👉 Turn on notifications in the “New Content” channel on OWS Discord!
New content will be added to the BB+ Scroll each week!
Subscribe to One Week Season in your podcast player!
I’ll use this space throughout August to provide deeper context on player blurbs that stand out to me. (Note: I’m away from my computer August 4/5, but I’ll be filling this up throughout the month. We’ll start here…)
Something that is often discussed in weekly DFS theory and strategy is the “story” your lineup tells. Basically, the idea is to think about what the recap of the day would have to be for each particular lineup to have been the “right” combination of players that gets you to a first-place finish on that slate. I like to do something similar when drafting Best Ball rosters, thinking about the “story” that the roster is telling about how the NFL season will play out with each pick that is made. As Hilow and I discussed on his pod a few weeks ago, these “if-then” statements can be extremely valuable thought exercises and help us see things in a different light than our competition.
At every selection in a draft, there are a variety of reasonable options available, and whatever choice you make also implies some things about the other players you passed on. Similar to price point or positional pivots on a regular DFS slate, we want to be aware of the scenario where your picks are “right.” Everyone understands team stacks, and most of the industry is focusing on late-season correlations and balancing exposures, but very few are actively trying to leverage the decisions made for a particular roster with their later-round picks by using these indirect correlations.
Contest: Best Ball Mania 3
Draft Date: July 27th
Round 1, 1.06: Stefon Diggs, WR, BUF – I could go on and on about why I love Diggs this year, but that’s not the important part. Any player with a first-round ADP has a ton to like about their season. The important part is understanding what this selection means for how I will attack the rest of the draft.
Round 2, 2.07: Aaron Jones, RB, GB – Jones was easily the top RB left on the board for me at this pick.
Round 3: Josh Allen, QB, BUF – When I selected Diggs at 1.06, I was effectively making a stance that Buffalo will continue as a top passing offense in the league and Diggs will return to his 2020 form with a monster, top-3 WR season. If that happens for Diggs, Allen is basically a lock to remain a top-3 QB as well. In many drafts, I like waiting until the 6th-10th round range to take my first QB, but in this “story,” locking up the odds on favorite for QB1 at pick 30 seemed prudent.
Round 4: Jaylen Waddle, WR, MIA – Waddle correlates with my first two picks during the playoffs, as Miami plays Buffalo in Week 15 and Green Bay in Week 16. Waddle also provides leverage off my decision to pass on Tyreek Hill in Round 2. It’s notable that I passed on Gabriel Davis at this pick. This was a calculated decision as I felt that if Diggs proves himself worthy of the 1.06 pick, then Davis likely will struggle to pay off fourth-round draft capital.
Round 5: George Kittle, TE, SF – This pick is where things really get fun. Taking Kittle here means I have taken care of both “onesie” positions through five rounds with elite options, which gives me great flexibility in my roster positions going forward. From a leverage perspective, Kittle helps me in a variety of ways based on picks I’d already made. The most obvious angle being that I passed on Deebo in the 2nd round and if Deebo is disappointing this year it likely would benefit Kittle. The second and less direct leverage this provides is based on the fact that with each of my first two picks I passed on Kelce and Andrews. I have already effectively made a bet that those two players will not have “had to have it” type seasons at the position, in which case, Kittle is as good of a bet as any to be the TE1 (or match the top options for a much cheaper price). Finally, one person in this league had already taken both Mark Andrews and Darren Waller – this means that having one of the few remaining “elite” TE options will be even more of an advantage in this league than normal.
Round 6: Amari Cooper, WR, CLE – I like Cooper as a high-ceiling WR on rosters where I have already selected at least two WRs before him. Taking Cooper here is a bet on Deshaun Watson being available for at least half of the year, which makes Watson a definite target later in the draft. If Watson were to miss the whole season, Cooper can still have a decent season. Additionally, so much of this roster is relying on a big season from Allen/Diggs so I can take on more risk/variance at my QB2 spot. Additionally, Watson will likely be used mostly on 3-QB rosters or 2-QB rosters with the other QB having a bye after Week 9 (a lot of rumors saying 6-8 game suspension is likely with CLE having a Week 9 bye). This will likely lead to the duo of Allen (week 7 bye) and Watson having low combinatorial ownership.
Round 7: Hunter Renfrow, WR, LV – Solid WR in what should be a good offense and he should have steady production with some potential for spike weeks. This fits well with my other WRs as Waddle and Cooper may be prone to down weeks due to their respective situations. I also already have some indirect leverage against Renfrow’s teammates as Davante Adams and Darren Waller have very similar ADPs to Diggs and Kittle at the same positions.
Round 8: Chase Edmonds, RB, MIA – Adding to the MIA bet, and the negative reports so far on Raheem Mostert’s knee. The market hasn’t adjusted much yet on a possible workhorse role for Edmonds, and I don’t want Tua to stack with Waddle due to already having Allen at QB and targeting Watson later, but taking Edmonds here gives me multiple options as correlations with my BUF and GB studs in Weeks 15 and 16.
Round 9: Devin Singletary, RB, BUF – Hilow had some great info about the use of a QB-RB-WR stack in winning Milly Maker lineups in a tweet he sent last week, and based on the board, I liked adding Singletary to this roster with that in mind.
Round 10: Garrett Wilson, WR, NYJ – Some picks are just about making the right pick. No one stood out as fitting a particular “story” here, which is okay. I added a high-upside player and rookie WRs usually have stronger second halves of the season.
Round 11: Isaiah Spiller, RB, LAC – Leverage off fading Ekeler in the first round. I also passed on Mike Williams in the third round so my story wasn’t that Ekeler fails and the WRs have monster years . . . it’s that Ekeler disappoints with the WRs also having seasons that don’t blow away expectations. That would likely mean Ekeler was injured or ceded a greater than expected workload to Spiller.
Round 12: Jameson Williams, WR, DET – Another bet on a talented player who I am hoping “pops” at the right time late in the year.
Round 13: Michael Carter, RB, NYJ – My RB5, Carter should have a passing down role with the potential for spike weeks from big plays or if Breece Hall misses any time.
Round 14: David Njoku, TE, CLE – High-upside and athletic TE, ADP fit at this pick (not a reach), and plan to stack him with Watson next round. He has the same bye week as Kittle which likely means low combinatorial ownership.
Round 15: Deshaun Watson, QB, CLE – Brings together the CLE stack with Cooper/Njoku, finishes off QB position (only needed 2 roster spots).
Round 16: Parris Campbell, WR, IND – A likely starter in Round 16 who also provides a bit of leverage on my decision to pass on Michael Pittman in the third round. My intent with this pick was to take Isaiah McKenzie but he was sniped right before my pick.
Round 17: Will Fuller, WR, FA – Julio Jones just signed with the Bucs, and his ADP shot up (as expected). My feeling is that Fuller follows suit very soon so this was one of my last chances to get him on the cheap.
Round 18: Jeff Wilson, RB, SF – I would usually go with a WR9 here rather than taking my RB6, but for this roster, I have already made a sizable bet on George Kittle having a healthy and productive season. Kittle is also a key to the SF run game, so if he has that kind of season, then having a piece of the SF backfield makes a lot of sense here. Wilson likely won’t be a regular contributor, but a three to five week stretch of heavy usage at the right time could be a leverage spot that puts me over the top.
THE STORY: Here are the basics of the “story” I need in order for this roster to carry me to the promised land.
FINAL THOUGHTS: I loved how this roster came together and while having *all* of these things happen may be a long shot, none of them are outrageous reaches or outside a reasonable range of outcomes. First place for BBM3 is $2 million and this entry cost $25. I will gladly take 80,000 to 1 odds on a parlay of the above situations happening. I hope you are able to take some things away from an insight into my thought process and I’d love to hear feedback on Twitter or Discord with any thoughts.
Good luck OWS fam, see you in the lobby!!
Welcome to The Oracle! :: The Greatest “Cheat Sheet” In Best Ball!
Each week in The Oracle, OWS team members will take on the key strategy questions from that week’s slate :: sharing their thoughts on how they plan to approach these critical elements from a roster-construction, game theory, and leverage perspective.
Training camps are in full swing and along with that comes big news and sweeping reactions. Among all the “noise” from this first week of NFL teams being back at practice, what is the one piece of information you have found the most actionable? Also, what is the one thing you think people are overreacting the most about?
Treylon Burks camp highlights. I kid, but seriously, this dude is the truth. In all seriousness, the most actionable pieces of early camp news has to be player health and level of participation. The pieces that come to the front of my mind are Hollywood Brown and a soft tissue hamstring injury that has prevented him from seeing any reps in a new scheme, the Baltimore running backs, Travis Etienne looking fully recovered and explosive, and Michael Thomas running routes at full speed (albeit with a slight change in gait).
To immediately make light of the fact I was completely joking about Treylon Burks camp highlight videos, my answer to this part of the question is exactly that – camp highlight videos. The level that the industry and field overreacts to mundane player camp videos is hilarious to me every single year. Most teams have only progressed through individual drills and 7-on-7s. No teams can even practice with full pads yet. Chill on the hype until we at least get to pads, OK? Cool.
The Allen Robinson steam seems very real to me. He is a player I have already taken very often in drafts and someone whose talent I have been high on since he entered the league. Now playing with the best QB of his career, Robinson could really take off. The impact this has on the rest of the team will be interesting to see – will his ascendence cut into the insane usage Cooper Kupp saw last year? Will having two of the top receivers in the league raise Matthew Stafford’s efficiency even higher than we saw in his first season with the Rams?
Miles Sanders “running with the second team” is not something I am concerned about. According to reports, he took four snaps with the second team last week in practice and now people are running with that “story” as a knock against Sanders. Until/unless we see Sanders taking snaps in a preseason game with backups, there’s nothing to see here.
Damnit I wanted to write about Allen Robinson but Mike beat me to it, so I’m just going to +1 that.
I’ll be totally honest here: not much. Camps have barely started and as Hilow notes we haven’t even seen teams practicing in pads yet. The single biggest takeaway from this question, to me, is “don’t overreact to super early camp news because most of it is likely totally irrelevant.”
That said…..I do like that Rhamondre Stevenson appears to be playing a lot with the first team in the early going. That’s one I’m going to keep an eye on. He showed plenty of talent last year when given the opportunity.
And, broadly, I like to watch the rookies in the early going. Guys like Treylon Burks, Drake London, and (especially) Skyy Moore are guys where, if they step into full-time roles early on in the season, they have the talent to smash their current ADPs, and I want to be in front of that if we start seeing signs in camp that they’re both A) ready for the NFL and B) going to be heavily utilized by their teams.
Most of my Best Ball contributions are going to start coming in the second week of August (as things get a little less crazy with the “work” side of OWS), but I wanted to drop into The Oracle (without reading anyone else’s responses) to explore some of these questions.
When I first started playing Best Ball in 2019 (which was the first year I got as deep into training camp as I now do), I was part of the overreacting crowd, and I would find myself chasing the rises and dips along with everyone else. I can’t tell you how much Darwin Thompson hype I chased…after I was already on track to be double-the-field from scooping him in the 18th round when no one else knew who he was. And realistically, that’s one side of things: news generally requires some sort of saturation point before it hits Rotoworld (and similar outlets) and starts sticking to the draft board. Sometimes, I’ll read positive reports on a player for weeks before a Rotoworld blurb is worded in just the right way that it turns up the public heat on that player. Other times, it’ll be a couple days. This is (frankly) the value of having some sort of Best Ball subscription, of course, if you’re not able to follow the news at the training camp level: you can stay ahead of the field on some of these player rumblings. But more broadly, it’s a reminder that you can’t get too caught up in the hype. If you think like an investor, you buy on players who are sliding due to uncertainty, and you sell on players who are moving up because of a report that was maybe worded a particular way by summary-sites, but that was really more of a passing observation made by a beat writer. You have to be discerning (you can’t just blindly move opposite the crowd), but more than anything, the key is to think deeply on the “why” for the movement.
Some recent examples ::
Diontae Johnson has started sliding because he isn’t practicing as he holds in. Now that we’ve defined why he’s sliding, you can ask yourself if you think he’s at risk of falling so far behind that he fails to perform if/when he returns. There’s no right or wrong answer, either, because you’ll have to make so many of these decisions, and you’ll be right sometimes, and you’ll be wrong sometimes. (So okay, there are wrong answers, when it all plays out. But on this side of the NFL season, a lot of these are close to 50/50 no matter how convinced one person might be on their stance. The deeper point is: know what question to ask. Know why the player is rising or sliding. Then you can make a more well-rounded decision on which direction you want to move.)
Others are less nebulous than the Diontae example. For instance: Miles Sanders slid for a few days because “summary sites” reported he was running with the twos at camp. That was the headline. It was less black-and-white than this in the actual reports (the Eagles always rotate backs at practice, especially in training camp; the reporter had just felt it was noteworthy that Sanders had seemed to spend significant time there, while clarifying that it could be nothing), and this alone allowed me to feel comfortable buying the slide on a player I’d spent very little draft capital on to that point. Further reports from the team shrugged it off as nothing, while Nick Sirianni later clarified that Sanders is their lead back. Within another day or two, these reports hit their public saturation point, and the slide became less persistent.
While I could go on and on, I think the bigger picture is that these things happen day in and day out. Some last a long time, and some last very little time. Even if you’re not able to pay attention beyond reading Rotoworld and OWS, get in the habit of asking why about player movement. What is the public perception of the player, more broadly? What was the headline, and what did the actual report say? Or like the Diontae example above, what is your opinion? Or think about Ja’Marr Chase last year. He was dropping passes in camp. That was the headline. But what was your opinion? Everyone had touted Chase as the best wide receiver prospect since Calvin Johnson. He hadn’t had a problem with drops in college. Was that going to persist as an issue deep into the season? Apply your own critical thinking, and assess the potential risk/reward of buying or selling a slide or a rise. What is the headline? What is the actual news?
A lot can change in a hurry in the NFL, and training camp is no exception. Give us two or three RBs or WRs who are currently going undrafted or taken in the last round that you think could jump several rounds by Labor Day:
Last year I wrote about projection fragility as it applies to DFS, and this year I’m taking it a step further to explore it in the best ball format. The core principle is the same, in those projections (which, under the hood, power season-long rankings) are for median values, but in tournaments (whether best ball or DFS), we don’t care about median outcomes. We need ceiling outcomes to win. For the non-mathematically inclined, a median value within a range of outcomes means that half the time, the actual outcome should be below the median, and half the time, it should be above the median. Here are examples of two types of distributions of values that can both have exactly the same median:
Which of those would you want in a tournament? Well, the answer isn’t that simple. You don’t ALWAYS want one or the other type of distribution because it depends on the rest of your roster. The important thing to know is that the barbell distribution has a broader range of actual outcomes. In a normal distribution, most outcomes are clustered near the median, whereas in a barbell, most of the outcomes are fairly far from the median. So if you’re seeking to add more variance to a roster, you’d want a player with a barbell distribution, whereas if you feel like your roster already has a lot of variance and you want to add some stability, you’d want a player with a normal distribution.
Welcome to the One Week Season 2022 NFL Best Ball “State of the Union.” This article is a view of the landscape from a big picture perspective of the format, tournaments, and industry at this time. The outlook and landscape of this area of fantasy football will change year to year, and many of the things are theoretical or opinion based, but it is an important exercise to think about from a big picture perspective to understand the game you are playing. Here are some key things to keep in mind as you venture into the Best Ball waters heading into the 2022 NFL season.
You don’t actually have to beat all of the people in your contest, you just need to win each of the smaller contests on the way to a championship. Just like the NCAA tournament, there is a lot of luck involved in your “draw.” It is possible, and possibly even likely, that two people who draft the *exact same team* of 18 or 20 players would have wildly different results. It would not be surprising for the entry that wins BBM3 or the DK Milly to have a duplicate roster that doesn’t even make it out of the first round or a subsequent playoff round. Last year, I had teams that finished 4th place in their Round 1 (Weeks 1 to 14) leagues that actually outscored teams I had that took 1st place in their Round 1 league. This continued in the later rounds as well, as I had a team in the big DraftKings $6 tournament that advanced to the finals with 175 points in Week 16 but another team that was eliminated in Week 16 that scored over 200 points. When you think about it from a “draw perspective,” you probably don’t want the best overall team for the first 16 weeks if you want to win in Week 17. Just based on the laws of regression and what we consistently see during NFL seasons, the recipe for success is likely a team that just “survives” the early rounds and then smashes in the finals. Obviously, so much of that is out of our control with the “draw,” which makes Best Ball more exciting and scarier at the same time. All we can do is use good process and game theory to give ourselves the best chance of winning and taking advantage of our “good draws.” Luckily, we’ve got some great articles and content to help you out with this!
The Best Ball Mania 3 on Underdog Fantasy and the Draftkings Milly Maker are by far the highest payout contests and therefore receive the most attention and discussion from content providers. The final rounds of these contests will feature 470 and 969 lineups, respectively. Something to strongly consider is that the “optimal” strategy for contests with fewer than 200 finalists will be very different than the optimal approach towards the big contests that are being touted and discussed across the industry currently. Most of our competition will likely not make this distinction and build differently as a result, meaning they will be building suboptimally for these contests. This is a huge opportunity to gain an edge as our expected advancement rates should be very similar across most tournaments for each of the first three rounds (regular season and first two playoff weeks) but our odds of taking down a tournament in Week 17 are much higher in some of the “non-headliner” tournaments. Examples of these tournaments:
There is a huge crossover from the NFL DFS community/industry to NFL Best Ball. The DFS industry is composed of so many mathematical minds and data-driven people. Those tendencies are naturally leading to a lot of dissecting the results of the only two years of large-scale Best Ball tournaments that we have as a sample size. This results in such a strong inclination to “follow the leader” and attempt to use strategies that worked in the last two seasons that it forms opportunities for us to exploit edges. The current landscape is very similar to the early years of DFS when everyone was just figuring things out (something Xandamere has also pointed out recently in our podcast series).
To sum it up, we are still early in the game, and with each year and more data, we should gain knowledge about what is optimal. A few years from now, when we look back at this point in the Best Ball market’s growth, I am not sure if my thoughts and philosophies from this year will look right or wrong, but I am almost certain that what most of my competitors are doing will be wrong – which is something we can and should be trying to exploit.
From a purely mathematical standpoint, we can have a decent idea of what we are actually risking and/or likely to see as a return. For BBM3:
These numbers are not a guarantee for anyone’s return and I am not saying you can’t make money from Best Ball. However, I do think these numbers can give you a good baseline expectation for Best Ball and help you make decisions for your bankroll and how you will attack these tournaments. Due to the long season and nature of the contests, it is reasonable to think of these baseline expectations as somewhat predictive.
(Note: What I mean by this is that for a regular season NFL Sunday main slate it would be much easier to completely whiff on having any teams “cash” if you max entered, but there should be a reasonable baseline performance expectation of teams advancing for anyone playing Best Ball who is at least somewhat competent. From there, things like player rankings/selection, draft strategies, roster construction, and game theory can then raise or lower those expectations for each individual depending on how sharp they are.)
Obviously, there is a ton of variance at play and it’s unlikely anyone will actually see a return exactly as outlined above, but if we know the “average” return would be about -70%, it becomes much easier to accept and seek the variance needed to actually win these tournaments and see positive ROI. A common issue for many people who play DFS is playing too safe as they get too focused on “cashing” for specific lineups. When some of those lineups min-cash they usually get close to 2x their money and are “rewarded” somewhat immediately, which reinforces the suboptimal plays and allows people to slowly bleed off money. The difference for Best Ball is that a lower percentage of teams are getting paid, you are waiting 5+ months to get paid, and the return for the “min-cash” type of teams is lower. For me, the takeaway here is that just staying with the herd and playing how everyone else plays is likely to keep you around that baseline expectation and just hoping to get lucky. If we stray from “popular” thought and are wrong, we are really only “being risky” with the roughly 30% of our investment that we would expect as a baseline return. Meanwhile, the upside from there for being right is immense. The risk-reward calculation is very clearly in favor of chasing game theory and leveraging the field. Luckily, we’ve got you covered there!
Most of the focus around Best Ball is on tournaments (much like DFS!) because they’re exciting and offer the chance to win life-changing money. But there are also smaller Best Ball contests available across all major sites ranging from 3 to 12 entries. These contests aren’t exactly “cash games” as we’re used to thinking about them in DFS, in that they don’t pay out the top half of the field (though I think it would be awesome if some site decided to implement that), but they’re fairly close; some of them pay out winners each week (like Yahoo), whereas others just pay out the top few teams (generally the top 20-30%) at the end of the season.
I wrote about this format last year and noted that I had an ROI of over 50% in “cash game” best ball contests. I’m not sure how long that’s sustainable, but I managed it again last season, so my sample size of 2 years feels pretty decent here. Will these leagues always be so soft? Probably not. With every new format of DFS, best ball, etc., we tend to see the games start off very soft as people learn how to play them well, and then those contests sharpen up over time. We’ve seen this pattern play out over the years with DFS as a whole, then with smaller formats like DK’s Tiers and Showdown, we’re seeing it right now with prop bets, and we’re seeing it with best ball. Can you make life-changing money playing these types of best ball leagues? Unlikely (unless you play a TON or unless your definition of life-changing is quite different than mine), but with a little preparation and a little practice, you can make best ball cash games a solid part of your overall DFS/sports portfolio and feel confident that you’ll make some decent money at it. For full transparency, my ROIs resulted in net profits of something around $500 for each of the last two seasons. I’m not buying a private island or anything, but for effectively no actual work, I’ll take the free money and use it to pay for some tournament buy-ins. As we like to harp on around here, every edge matters.
Let’s first consider what makes these leagues soft. First, the format is still relatively new. And, given that the sample size is an entire football season, we just don’t have good data around what’s successful. This means people are essentially flying blind – it’s like the beginning days of DFS in which people were just focused on figuring out who the good individual plays are and not thinking at all about how those plays come together to form an actual roster. The second reason these leagues are soft is that what little best ball content exists is all focused on the big tournaments. People go into these leagues building things like hyper fragile or zero-RB builds that might be appropriate to a massive tournament but which take on too much variance for a cash-type format. So, people don’t have much of their own data or experience to help them figure the format out, and the existing content isn’t going to help them.
The important thing to understand is that best ball tournament builds are built around a series of assumptions. If you draft Christian McCaffrey in the first round, you’re assuming that he’s going to be healthy and productive all year, and so you aren’t likely to invest a ton of additional draft capital in the RB position (nor would you really consider drafting CMC’s backup), because if he gets hurt you almost certainly aren’t in a position to take down 1st place. Best ball tournaments are about leaning into and smartly leveraging variance.
Cash game leagues are the opposite: they’re about minimizing variance in order to increase your odds of a profitable finish. You don’t need to beat tens of thousands of other users. You only need to beat 9 or 11 (depending on site/format).
Here’s the meat of things – a list of guidelines you can use as a framework which, along with a solid list of rankings (our friend Hilow has put those together for OWS right here), should set you up for a profitable season of cash leagues.
Does this seem pretty simple? It should. These leagues are soft and armed with just a solid list of rankings as well as the principles laid out above. You should be able to crush them. You’re welcome for paying for your OWS subscription and some extra pocket money for the season. 🙂
$10,000,000. $3,500,000. $1,000,000. These are the prize pools of the largest Best Ball contests on the three largest Best Ball platforms – Underdog, DraftKings, and Drafters, respectively. Even FFPC, or Fantasy Football Player’s Championship, is running a large-field Best Ball tournament for the second consecutive year, which has a prize pool of almost $1.1 million. The point is Best Ball is currently the fastest-growing fantasy football format and there is a lot of money to be made this season. But that’s where a lot of people will stop when talking about Best Ball – “See big prizes, take stabs.”
Each format has its own scoring settings, advancement rates, playoff weeks, and so on, which we must understand as we formulate a plan of attack. The rest of this article is going to take a methodical and statistical approach to define a winning strategy, focused on the two largest contest offerings – the Best Ball Mania III (BBM3), found on Underdog, and the Best Ball Millionaire, found on DraftKings. We’ll start by taking some time to digest the scoring and positional intricacies of each format, continue with a look at “the numbers” for each of those contests, develop a theoretically optimal plan, and explain the process behind our findings. The ultimate goal of this piece is to ensure that readers leave with a better understanding of how to optimize a portfolio of rosters across the two largest Best Ball contests available, and I want to preface it with a statement:
What we are going to be discussing here today is 100% theory. Best Ball is in its infancy, and it is about as far from solved as can be. What we are trying to do is leverage field tendencies based on observations and statistics in order to place ourselves in the most optimal position to win the most money when we get things right!
Let’s start with the basics. Both contests require you to “start” (in the Best Ball format, the computer automates your lineup based on the optimal combination of players in a given week) one quarterback (QB), two running backs (RB), three wide receivers (WR), one tight end (TE), and one flex (FLEX). That means that each week, your “starting roster” consists of eight players, optimized from your entire draft. Drafts on Underdog consist of 18-rounds whereas drafts on DraftKings consist of 20-rounds, an important distinction which we will get to shortly as it will influence our decision-making processes between platforms. This means that your “bench” consists of only 10 players on Underdog, while on DraftKings you can carry an additional two players.
The first round of each contest is a 14-week mini-league consisting of the 12 people that draft together. The top two teams from each draft, or mini-league, based on cumulative points earned over the first 14 weeks, will move on to the quarterfinals round in Week 15. We start to see some differences between the two contests once you reach Week 15 – on Underdog, one team from each new 10-person league will advance to the semifinals in Week 16, while on DraftKings, one team from each new 12-person league will advance. Once you reach Week 16, teams will be placed in leagues of 16 on Underdog and leagues of 12 on DraftKings, during which only one team will advance to the finals from each site. The finals played during Week 17 of the NFL season, consist of a 470-team, GPP-style contest on Underdog and a 969-team, GPP-style contest on DraftKings. Again, we’ll break down the importance of these advancement rates and the size of the finals pool further below.
The major distinctions in scoring settings between the two sites are as follows:
-DraftKings: Full point per reception (PPR), Underdog: Half point per reception (Half PPR).
-DraftKings: Bonus points (3) for 300-yards passing, 100-yards rushing, or 100-yards receiving
-Interceptions and fumbles lost are both minus one (-1) point on DraftKings while fumbles lost are minus two (-2) points on Underdog.
Finally, the buy-in on Underdog is $25 and the buy-in on DraftKings is $5, with a total of 451,200 and 837,200 entries on each platform, respectively. The first-place prize on Underdog is a whopping $2 million while the winner of the DraftKings contest will walk away with a cool $1 million. The barrier to entry values (cost of max entering each contest for 150 entries) and different first place prize money values will come into play shortly as well.
With the basic understanding of how each contest is run out of the way, we’ll now take a look at some important statistical differences between the contests themselves. I’ll break up the modeling into each unique platform so we can start to see deviations in our optimal approach.
From a purely statistical sense, without measuring skill or variance, the chances of advancing a given team to the finals in Week 17 on Underdog are approximately 0.104%, or (2/12)*(1/10)*(1/16). Once you reach the finals, the chances of winning the contest, for a grand prize of $2,000,000, are approximately 0.213%, or 1/470. The “min cash,” or minimum amount won, for making the finals on Underdog is $1,000.
From the same purely statistical sense, again without measuring skill or variance, the chances of advancing a given team to the finals in Week 17 on DraftKings are approximately 0.116%, or (2/12)*(1/12)*(1/12), which is slightly better than on Underdog. But once you reach the finals on DraftKings, the chances of winning the contest, for a grand prize of $1,000,000, are approximately 0.103%, or 1/969. The “min cash” value for making Week 17 on DraftKings is $250.
Two years ago, we saw a hyper-fragile team built by Justin Herzig (@JustinHerzig on Twitter) ship the BBM on Underdog. What happened the following year? According to Hayden Winks (@HaydenWinks on Twitter), hyper-fragile builds (rosters with four running backs and additional emphasis on the other three positions) increased almost 3x the previous year (just over 6% of total rosters in play in 2020 to just under 17% in 2021). Last year, Liam Murphy (@ChessLiam on Twitter) famously paced the field in the BBM2 by placing additional emphasis on Week 17, eventually shipping the contest with a gnarly Rashaad Penny/Amon-Ra St. Brown mini-Week 17 correlation and Cincinnati stack. That has translated to the field now being privy to the fact that Week 17 is where all the money is made, which means we should be paying additional attention to it. But the field seems to be clumping together in a very narrow grouping in Week 17, correlated emphasis, giving us solid leverage potential having identified this as the case.
The remainder of this article will dive into field tendencies and leverage opportunities, paying attention solely to Week 17 (where all the money is made in these massive Best Ball contests). These methodologies are intended to be combined with your own, with what you have learned about roster construction, variance, leverage, and other content from this year’s Best Ball product at OWS. It is essential to understand that so much more goes into roster construction than simply focusing on Week 17, but that is precisely what we’re going to do in this piece – concentrate only on Week 17.
The glaring issue with how the field seems to be interpreting this newfound strategy is that we’re now basically seeing a split field between more recreational players that are looking to leverage isolated snapshots in ADP (which is neither good nor bad; it just is – for more information on dynamic ADP and how to best leverage it in varying draft windows, I highly recommend you check out the first installment of our Best Ball podcast series) and “the sharps” who are attacking Week 17 relentlessly through team stacks, game stacks, and correlation.
I will not go into why Week 17 is so important in this piece as it should be widely understood through the rest of the data, interpretations, and teachings throughout the site this year, but we will dive headfirst into some of the weaknesses from the field, with particular attention to Week 17 and general field trends. What we’re seeing this year is an overemphasis on Week 17 and primary correlations. You see it all over Twitter, you see it in your draft lobbies, and you see it with the content that is being put out around the industry. Week 17 or die! And while that isn’t necessarily wrong (it’s actually super right), the general interpretation is to over-stack potential game environments in Week 17. Examples of this include QB/WR/WR/bring back, QB/WR/TE/bring back, and QB/WR/2 x bring backs. But what is the historical precedent for over-stacking to be optimal? If we look at the Milly Maker winners from 2021 (18 weeks, with more than 400,000 entrants in each contest), over-stacked rosters with a quarterback, two pass-catchers, plus a bring-back won only four of the 18 total weeks. A mega-stack with a quarterback, two pass-catchers, the team’s running back, and a bring-back also shipped once. That means that five of the 18 total full slates were won with an over-stacked roster.
Put another way – 13 of the 18 weeks were won without over-stacked rosters, which is essential information considering the field is mainly overemphasizing over-stacked best ball rosters in an attempt to leverage the field for Week 17 where all the money is made. Five of the 18 weeks were won with a QB/pass-catcher/no bring back roster, four of the weeks were won with a QB/pass-catcher/bring back roster, and four of the weeks were won with a QB/RB/pass-catcher/no bring back roster. Since the field is placing heavy emphasis on both over-stacked rosters and team stacks not including the team’s running back, a situation exists where we can build smartly for Week 17 without making suboptimal roster construction errors by simply including a team’s running back (or backup running back – more on this below) in your team stacks.
Of the 18 Millionaire Maker winners on DraftKings last season, four utilized a quarterback, running back, and pass-catcher stack with no bring back. Yet only a tiny percentage of the Best Ball field is building their rosters in that way. Now, that may or may not be optimal, considering you first have to make it through Weeks 15 and 16 before we’re looking at these contests through the lens of an isolated slate, but there is a massive amount of leverage to be gained from building teams in ways the field is not. Basically, don’t feel compelled to force a Week 17 bring back on your primary stacks, and don’t be dissuaded from including a team’s running back on your primary stacks because the data shows there are multiple ways to beat large field DFS contests, which is what Week 17 boils down to.
We’re now in Best Ball summer, and there’s money to be won everywhere. The Best Ball prize pools from the major providers doubled from 2020 to 2021. It’s already more than doubled again from 2021 to 2022. Best Ball is booming.
Underdog is the most popular place to play, and for good reason. The user experience is second-to-none, the mobile app works great, the tournaments are gigantic and fast-filling, and they track much of your data for you. Plus, they make their data publicly available for free allowing for thorough analysis. But, because they’re offering so much money and providing such a great platform, the competition is stiff.
We love DraftKings. The vast majority of DFS content centers on DraftKings. But their Best Ball product sucks.
They don’t provide the user with any of their data, so you’re on your own for exposures and there’s no way to upload your own rankings. If you want to adjust the rankings, you literally have to drag players manually up and down a list into the correct positions and expect that to take an hour each time you make any sizable changes. There’s no draft board, so good luck keeping track of what your opponents are doing while drafting. The roster page offers far less information than on Underdog. And the mobile app is buggy and crashes if you switch between apps too often. Plus, the only data available is your personal contest results.
But the prize pools are mouth-watering, and the competition’s mouth-breathing.
Because of these issues, there’s a profound lack of Best Ball content focused on the DraftKings platform. There’s simply a lack of thought behind how to succeed in these tournaments. Luckily, I qualified for the finals of last season’s $300K Play Action tournament ($3 entry fee, max 20 entries, $50K to first), so I had at my disposal the raw data from that contest’s Week 17 championship. This data does not include anything about when a user’s team was drafted, or what pick they got each player. It’s simply their roster and the points scored in Week 17. But, by contrasting this data with data from a similar tournament on Underdog, I was able to uncover some important takeaways to help you win against a softer field of competition.
Even with less savvy opponents on DraftKings, you cannot expect that you’ll be able to draft like you’re playing on Underdog and see the same level of success. I’ve seen plenty of otherwise smart players making basic mistakes in the DraftKings tournaments because they’re too comfortable with Underdog and not keeping in mind the platform differences in scoring system and playoff structure. Here are the main takeaways from this analysis:
Hey, Guess what? The Chiefs Play the Broncos in Week 17!
If you’ve tuned into any media concerning our big Best Ball tournaments, you’ve encountered a bombardment of discussion revolving around the huge importance of Week 17 matchups. By that time, the field will have been whittled down to only a few hundred, essentially turning into a mid-sized GPP tournament. On Underdog, 470th place will earn $1,000 while 1st place will rake in a cool $2,000,000. Quite a difference. Week 17 has been hyped so much that some voices in our business are beginning to pivot and rage against the very concept of its utmost importance. Do not listen to these people. I don’t care that there’s only a .1% chance of advancing a team to the final round… that’s where the money is!
Just promise me you won’t sacrifice too much value because you reached 14 spots to take a bring-back opponent from your 3rd quarterback’s Week 17 matchup.
You do want to structure your rosters to maximize your players’ ceiling weeks while minimizing the damage when they inevitably experience a floor.
Correlation in Best Ball rosters is obvious when you think about stacking. But under the surface lies an edge in building rosters with positional players that complement each other from a floor/ceiling standpoint. You maximize your players’ ceiling weeks while minimizing the damage when they inevitably experience a floor game.
There’s a real-life example in my household. Mrs. Sonic and I function well together emotionally because I have massive highs and lows, and she is solid as a rock, maintaining a level head despite her surroundings. My volatility leads to plenty of excitement…while she keeps our lives from going completely off the freaking rails.
I’ve also done 229 Best Ball drafts already this season, and considering “divorce lawyer” hasn’t appeared in her internet history yet, I think I’ve married well.
Many assume that players with a boom or bust reputation are perfect for Best Ball, and there’s some truth to that. They’re certainly better for this format than managed leagues where you can completely screw yourself when you roster a dude like Marquez Valdez-Scandling, and he gives you a 1/7/0 line, only to bench him the following week when he catches multiple bombs.
How do you win at Best Ball 2022? No, seriously can you tell me? The prize pools on Underdog and DraftKings are beyond massive, the entry fees are super low, and it just feels like both sites want to take our money. But how can we compete with a half million or more entries without just getting flat-out lucky? Well, I hate to be the one who says it but I think we’re all going to continue to be searching for the answer in cracking the Best Ball code into next season.
What we can do, as many wise adults will say, is focus on controlling what we can control: our own behaviors. Best Ball is such a nascent industry and format that we really don’t have enough signaling data to derive advice from. Sure, we can take the winners from the mega tournaments last year and glean some insights (more on that below) but it’s still a small sample. And with so much riding on how we conduct the draft (literally everything), even the players who fall to us or don’t fall to us only represent just a few variables in this equation. We have to try to consider how the other managers are drafting, which is impossible to predict. We have to try to consider which players may get injured, which is impossible to predict. We also have to try to consider the best playoff matchups in Weeks 15-17. It is difficult to project which teams will be in the playoff hunt and which will not. Difficult, not impossible . . . hey, that’s progress but drafting Best Ball rosters is still as uncomfortable as can be.
Maybe I can offer you some hope here. Don’t take this too seriously. Keep your bankroll in check. And get as frisky as you would if you were bottling up 17 weeks of NFL DFS energy into one Best Ball draft window. If we’re going down this season, we’ll go down swinging. Let’s talk about some methods of inspiration to take us down the paths less traveled.
Hello OWS family, and welcome to the 2022 season! It feels like last year just ended, but Week 1 prices are out, Best Ball drafts are in full swing, and training camp has begun! This article is going to touch on several Best Ball strategies that are all viable depending on how your draft progresses. The most important concept to remember is that you shouldn’t go into a draft with the idea of doing one specific strategy. You should let things unfold and pick the best strategy for that draft.
Preferred Roster Construction
Hero RB gets its name from the idea that you select one RB in the 1st or 2nd round (your hero) and then ignore the position for at least the next five picks. It’s not uncommon in a hero RB build to wait until the 9th round before selecting your second RB. Be brave. The idea of the strategy is that by waiting on RB, you can lock up elite players at the “onesie” positions while adding your favorite high upside WRs in the early middle rounds. This strategy also leverages the idea of avoiding the “RB dead zone,” roughly thought of as rounds 3-7. Below is an example of what a Hero RB team might look like:
This team used its 7th/8th round picks on RBs (plus Hines falling). It made me feel uncomfortable drafting five RBs. When I use a Hero RB strategy and draft my second RB later, I’ll typically roster six RBs. This squad also lacks an elite TE, which made me want to use an extra roster spot at the position. You can see the benefits of Hero RB as this team was able to come away with a fearsome foursome at WR of Evans/Brown/Robinson/Cooks, while not looking entirely unstable at RB.
Preferred Roster Construction
Superhero RB (these names sure are creative) is the same thing as Hero RB, doubled. Instead of selecting one RB in the first two rounds, you aim to use both your first two picks on RBs and then lay off the position until at least round 9. You want to leverage avoiding RBs in rounds 3-7 while packing your bags full of high upside WRs. The goal of Superhero RB is to be less vulnerable to your top RB getting injured, while also hoping that you manage to nail two studs who fill your RB slots all year. Below is an example of what a Superhero RB team might look like:
This team used its 7th round pick on an RB for stacking/falling ADP purposes, but only drafted five RBs, and didn’t take 4th/5th RBs until the late rounds. Superhero RB teams tend to come from early draft slots where you can land two studs (Taylor/Jones) who you hope to fill your RB spots most weeks. While this team managed to land two studs at RB, and strong players at the onesie positions, it’s clearly weaker than its Hero RB counterpart at WR, sporting Allen/Smith/Lockett/Moore as its top four options.