually enjoyable. Join our mailing list to stay in the loop to stay informed, for free.

    Latest stories

    - Advertisement - spot_img

    You might also like...