6 more key notes on Underdog's Eliminator format
I can't stop thinking about the best ways to build Eliminator teams
Going to try to be relatively quick today as it’s my kids’ last day of school and we’re hopping in the car for an 8-hour drive to their great-grandma’s house this afternoon. But I can’t shake all the subtle considerations of the Eliminator format on Underdog, and I wanted to write up some more things I’ve been considering before taking off on that trip.
For the background on how I’m approaching this format, see my initial piece. I talked through everything from the simple — like how to play the bye weeks — to the more complex — like an interesting opportunity to play the format like more of a cash game by loading up on late-season byes because you will have a profitable roster if you go out in the heavy bye weeks of Week 12 or Week 14. I’ve removed the paywall on that post, and will leave the content this one free for all readers as well, as I’ve been discussing this format often on my new YouTube channel (go subscribe to the channel, it helps me out!) and decided to just make my Eliminator content open to all.
The way I’d describe my general argument in that initial piece is that you need to be willing to build a more upside-based roster than you’d initially think for a format where weekly consistency is the name of the game. Floor feels important, but I made the case in a variety of ways that the way to achieve weekly floor isn’t positional balance or targeting the boring (read: fragile) profiles that only get drafted due to projected volume, but rather by building a roster that has enough upside that it can win its pods most weeks, and then when it has down weeks or players on byes, it can still survive that as well, because the team is just that good.
Every fantasy team I’ve ever built that put up consistently strong weekly scoring did so because it had league winners on it, and it was capable of putting up 25- and 30-point weeks in several different lineup spots. I have also built a ton of teams where I’m trying to delicately stack small wins, and while some of those teams have been above average over the course of the season, they tend to be less dominant in their weekly scoring profile, which does tend to lead to some subpar weeks.
In the time since I published that, there are six key things I’ve come across or considered further that I want to add to what I wrote in that column. Many of these things were sparked from a great conversation with Shawn Siegele this week over at Stealing Bananas that you’ll need to listen to if you’re into this format. We drafted a team in this contest as part of the show, and talked through several interesting considerations during it.
The first of these six points drives home that you need a dominant team.
1. You might not even get to the byes
In my first post, I shared a table that looked through the advance rate of different pods by week, as well as when the payouts hit. What I felt like I needed to add in the time since is one that shows just how much of the total field has been eliminated as we enter each week.
I’ve added that column (third from the left) in the below table, but here are some quick notes:
By the time we hit the first bye week in Week 5, only 37.3% of the total number of Eliminator teams that were drafted this year will still be active.
To get your team into the money and earn a payout, you need to make it to just Week 8. But only 13.1% of teams will do so.
The heaviest bye weeks are Week 12 and Week 14, which both include six teams, many of whom are very interesting ones to target for fantasy. In the first post, I talked through how intentionally stacking these byes can increase your probability for just advancing into a range where you could make a nice profit ($50 on your $10 buy-in if eliminated in Week 12, or $100 if eliminated in Week 14). But note that just 2.3% of teams will even make it to Week 12, and just 0.8% to Week 14. So even if you stack those late byes, you’re by no means guaranteed to survive that long (which for me has been freeing in stacking those byes).
I understood this on some level when I wrote my first piece and tried to push you toward being willing to embrace riskier builds like only two players at each of the onesie positions, but even I was taken aback by how likely I am to lose a ton of these Eliminator teams in the very early part of the year.
The immediate conclusion for me was to be less worried about bye week specifics and more worried about building strong rosters. A path through this gauntlet feels more likely by building a really strong team that gets some pod luck in the one week where it is particularly weak due to byes, rather than a weaker team that was overly cautious with floor picks and bye week considerations.
2. Early-round vs. late-round bye balance
Those notes I ended the last section with do not mean you don’t consider byes. They simply mean that dropping down a tier purely for bye week reasons is the tricky thing.
This second note for me has been the realization that late-round picks aren’t likely to contribute most weeks anyway. This has impacted my thought process regarding achieving “balance” with my byes.
In the first piece, I showed how if you perfectly balance bye weeks across your roster, you’ll average 2.25 players that had each of the eight bye weeks. I’ve used that thought process as a justification for perhaps stacking a few early picks on the same bye, if I don’t suspect I’ll need to go back to that bye later, and thus I can finish the draft “balanced.”
But that mentality fails to recognize the strength of the picks in various points of the draft. I’ve done this before with Week 9 byes, for example, where I took an early 49er and then stacked that player with Brock Purdy and I also grabbed Jaylen Warren as a RB target. I know that the 49ers and Steelers are the only teams on Week 9 bye, and I don’t have a lot of late-round targets for these teams, so taking three players for that particular build in my first 9 or 10 picks with that bye wasn’t a major concern for me because I was likely to finish that build with just three Week 9 byes.
But the reality is late-round picks aren’t expected to contribute every week in the same way early-round picks are. For late-round picks, you’d likely consider it a win if you get five usable weeks all season. In that way, while we do know one of the weeks they’ll give us a zero (their bye), they in some ways should be thought of as having a whole bunch of byes. They aren’t a reliable bet for weekly production, just spike-week depth.
This realization meant for me that when I take three Week 9 byes as part of my first 9 or 10 picks, I’m putting a lot of my “reliable” scoring — an assumption I should make if this team is going to be good is most of the early picks are somewhat reliable — out in the same week. It’s not just three players on bye, but a higher percentage of my bankable points — including key pieces at both onesie positions if it’s George Kittle I’ve stacked with Purdy — and that is going to make Week 9 a difficult one to survive. That’s still maybe the right bet in some contexts, because again you could run into some pod luck, but it’s still a consideration worth thinking through.
The other major reason I’m emphasizing this idea of balancing byes in the earlier rounds separately from the later rounds, is there’s a real benefit to having your byes balanced early in drafts: Optionality. If you successfully navigate the first 8 or 10 rounds of your draft with basically no bye overlap save for maybe an intentionally stacked QB, or two players at different positions (like a Week 12 bye in both your RB and WR rooms, but not bye overlap within either position), you free yourself up to really take the best pick in the remaining rounds of the draft.
I started with the note that overlapping byes in the early part of your draft probably does have an outsized impact on your ability to have a solid week that week because I think that’s important, but this note about allowing yourself the optionality to target any particular bye in the back half of your draft is the way more practically applicable note. It’s really tough when you have too many, say, Week 6 byes in the early part of your draft (easy to do with Miami, Kansas City, and the Rams all on Week 6 bye!) and then are staring at a falling Ty Chandler who is a perfect fit for your RB build but he’s yet another Week 6 bye, or something like that.
3. How bye imbalance impacts pod strength
Taking this idea of bye balance a step further, one of the major epiphanies I had while talking on yesterday’s episode of the Stealing Signals YouTube show was how pod strengths will be impacted by how many NFL teams are on bye that week.
In other words, I initially thought through how I could get close to 2.25 players on bye for each of the eight weeks where there are byes — which meant going out of my way to target Week 7 and Week 9 byes (because there are only two teams on bye both weeks) and breaking ties away from the heavier Weeks 12 and 14 — but the overall field is dealing with that same environment.
In other words, in a perfectly equitable contest, the teams you are up against in your pods will average more players on bye in Weeks 12 and 14 than in Weeks 7 and 9. Thus, even if you have more Week 12 or Week 14 byes on your roster, and are going into those weeks short-handed, your opponents are more likely to be short-handed those weeks as well. Your relative strength is fine.
The other way to think through that is to draw back to my Purdy/Kittle/Warren team from the last section. I chose that one in the last section for a reason, because the issue I ran into was compounded here, as well. Having those three Week 9 byes stacked in my earlier rounds is a major problem because in Week 9, the other teams I’m playing against are unlikely to be hit with heavy byes. There will be some teams that have Christian McCaffrey and other key 49ers’ players, to be sure, but I’d wager that if I do make it to Week 9 with that team, I’ll likely be the only team in my pod size of seven to have three players in my first 10 picks on a bye. That makes it that much more difficult to finish top five in the group and advance that week.
By comparison, on some of the rosters I’m building where I’m willing to be a little heavier on Weeks 12 and 14, it’s worth considering that if I do get all the way to those weeks, it’s fairly likely other teams in my pods will be dealing with bye week issues, too, because there are six teams on bye each of those weeks, including many good players who — if they have strong seasons — are likely to be a major reason why teams make it that far, given they won’t have had a bye up until that point.
Maybe those other teams in my pods in Weeks 12 and 14 will have a worse combination of byes and injuries at certain positions are are more prone to taking a zero. Maybe they’ll have been more dependent on Week 12 and 14 players just to get there, and are now having to survive a week without them. Maybe pod luck in Weeks 12 and 14 will be more variable as a result. Probably.
All this said, I’ve stopped thinking about bye week balance in terms of the 2.25 per week average. I’m more willing to get to three or four Week 12 or 14 byes, as needed, and I’m not pushing to get anymore than two Week 7 or 9 byes onto my roster. I still want to be somewhat balanced, but I’m considering what I’ll be up against those weeks.
4. Divisional bringbacks
It’s been discussed for Weekly Winners contests and others how stacking up divisions allows for two game stacks through a year. This idea of having a divisional bringback for a primary stack can be really impactful in the Eliminator especially.
Consider a build where you wound up with a late-round Bo Nix stack with another Bronco, and also had guys like Josh Palmer and Brock Bowers on the roster. This example is meant to focus on players without high ADPs, but can obviously relate back to higher-drafted players.
But the Broncos play the Raiders in Weeks 5 and 12, so in both of those weeks you have Nix with a pass-catcher and Bowers as a bringback to help with weekly upside. The Broncos also play the Chargers in both Weeks 6 and 16, so now you have similar but with Palmer as a correlated bringback either week. It has only taken four roster spots, but you’ve built in some correlated upside in four different weeks, and that kind of thing could be huge if you hit on the blowup game on a week where you are bye-heavy — say, in Week 6 Nix, his WR, and Palmer all have strong performances in a surprisingly high-scoring affair.
One key point here is all Week 18 matchups are divisional contests, so only two of each team’s three divisional opponents provide this edge. For the Broncos, because they match up with the Chiefs in Week 18, they only play the Chiefs once in the fantasy-relevant weeks, and you don’t get this double dip benefit. The Week 18 schedule becomes one to consult when playing this strategy.
5. Three-QB builds (all stacked)
In the last piece, I emphasized that I was comfortable with only two players at either of the onesie positions. As I’ve drafted more of these teams, I remain very comfortable with two strong TEs, because the weekly floor at TE is going to be low regardless (making it more viable to survive a week even if I have to take a zero at TE than I might be able to with a zero at QB). And I want to be prioritizing elite TEs across formats in 2024 as it is, and it frees up a roster spot to not use three on both onesies.
But I’ve really come around on the three-QB builds. I’m really starting to like some of the late-round QBs I’ve written about recently, and there are huge potential advantages to the build types with three stacked QBs.
We don’t want to be overstacking because of how it creates logjams on those bye weeks, so having three single stacks is a more balanced build from a bye-week perspective.
We have three shots at weekly upside spikes, which is the idea of three-QB builds in normal formats, where you can get different ceilings across the three different playoff weeks. This carries over stronger in Eliminator, where we’d like access to different weekly ceiling every week if we could get it.
We can really maximize the divisional bringback lever discussed above, because we could theoretically have two separate divisional bringbacks for each of the three single stacks, using four roster spots to provide us four mini game stacks for each of our three QBs, or 12 mini game stacks across 12 roster spots (six of which would be the single stacks) if executed perfectly equitably.
I still think two-QB builds could be viable with an elite at the position, but I would really want to bet on that elite to be a star all year, and I’d likely want multiple divisional bringbacks and also some Week 17 correlation built in with that QB. If I’m betting on Jalen Hurts to be so good this year he provides me usable weeks almost every time out and I don’t really need the three QB stacks, then I’m betting on Eagles games to be fun for fantasy and I’ll want bringbacks that build out that particular bet (with some correlation built into my QB2 pick, as well, ideally).
But as I just wrote, I’ve really come around on balancing this across three different QBs because of how it can impact overall flexibility of scoring and give you weekly upside outs through the gauntlet of weeks. Building in these game stacks would be particularly important for weeks where I have too many players on bye.
6. Stability plus upside players
In trying to think through some player archetypes I’ve been targeting in Eliminator, one key feature for me has been players that aren’t solely “floor” plays, but do have some claim to early volume and role stability. In much of what we do, we talk about the concept of antifragility, which aims for players who can not only withstand a shock to the system, but who thrive because of it.
I don’t want to avoid truly great antifragile bets in any format, because when they hit they define seasons. That said, in Eliminator formats I’m leaning toward rostering players that don’t need a shock to the system to elevate; they might have the inside track at a strong season merely through inertia.
A key point in these builds for me — because I talked about in Part 1 leaning toward Zero RB and Zero RB adjacent strategies in Eliminator — is building out my RB room. The Cardinals are a great example of what I’m considering here, where Trey Benson is your classic antifragile bet who could be a league-winner if James Conner misses time.
And yet, Conner fits that archetype of back that is priced down in today’s fantasy football landscape relative to how pricy it used to be to target “guaranteed touches” at the RB position. He also seems to have the inside track at the lead RB job, and is coming off the best rushing season of his career. For him, the upside scenario doesn’t require a shock to the system; it requires him to be what the Cardinals are hoping he will be, and keep Benson on the sidelines for much of the latter’s rookie year.
Similarly, I’ve talked seemingly ad nauseum about how Raheem Mostert is an eighth-round pick who led the NFL with 21 TDs last year. If he doesn’t get hurt, it’s hard to imagine he isn’t a win there. Jaylen Wright is another massive antifragile bet, and I’ve been willing to make that bet on Eliminator teams after ADP, because he can still be huge. But Mostert is the cleaner bet for this format, where I feel comfortable with the early season role and some benefit to my weekly scoring in a floor sense, plus I can see the obvious upside just by looking at his 2023 stat line (which he likely won’t replicate, but is still evidence that ceiling clearly exists here).
These types of bets are harder to delineate at other positions which are more talent-based, but I mentioned Josh Palmer earlier, and he and probably more realistically Quentin Johnston are good examples of this. I don’t know that Palmer has true upside, but QJ’s case is that he does reverse the concerning Year 1 trends. Short of that, though, he does look like a guy who will at least be given plenty of routes, and the role certainty probably provides some degree of floor usability where he runs into some decent weeks even if he doesn’t break out, simply due to the more consistent playing time that seems likely to come for him this year.
Am I betting on QJ to break out in 2024? Not really. And he could be a problem for some of my Eliminator builds. But if I’m going to take him in some format, this is one where his Week 5 bye is helpful, and I do think his expected role provides some weekly help in that it could be the difference between me taking a zero in a given week that might lead to me being eliminated.
That type of thinking often winds up with you landing on bad picks, and I try to drive home in other formats that taking a zero at WR one week rather than, say, five points, isn’t likely to dictate whether your roster advances after Week 14 or not. But in an eliminator format, there are cutoffs you need to clear every single week, so the likelihood of a five-point week being just enough to stay above a cutoff one week ratchets up merely due to there being so many more of those cuts.
Hence, I’m more likely in this format to take QJ in another purely because I feel better about his projectable routes. It doesn’t mean I’d reach for him, or prioritize him over a player I felt much better about from a skill level standpoint.
That’s all for today. Because there’s just so much great strategy stuff to try to contemplate in the Eliminator realm — including a lot of schedule considerations, like those game stacks on weeks where you’re heavy on a bye — one thing I’d emphasize is slow drafts can be great for this format. It’s probably also possible to get really good at the important considerations in fast drafts, but I’ve enjoyed having the time to think through options in the slows for Eliminator.
Until next time!