The winning strategy in guillotine league football is Bid Low, Bid Late, which calls for spending as little as possible until late in the season. In order to survive late into the season, players should aim to start with a resilient team.
This article suggests seven tactics for drafting in a guillotine league to build a maximally resilient team in support of the Bid Low, Bid Late strategy. These are the tactics I am suggesting:
- Bye Diversity
- Singleton the Singletons
- Late Bye
- Crowded Bye
What follows is a work in progress, representing my best hypotheses, and not a backward-looking analysis of proven tactics. The guillotine league format is only a year old, and the one league I have played in involved an auction, not a draft. I am implementing these tactics, to various degrees, in the leagues that I am playing in this year. I expect to report back on how well they worked out.
Fanball Guillotine Leagues
The tactics I am describing here are tailored to Fanball guillotine leagues. These leagues have 17 teams, each of which drafts 12 players. Starting lineups consist of seven players: QB, 2 RBs, 2 WRs, TE, and flex (RB, WR or TE). Scoring is conventional PPR, except that passing touchdowns are worth six points, not the conventional four.
The season is structured differently than conventional fantasy football seasons. The first thirteen weeks are the “regular season,” where each week, the lowest-scoring team is eliminated, and its players are dropped to waivers. Weeks 14-16 are a three-week playoff, with the winner determined by the cumulative three-week score. The top three positions pay out.
Overall, Fanball’s settings seem to be aimed at forcing teams to spend FAAB more and earlier. Shallower benches mean more turnover, and a three-week playoff involving four teams puts a cap on the value of bidding late.
This should mean an adjustment to the Bid Low, Bid Late strategy, but not an abandonment of it. I think bidding low and late remains the best strategy, for all of the reasons I have stated in my articles: (1) A good team is not necessary to avoid elimination early in the season – it is mostly luck, unless your team is truly terrible; (2) Free agents become better and more numerous as the season progresses; and (3) Free agents become cheaper as the season progresses.
However, the Fanball settings give Bid Low, Bid Late less of an advantage than a winner-take-all league, and it might mean that the optimal strategy is to spend more and earlier than in a winner-take-all league. The diminution in importance of strategic spending means that more traditional fantasy football skills, like player assessment and skill in week-to-week lineup setting, are increased in importance.
It seems to me that Fanball’s settings, particularly the short bench, also make it more important to draft a resilient team that could survive the early-to-mid season, and reduce the need for FAAB spending.
Tactics for Building a Resilient Team
An important note about these tactics: They are not meant to be used in a heavy-handed manner to pick players with substantially worse projections over better players, just because some of the tactics favor the worse player. Rather, they are considerations to apply when choosing between players who are relatively closely matched. I would put them at slightly above tiebreakers: I would pick a slightly worse player over a slightly better player in the same tier if the tactics weighed decisively in his favor. For example, I slightly prefer Brandin Cooks to Julian Edelman in conventional PPR, but in guillotine I would pick Edelman because of of better week-to-week consistency. Call it a tiebreaker+.
Tactic 1: Anti-Volatility
In regular league play, I don’t normally mind picking boom-or-bust players. I figure that over the course of a season, in a 9-man lineup, the ups and downs don’t matter that much. But in a league where the lowest score in any of the first thirteen weeks gets you eliminated, and especially in a league with only 7 starters, a high amount of volatility can kill a team. For that reason, I tried to avoid players that I perceived to be high-risk, even at the cost of high reward. “Neither your honey, nor your sting,” as the ancient Jewish maxim counsels.
I distinguish three types of volatile players: First, those who have a wide range of outcomes when they start. I’m thinking of players like Ted Ginn and other downfield runners for whom the difference between a terrific day and a terrible day might be one or two missed connections. Other type-1 volatile players might be a touchdown-dependent tight end, a goal-line running back, and any player whose usage depends on gamescript.
The second type of volatile player is one who may or may not play during the season due to suspension, holdout or roster status, or a player whose length of absence is unknown for these reasons. As of this writing, Ezekiel Elliott and Melvin Gordon are type-2 volatile due to potential holdout, Josh Gordon is type-2 volatile due to a suspension which may be lifted, and Jay Ajayi is type-2 volatile due to being a presumably good player who is not on any roster.
The third and last type of volatile player is the injury-prone player. This is someone who performs well when he is on the field, but might be expected to miss significant game time. Last year I would have said that Jordan Reed is type-3 volatile, but now I’m not so sure he performs well when healthy anymore. Will Fuller is an okay example of such a player, but he also has a significant element of type-1 volatility, as a player who flourishes on long receptions and touchdowns.
I am most worried about type-1 volatility, since that is the player who I am most likely to start and get a terrible week from. In contrast, the effects of type-2 and type-3 volatility can largely be eliminated by not starting such players in games where they are expected to score 0 points. The nightmare guillotine league scenario is starting a John Brown type, getting a 1.5 point game from him, and being the lowest-scoring team by 3 points.
I am least worried about type-3 volatile players. I would draft a player like this so I can play him when he’s healthy, and play someone else when he is injured. As long as his performance when he starts is fairly consistent, the cost of drafting him is a bench spot. Of course, it is better not to burn a bench spot, but the cost of doing so is much less than the cost of starting a player who performs terribly.
As far as type-2 volatility (suspension/holdout/unrostered), I treat it as follows: Since I don’t think I know how to evaluate the chances of a holdout, I discount for it. I take late-round fliers on unrostered players who I think likely will end up usable, like Jay Ajayi.
For suspended players, I distinguish between definitely suspended players and maybe suspended players. I significantly discount players who are definitely suspended for four or more games, not because of volatility but because of the next tactic I’ll discuss, Anti-Stashing.
I do not mind paying the appropriate price for a player who may or may not be suspended at the beginning of the season. For example, I think drafting Josh Gordon in the late rounds is a good idea, and I have done so several times in my early Fanball drafts. If he is surprisingly reinstated, the dart-throw has paid off. If there is no hint of reinstatement when the season is near, I can drop him for a usable player.
(For further reading on volatility, including the fascinating paradox that more consistent players score fewer fantasy points than their volatile counterparts, yet are more valuable in fantasy, see The Predictability Paradox.)
Players to elevate: Reliable, high-volume receivers like Julian Edelman and Keenan Allen.
Players to fade: Boom-and-busters like Ted Ginn, Mike Williams. One-dimensional RBs like Jordan Howard and Tarik Cohen.
Tactic 2: Anti-Stashing
This one is pretty simple. Because there are only five bench spots on a Fanball guillotine team, it is not a good idea to waste a spot on a player who is not going to produce until later in the season. This rules out players who are injured (or semi-injured) to start the season (like A.J. Green), players who are suspended at the beginning of the season (like Kareem Hunt), and any rookie WRs, TEs and QBs, unless there is reason to think they will be productive early in the season. This is based on my impression that non-RB rookies rarely produce much in the early season.
If you think a player has a good chance of breaking out in the second half of the season, why draft them now, when they are likely to be available cheap later in the season? Remember that the number of rostered players shrinks as the season goes on, so that of the 204 players rostered when the season starts, no more than 108 will be rostered by Week 9.
It seems much better to roster a productive backup for bye-week fill-ins and injury replacements, or a player who might break out at the beginning of the season, like a rookie running back or someone who is an injury or a poor performance away from possible ascension. In the later rounds, I would prefer a John Brown, Nyheim Hines, Brian Hill, Albert Wilson, Devin Singletary, or any player contending to be relevant on a high powered offense (like Marques Valdez-Scantling or Devin Funchess, or in the last rounds, Jake Kumerow, Rishard Matthews or Deon Cain) over a hopeful rookie breakout like Noah Fant or N’Keal Harry or a suspended player like Chris Herndon.
What about Kareem Hunt? I would, with some misgiving, take a flier on him around round 10. My problem with Hunt is that he may not have value until so late in the season that he is not helping to get over the weeks 6-10 hump. He is suspended for 8 weeks. Unless Chubb has flopped or is injured, I would be wary of starting Hunt week 9 until I’ve seen his usage. That means he might be usable for 3 weeks, weeks 10-12, until I’ve replaced him (since I’m expecting to have 3 of the top 10-15 RBs by the late season). Burning a bench spot for 9 weeks seems like a high price to pay, which lowers his draft value.
Players to fade: A.J. Green, Kareem Hunt, Chris Herndon, Kyler Murray.
Tactic 3: Anti-Stacking
For the same reason that stacking is useful in daily fantasy sports (DFS), I believe that anti-stacking is useful in guillotine leagues.
Successful DFS players, trying to beat a field of competitors, shun safety in pursuit of outlier scores. Often, this takes the form of stacking players from the same team, to benefit from the fact that a high-scoring offense tends to mean multiple high-scoring players.
In guillotine leagues, you are trying to resist elimination by avoiding an outlier score. You want to avoid players from the same team in order not to have multiple players dragged down by a team’s low-scoring week.
As this article shows, there is a meaningfully strong correlation between fantasy points scored by a starting QB and his top wide receiver, with weaker correlations between a starting QB and his top tight end and between a starting QB and his 2nd and 3rd best receivers.
This correlation means that in a guillotine league, I would want to avoid starting a quarterback and a receiver from the same team, if I can help it, especially a quarterback and his best receiver. The high correlation means that when the top WR turns in a bad game, it is likely that the QB will score poorly as well. And two players scoring poorly significantly increases the chance of elimination.
For draft purposes, this means that when I draft my QB, I try to avoid getting someone from the same team as my tight end or (especially) my wide receivers who are the WR1s on their team.
So if I have Michael Thomas and I am facing a choice between Drew Brees and Matt Ryan, I would go with Ryan (if I value them similarly, as ADP currently does). If I have TY Hilton and can draft Luck or Rodgers, I would take Rodgers. If I have Zach Ertz and I have to choose between Carson Wentz and Russell Wilson, I would go for Wilson. And so on.
If I draft a top QB-WR or QB-TE stack, I am probably going to play the stack week to week, which means I am increasing the risk of a bad outlier performance from my team. To put it simply: if Brees doesn’t throw any touchdowns, Michael Thomas doesn’t catch any touchdowns. If Brees has a bad week for fantasy scoring, either because he couldn’t complete many passes or because the game script called for lots of rushing, then it is likely that Thomas had a bad week for fantasy too.
If I start Drew Brees and Julio Jones week to week instead, their performances are not correlated. I will get fewer outlier high scores and low scores, reducing the risk of elimination.
I do not worry about stacking a QB and a running back. The correlation between their performances is pretty low.
What about negative correlations?
You might be thinking: on this logic, shouldn’t I draft, then stack, players for negative correlation?
This is a good idea in theory, but in reality, there are just no meaningful negative correlations between teammates. The strongest negative pairing is between RB1 and RB2, and it is not a strong correlation. It is also not easy to find two RBs on the same team who are startable (now that Ingram is on the Ravens), and I do not think it makes sense to force a stack like Mack-Hines or Montgomery-Cohen when there are likely better options available. You also may not have the stomach to cap your team’s upside by stacking RBs. The one platoon I can think of that I *might* like is Carson-Penny, but only if they are both actually startable and are my best options, or close to it.
What about strong negative correlations between players on opposing teams? These exist, and could be useful in theory. But they all involve team defenses, so they are not useful in Fanball guillotine leagues, where defenses are not played.
There is also a significant positive correlation between QB scoring on opposing teams. This is another piece of information that is useful in theory but not in practice, because you cannot start more than one QB in this type of league. Even in superflex, this correlation should not affect draft strategy because any two QBs will oppose each other at most twice a year, not enough to inform draft strategy.
Tactic 4: Bye Diversity
Bye diversity means spreading out the bye weeks so that your top performers are not benched on the same week. This dovetails with anti-stacking, since when two players are benched on the same week, their scores are 100% correlated (and not in a good way).
It seems to me especially important to avoid multiple byes up to around week 10. By week 11 or 12, I would expect to have a stacked roster anyway, so finding two good substitutes should not be a problem. Moreover, there will be many starters on bye those weeks, which means many of your competitors will be at least one and maybe more players short.
In contrast, weeks 4, 5 and 8 each have only two teams on bye, and they’re not teams loaded with talent. I would be especially wary of multiple byes on one of those weeks, particularly week 8. See Tactic 7, Crowded Bye, for more discussion.
Obviously, with a 12-man roster, there are going to be some players with the same bye weeks. What you should try to avoid is having coinciding bye weeks from your anticipated starters, especially in weeks 4-10. Maybe starting your draft with Mixon-Gurley (bye week 9) or Elliott-Cooper (bye week 8) is not the best formula for resilience.
Tactic 5: Singleton the Singletons
This tactic preaches drafting just one quarterback and one tight end. There are only five bench spots in Fanball guillotine leagues, and ideally they are all used to store wide receivers and running backs, either as backups or as breakout candidates.
Singleton the Singletons is in contrast to approaches such as “Quarterback by Committee,” where you draft two quarterbacks late and play the one with the more favorable matchup each week. I don’t like this approach in guillotine leagues with shallow benches, for two reasons.
First, with only five spots on the bench, each bench spot should be used to maximize the resilience of your team, in order to allow you to avoid having to spend your FAAB early, which in my view is the worst thing you can do in a guillotine league.
Second, I am skeptical of people’s ability to predict matchup strength before the season starts, which dilutes the value of committees.
But even if you think that you are able to predict a quarterback’s strength of schedule before the season starts, you should avoid the bench-burning Quarterback by Committee strategy. Instead, use the following strategy: draft a quarterback late who has a strong schedule in the early-middle weeks, say weeks 4-7. Then around week 8, buy a good quarterback for cheap when the market is flooded. By week 8, there will be only 9 other teams, and several of them will probably have realized the wastefulness of holding multiple quarterbacks on their roster.
This will allow you to save your higher draft picks for RBs and WRs, leave all of your bench spots for RBs and WRs, and spend cheaply on your QB while ensuring good QB coverage from week 4 through the end of the season. (You don’t really need a good quarterback in the early weeks, when survival depends much more on probability than on whether you started Patrick Mahomes or Nick Foles. For discussion, see my article Rely on Probability, Not Spending, to Survive Early.)
As far as tight ends, I think the best strategy is to not stress the position too much and to implement the Late Bye tactic. Try to hit on a late rounder (Jordan Reed? Mark Andrews? Kyle Rudolph?) or else just count on picking up a good tight end off waivers late in the season.
Players to elevate: Late round tight ends and QBs.
Players to fade: Early and mid-round tight ends and QBs.
Tactic 6: Late Bye
This is more for the singleton positions, and especially for tight end.
I think it is a bad idea to have a tight end taking up a bench spot. And that means that if you are drafting a tight end, either draft one who has a late bye, or draft one who is dispensable, so you can drop him for his early or middle bye and pick up a relatively cheap equivalent replacement.
This is why, as much as I like O.J. Howard, I would underdraft him because of his week 7 bye. It’s also why I fade George Kittle.
I would try to avoid QBs with early byes because of the same dynamic. If I spend an early-mid round pick on Andrew Luck, I am probably planning on holding him past his week 6 bye, and that means wasting a bench spot or scrambling for a cheap substitute.
To a lesser extent, I would use this tactic for the RB and WR positions. Because I would expect to always have backups on my roster for these positions, I am less worried about bye weeks, as long as I have implemented Tactic 4: Bye Diversity. But faced with a choice of two equal WRs or RBs, I would take the one with a later bye.
Players to elevate: Tight ends and quarterbacks with late byes.
Players to fade: Tight ends and quarterback with early-mid byes.
Tactic 7: Crowded Bye
I am not obsessed with byes… I just think that they are very important in guillotine leagues. I have argued that you should not have too many top players with byes on the same week (Bye Diversity) and that your QB and especially tight end should not have bye weeks too early (Late Bye). This last bye-related tactic, Crowded Bye, counsels that you draft players whose bye weeks coincide with many other starting players’ byes.
I think this tactic is best illustrated by looking at week 8. This season, only two teams are on bye that week – the Ravens and the Cowboys. That means there is only one superstar player – Ezekiel Elliott – on bye that week (and there are only a handful of likely starters on bye – Amari Cooper, Mark Ingram, and maybe Mark Andrews if he lives up to the hype).
That means that if you own Zeke, no other team is as disadvantaged as you are by predictable player unavailability that week. If you are looking at Elliott in the first half of the first round, you might prefer to draft McCaffrey and his week 6 bye, when Odell Beckham, Nick Chubb, James Conner, Juju Smith-Schuster, and Mike Evans are also on bye.
Or maybe you draft Kamara, whose week 9 bye is shared by Michael Thomas, Julio Jones, Devonta Freeman, Joe Mixon, A.J. Green, Todd Gurley, Brandin Cooks, Robert Woods and Cooper Kupp. Or David Johnson, whose Week 12 bye is shared by Tyreek Hill, Damien Williams, Pat Mahomes, Melvin Gordon, Keenan Allen, Dalvin Cook, Stefon Diggs, and Adam Thielen.
There are two other weeks with only two teams on bye: week 4 (Jets and 49ers) and week 5 (Dolphins and Lions). I am more worried about week 8 than about these early weeks. Weeks 4 and 5 are still early weeks, where a team that has to bench Le’Veon Bell, George Kittle or Kerryon Johnson can count on beating out one of 12 or 13 other teams by some combination of better team and better luck, as opposed to having to beat out one of nine other teams in week 8.
Keep in mind that the later it is in the season, the fewer players are starters, which concentrates the talent. In week 4, your opposition starts 91 players. In week 5, your opposition starts 84 players. In week 8, your opposition starts only 63 players. One out of three players who are startable in week 4 are not startable in week 8.
Keep in mind, too, that if you are employing the Bid Low, Bid Late strategy, your team will be quite vulnerable in week 8, as you patiently, with great restraint, bide your time while the rest of your league overpays for players. You are largely relying on chance and resilience to survive until week 10 or so. Trying to survive week 8 without a top player is like kneecapping chance while she is on the way to help you.
Players to elevate: Players with byes in weeks 7, 9-12.
Players to fade: Players with byes in week 8.
To build a resilient team, avoid volatility (Anti-Volatility, Anti-Stacking, don’t waste bench spots (Anti-Stashing, Singleton the Singletons), and minimize the stress imposed by bye weeks (Bye Diversity, Late Bye, Crowded Bye). Just remember that these tactics are a complement to, and not a substitute for, drafting a good team, and should be used only as a tiebreaker or tiebreaker+.