How I Took 5th Place Out Of 73 In My First FanDuel NFL Contest

As a complete newbie to one-day fantasy football contests, I spent most of my time last week just getting used to the FanDuel interface, and getting ready to dive into this new world.

It was a super busy work week here at TR, so I had my hands full just learning what the various game types are, figuring out how to enter contests, how to pick my team, what tracking features FanDuel has, etc. I also did some Internet research to try to find some resources to help me learn about daily fantasy strategy.

Week 1 Recap

As I mentioned in my first post yesterday, I’m starting the year with a $250 bankroll, and am going to be tracking my results and the growth (or shrinkage) of my bankroll all season. My username on FanDuel is tr_david, if you want to track me down on the site.

Remember, I am not an expert at this; I have no prior experience playing daily fantasy football. So I’ll be pleasantly surprised if I am able to increase this initial bankroll by the end of the season. And I fully anticipate losing some money along the way, especially in the beginning.

In Week 1, I entered two very low stakes leagues just to get my feet wet.

Bankroll Entering Week 1$250.00
ContestEntrantsMy RankEntryWonNet
NFL 50/50 League10032$2.00$3.60+$1.60
NFL 50/50 League10084$2.00---$2.00
Total Net Winnings-$0.40
Bankroll After Week 1$249.60

Besides the two games I played on my own account, I was also in charge of picking the official TeamRankings lineup for the private tournament we hosted. That ended up being my best performance in Week 1 — I managed to get TR into 5th place out of 73 entrants!

So I thought it would be helpful to share my strategy below. Let me know what you think in the comments.

TR Tournament Strategy Breakdown

As a complete beginner, I decided I needed to use some kind of lineup optimizer tool in order to have a chance of doing well. But I wanted to give our entry some kind of a TR twist, so I decided to use our algorithmic betting picks to decide which teams to select players from. Or, more accurately, I used our betting picks to determine players and teams to exclude from my drafting.

In order to do that, I used the Rotowire Fanduel NFL Weekly Lineup Optimizer*. The tool uses their weekly NFL player projections to create optimized lineups based on the player salaries and scoring system at FanDuel.

[*Note: I’m not really endorsing this tool. I do like the interface, but I clearly haven’t used it long enough to figure out if the projections are any good.]

The key with this tool is that you can you exclude specific players and teams, or lock specific players into your lineup, and then get an optimized lineup based on the constraints you set. So I excluded two types of teams:

  • Teams from our playable (2-star and 3-star rated) Under picks listed on our NFL Over/Under picks page
  • Teams that we were making a 3-star point spread pick against on our NFL spread picks page

Just FYI, you do need a premium subscription to TR to see our current week betting picks on those pages, but you can also use the pick tables at the bottom of the page to see how our betting picks did in prior weeks. It is possible, though, that some of our Week 1 betting predictions may have changed between when I grabbed them for this research, and where they finally “froze” right before game time.

The Logic Behind My Theory

My reasoning here went something like this:

  1. I hadn’t yet researched how FanDuel sets player salaries. I assumed that FanDuel would try to set them in a way that would equalize the perceived value of using each player. This is similar to how sports books often set an opening line for a game, at a value that they hope will roughly balance the money bet on the two teams. [EDIT: After further review, I think this is wrong. See EddieSpaghetti’s comment below the post.]
  2. If FanDuel is good at setting player salaries (which I would assume they are, given they seem to be running a successful business), and the betting markets are good at setting spread and over/under lines, then it would stand to reason that both FanDuel and the betting markets hold similar opinions about how high scoring a given game or team will be.
  3. If our prediction for a game is different than the betting market consensus, the players playing in that game will have to either over-perform or under-perform relative to expectations in order for our betting pick to be correct.
  4. Over the past five years our 3-star NFL point spread picks and our 2-star-and-above NFL over/under picks have been profitable.
  5. Given all of the above, it seemed like a strong prediction from our model might be an indication that the players on the corresponding team(s) might be mis-priced by the market, and also by FanDuel.
  6. Ideally I would have wanted to shade the individual player projections in one or the other direction, and then optimize a lineup. But that wasn’t possible, so simply excluding players from teams that we projected to do poorly seemed like a reasonable, quick alternative.

So to recap, based on my reasoning above, here’s what I did:

  1. Went to the Rotowire NFL FanDuel Lineup Optimizer.
  2. Using the “Exclude Teams From The Player Pool” section (bottom right), I excluded:
    • Teams involved in games where we rated the Under as a 2-star pick or better**
    • Teams that were were making a 3-star pick against
  3. Hit the “Show Me The Optimal Lineup” button.
  4. Used the resulting lineup exactly as-is. You can view the actual lineup on FanDuel.

[**In retrospect I should have scrolled through the “Excluded Players List” (top right) and added the defenses from those Under games back into the player pool, because an Under pick ought to be a good sign for a defense. But I didn’t think of it at the time.]

Adjustments Prompted By My Approach

I also entered Rotowire’s default optimized lineup in a contest using my personal account. You can view that lineup here.

As you can see, the changes between that lineup and the one mentioned previously are:

  • At QB, sub Alex Smith out (we picked TEN +3 ATS), and Matt Ryan in
  • At TE, sub Cordarrelle Patterson out (we picked MIN-STL Under 43), and Zach Ertz in
  • At D, sub New York Jets out (we picked OAK-NYJ Under 40.5***), and Buffalo Bills in

[***See my previous ** note — probably shouldn’t have excluded the Jets. Oh well.]

The most important change here was subbing Matt Ryan in for Alex Smith. Ryan ended up scoring almost 19 more points than Smith, and that’s the main reason this lineup beat the default FanDuel lineup.

The Outcome & Future Thoughts

This strategy seemed to go pretty well in Week 1, but obviously that’s such a minuscule sample that it’s essentially meaningless. It’s also worth noting that our top betting picks had a very good Week 1. So while that may be some validation of my general theory here, it also means that if our betting picks have a bad week — or more precisely, when they have bad weeks, which they certainly will — this daily fantasy strategy may have a bad week as well.

At any rate, I’ll be curious to see how Week 2 goes. I’ve entered a few contests with a Thursday start date, and a plan to enter a few more that start Sunday. Next week, I’ll recap how those turned out, and what strategies I used.