Chasing The FanDuel Sunday Million
Chasing The FanDuel Sunday Million
*Skip to Week 3 Results if you read the introduction last week.
One of my goals for this NFL season was to find more time to do data analysis and turn that into creative content. I feel like I've been able to do that in a few ways this season:
Still, I wanted to do a regular piece that was a bit more fun and hopefully is able to portray how I go about using a lot of the information on the site.
I figured, hey why don't I try to win a $1million in the process? After all, I'm surrounded by million dollar winners and starting to feel a bit left out.
SportsGrid CEO Jeremy Stein, who is making lineups for the DR Sweat promotion, has done it TWICE.
I was live sweating with Drew when he pulled it off:
Each week I'll go through the process of how I made my lineups, and the macro (roster construction) and micro (specific player pool decisions) strategies employed. I'll then evaluate the results and how I likely got screwed by variance with no other explanation for my failings.
I'll document the results in here, although the goal isn't to grind out weekly profit in this format. The goal is to give myself the best chance to win life-changing money, something I feel got lost in my macro strategy this week (more on that later).
In general, I'll keep my entry fees in the $1,000 - $2,000 range. So, Week 1 that allowed me to max enter 150 times. In Week 2, the entry fee was $15, and I entered 100 lineups.
Here was the Week 1 recap: https://dailyroto.com/chasing-the-fanduel-sunday-million/
Now, onto Week 2.
Week 2 Results
Entry Fees: $9 x 150 entries = $1,350
Entries Cashed: 57 (38%)
Gross Winnings: $922
Best Entry: 715 / 380,405 ($50)
Net Profit: -$212 (-31.7%)
Lineup Optimizer Strategies
One of the challenges I had in putting together lineups this week was how to best utilize the optimizer. Last week, I talked about running stacks in the DailyRoto optimizer by QB. This allowed me to tailor each optimizer run to the QB or game I wanted to stack, so I was able to alter exposures and boost skill players specifically in that game. There are two main positives in running stacks this way:
- More organic stacking. Since you're boosting the relevant players to either pair with your QB or to come back on the opposing team, you can be a bit looser in terms of the stack constraints you use. Instead of telling the optimizer exactly what stack you want (“Give me 2 WR/TE paired with the QB, and 1 skill player coming back the other way”), you can simply run lineups with a minimum number of requirements, and let stacks happen organically. As a result, you can get a mix of stacks pairing your QB with 1 WR and some with 2. Similarly, if there's not a ton of value in the opposing team, perhaps you'll get some skill players on the opposition running it back, but not in every single lineup.
- Aside from having more diversity in the types of stacks you're allowing, which you'll still put some constraints on to ensure a base level of sanity (like using global team limits), it's easier to get exposure to peripheral players that are worthy of being in a small percentage of stacks but not as one-offs anywhere else. This the DeSean Jackson lesson that I learned in Week 1 and employed with success as I outlined in the Week 2 Sunday Million recap.
The drawback of running stacks by QB is twofold:
- It can be a little bit tougher to visualize and manage your exposure to players outside of the stack since you're running a fresh 10-20 lineups at a time rather than a full 150 lineups at once. We're currently working on some features that should make this easier moving forward.
- It can be time-consuming under the gun. It's not really a surprise to anyone that running a more customized 15-20 team stack X amount of times until you have your full 150 lineups is going to be more tedious than setting up rules, exposures, and boosts that apply to all 150 lineups at once, allowing you to submit a run just one time. This generally isn't an issue on Saturday nights, but if information is changing Sunday mornings, it's much easier to rely on settings geared towards running all lineups at once rather than having to re-do X amount of individual stack runs.
Initially, I was going to run stacks by QB again this week. However, on Saturday night with some uncertainty around Leonard Fournette's status and an inclination that we could see a meaningful shift in our ownership projections, I opted for flexibility.
The good news is that when Fournette was ruled out and our ownership projections were updated Saturday morning, I was able to quickly run 150 lineups that followed the QB exposures I wanted and the groups that I had set up for each QB that ensured stacking.
However, I think the drawbacks outweighed that flexibility. I'll talk more about Calvin Ridley below, but he fit the “DeSean Jackson” role this week. When I was running stacks by QBs, I was at least getting some Calvin Ridley on my Matt Ryan teams as a result of boosting all ATL skill players. Since I didn't want to boost him and use him as a one off when running a full 150 lineup set, I didn't get any in my Matt Ryan stacks when running lineups Sunday morning. That's of course a specific situation that looks worse in hindsight than foresight, but ultimately, I think moving forward that I prefer the balance between customizing settings for each stack with allowing specific stack types to manifest organically that is offered by running lineups specifically by QB.
I should also note that we're constantly improving our optimizer, and have a big update being pushed this week that should make however you choose to build your lineups more streamlined.
What Went Wrong
As noted above, I would have gotten at least some Calvin Ridley had I run stacks specifically by QB. That would have been a win for how I set up stacks, though, not on the evaluation of Calvin Ridley himself. Our projections did not like Calvin Ridley, and personally, Ridley has been one of the rookie WRs that I'm least interested in, due to a combination of prospect age and athleticism.
Ridley was our third highest projected value of the week among the Atlanta WRs. While a disappointing game from Julio Jones, left more statistics available for Ridley, Jones' projection was solid given his massive share of team air yards through two games and combination of historical efficiency and expected efficiency based on this year's opportunities. But what about the gap between Sanu and Ridley? That was the real problem and what most prevented me (and others) from getting Ridley in Ryan stacks, aside from manually boosting Ridley.
If you're taking the "hey we should have seen this coming" viewpoint, after getting out-targeted by Sanu 6 to 2 in Week 1, Ridley out-targeted Sanu 5 to 2 in Week 2. However, after three weeks, Ridley and Sanu have seen an identical amount of targets. Ironically, Sanu saw a season-high 7 targets in Week 3, while Ridley saw a season-high 8 targets:
It's also not as if the playing time for Ridley jumped a bunch in Week 2, which could have indicated more trust and responsibility for him moving forward. In fact, Ridley's Week 2 snap share remains his lowest after three games while his Week 1 snap share remains his highest:
Ultimately, this is an L. The market was somewhat on Ridley (8.8% ownership in the Sunday Million). We weren't. He broke the slate. Still, I have a tough time seeing a big mistake in the projection. Arguably, we should have evened up the volume between Ridley and Sanu, but even that wouldn't have pushed Ridley into "play" status. With the information at hand, I regret the way I built my teams that more or less disallowed Ridley (as covered above), but the Ridley projection itself seems fine, maybe even only a little bit low unless you severely played up the Julio Jones calf injury angle (was not on the injury report). There was no context in which I was coming close to the field on Ridley exposure this week, and you just have to live with these type of outlier results.
Speaking of outlier type results, the Bills beat the Vikings. By a lot. To say this was unexpected, is to put it mildly:
Not only did the Bills win, they completely dominated the Vikings leading to an absurd two rushing attempts from Latavius Murray in a game that he was not hurt.
Perhaps I'm feeling stubborn on the Monday morning after, but from a projection standpoint, I'm not doing any mea culpas on Murray. Murray was our third highest projected RB value on the week and 10th highest projected RB overall.
Initially, when the Cook news broke, my plan was to lock Murray in all of my 150 lineups. Since regretting not taking a bigger stand on James Conner in Week 1, I've been wanting to find a spot to get very aggressive since. The opportunity really hadn't presented itself. So, heading into Saturday night, when I make the bulk of my lineups, my intention was to lock Murray.
However, on Saturday night in our subscriber Slack chat, SportsGrid CEO Jeremy Stein estimated nearly 50% ownership on Murray. From my perspective, a sub-30% field ownership on Murray meant I wanted to go all in. While there were certain routes to him failing (which we saw realized), the combination of value in a vacuum along with the ease in roster construction alongside more expensive stacks made it worth it at that level. However, at 40-plus percent and for sure at 50%, it seemed like an unnecessary overweight. There was not a scenario in which I was fading Murray in low stakes MME this week, but had I felt those ownership levels would have been realized, I would have probably played an underweight position at around 30% exposure. While the tipping point on the decision from lock to not lock was somewhat feel based for me, I used our Range of Outcome Projections to simulate the RB position on FD. These are the probabilities I had for Murray finishing Top X at the position in value (so salary taken into consideration):
Ultimately, after that Saturday night conversation, I couldn't make up my mind on what the field ownership would be. So, I decided to allow for 100% Murray, but by utilizing the shuffle and MAX public ownership constraint in our optimizer, I knew I wouldn't quite get 100% Murray. Letting things happen more organically (using a 25-30% ownership projection on Murray to feed into the MAX public ownership constraint), I ended up with just over 60% Murray exposure. If you're unfamiliar with our MAX public ownership constraint, it uses our public ownership projections, sourced by Unsourced Fantasy Collective, to make sure your lineups don't exceed a certain combined public ownership. This helps to ensure GPP lineups aren't overly chalky.
Needless to say, I was disappointed when Murray's actual ownership came in at 42.9%. I was glad I didn't lock Murray, certainly, but I also knew from my sims that the odds of him finishing inside the Top 5 RB values on the week (32.3%) were meaningfully less than what the field owned him at. Like I said above, there wasn't a scenario I was fading Murray because of the upside at his price (and gut wise, I felt it was even higher than what our projections indicated), but had I known for certain ownership was coming in over 40%, I probably would have capped my exposure in the 25-30% range.
What Went Right
Overall, I felt like I did a good job of identifying high upside stacks and getting a piece of the most important QBs to own without spreading myself too thin. Before making my lineups, I outlined my stack counts in a Google Doc, and my actual exposures ended up almost exactly to this plan:
-Mahomes 20 (20)
-Jimmy G 20 (40)
-Matt Ryan 15 (55)
-Rodgers 20 (75)
-Wentz 15 (90)
-Goff 20 (110)
-Watson 20 (130)
-Cheapos (Bortles/Tannehill) 20 (150)
While I didn't have Brees or Newton exposure (ironically, on DK I expanded my list to include those two), 7 of the 9 QBs I had exposure to finished in the Top 10 on FD. The only misses were Wentz and Bortles, who I had less exposure to than most of my other stacks.
One of the spots I was most overweight on was a Rams stack, one of the contrarian stacks we outlined in the past week's premium content:
Given the concentration of this offense, I forced all of my Goff stacks to include two of Cooks, Woods, Kupp, and Gurley, using our grouping functionality:
Here were the results:
-Goff (2.7% ownership, 26.6 FD points)
-Woods (3.5%, 28.7)
-Gurley (18.1%, 22.1)
-Kupp (6.2%, 14.8)
-Cooks (5%, 12.5)
Woods was our top projected value of the Rams WRs.
All in all, another down week, although not surprising given the top heavy nature and large field GPP. We'll be back at it again next week.
Be the first to know
Want FREE Fantasy and Gaming Advice and Savings Delivered to your Inbox? Sign up for our Newsletter.