Here it is: the last weekend of NFL DFS. I originally thought that breaking down how I won a qualifier on a past two-game slate could serve useful to the readers. As I broke it down, I couldn’t help but think to myself, “I don’t see any useful information in this.”
Instead, what I thought I found was another reminder at how important it is to look at each slate independently. Before getting into what I mean by that, I have edited down the breakdown of the lineup and will include it here:
I will be scaling back a considerable amount – playing only a fraction of what I would play during the regular season, but playing nonetheless. I had success the last time I played a two-game slate, leading me to believe that looking back at that lineup could potentially give us all some direction on attacking this weekend. I took a look at the lineup in an attempt to gain actionable intelligence and what I originally found was nothing. I didn’t think this lineup, the plays, or way it was constructed served to help me in any way this weekend.
The contest I entered had just over 36,000 entries and the slate featured the New England Patriots at Houston Texans and New York Giants at Miami Dolphins. First place won a seat to the DraftKings FFWC.
The strategy that went into this lineup was fairly simple. I thought I had a decent idea on how the game flow was going to go in both games and I decided I wasn’t going to try to go super-contrarian or pivot off anyone that I thought would be a good play.
The first thing I did was decide on which defense I wanted, which has never been a starting point of a lineup for me in the past. Once I quickly settled on the New England Patriots, I decided that I would entirely fade the opposing team. Typically, that would be obvious, but with only four teams to pick from, I am not opposed to playing a skill player and the opposing defense. The only real fade on the Houston side was DeAndre Hopkins.
I knew I wanted to take a RB on New England to gain some access to the potential game flow correlation. I wish I could remember how I ended up settling on James White over LeGarrette Blount, but I can’t. I thought Gronk was a must play because of how soft the TE position was. Also, I thought having already rostered a New England running back would give me some separation from the other Gronk lineups.
This left me with the Giants-Dolphins game. I knew I wanted to go with a game stack that weighed more heavily on the Giants. This was an obvious stack and I thought a lot of people would be on it. I took Eli Manning at QB and stacked him with Odell Beckham Jr., Rueben Randle, and Will Tye.
This was going to be a popular offense to stack but I thought I could gain some distance by going with the “triple stack.” This is something I do at times – when I think an offense will be a popular one, I will try to differentiate myself by going all in on that offense. If I don’t want to come off of an offense that most lineups will have exposure to, I will try to gain the edge by taking on more exposure than the field. Also, I thought going with a second TE in the flex would gain even more separation from the other Gronk lineups.
I was left with a RB and WR slot to fill and I knew I wanted it to be from the Miami side of the ball. The decision was practically made for me to go with Jarvis Landry and Lamar Miller. Again, I went with a RB and pass catcher from the same team. The thought of absorbing negative correlation to gain separation wasn’t a thought in these two players, but it was a nice addition. I ended up leaving $1,200 worth of salary on the table.
Looking back, I made six small moves in the effort to be different while not coming off players I liked. You would think that I did enough to gain the separation needed to have a chance to win, but guess what – I tied for first. Another player had the exact same lineup as me. My roster construction played little into finishing at the top of the leaderboard, but rather it was the fact that I had the nuts at every position.
I had chased a seat into the FFWC all year and at many different stakes, so there is no denying that I was more than happy with this outcome. Having said that, I still don’t think I should put much weight into anything that went into this lineup construction.
You are probably wondering “Then why the hell did you write this?”
Aside from the obvious reason – that I wanted you all to see that I had won – I think it points to an interesting thought. That fact that this lineup construction was successful does not, in itself, mean that it should be repeated or become a strategy that is often employed.
I have mentioned in the past that I consider my “process” to be a fluid one. I do not change the way I examine players. Early in the week I look at salaries, Vegas, matchups, and a few other things to get an idea of the most obvious values. I spend a good deal of time listening to what the “group think” of the week might be. I do a lot of work off the Players Models page and rarely touch where I weigh certain data. I have developed a manner of using these tools that I think is good and repeatable. I don’t think there is any argument in how important that is when you want to be a profitable player.
The fluidity comes in when I am deciding how I want to attack the games that are available to me. This is something that I try to keep independent of other weeks. We just took a look back at a successful two-game slate and I think the only thing that was done that I could consider repeatable was that I looked at that week independently.
When making that lineup I quickly recognized two things. I wanted to play New England’s defense against the Texans and that the other game featured two defenses that performed poorly against the opposing team’s top-receiving option.
Spending 60 seconds in the Trends Tool allows for me to check and make sure that these thoughts are something I should take action on.
Playing New England’s defense, fading Houston, and attacking the shootout potential in the Dolphins and Giants game made more and more sense the further I looked into it. The lineup construction began to take care of itself at that point.
Much is made of “process over results.” If we have a solid process, we should not change what we do solely because we had a poor week. In the same light, I don’t think because I had a good week the last time I played a two-game slate that I should attack this one the same way.
Our process should be repeatable, data driven, and profitable, but the way we attack a slate should be done with a little more independence. We use the data and information that our process has given us to find the optimal roster construction for that given slate. Diving into past results and examining the thoughts that go into a lineup are incredibly useful, if used correctly. We just need to be careful and not allow ourselves to mimic the way we attacked a slate just because it was successful.
This weekend will pose more questions than Week 14 did, and I need to look at it independently. I have already begun looking into different trends in an attempt to find an edge on the field and point me in the direction of which way to attack roster construction. It is the last week left for us to have the fun that is NFL DFS research. Rather than point out some of my early findings for this week, I think I will let you get your hands dirty in the Trends tool – there is no reason I should be the one having all of the fun.
