So long and thanks for all the fish. – Douglas Adams
When Zero RB, Antifragility, and the Myth of Value-Based Drafting came out in November 2013, I certainly had no idea of what was about to follow. Now with the debut of Fantasyland’s Zero RB episode, it felt appropriate to offer a Hitchhiker’s Guide. Hopefully it answers questions you might have had after listening. As always a big thanks to RotoViz and also to our other esteemed guests, Brian Burke, Steve Palazzolo, and Brian Owens.
What is Zero RB?
Zero RB is a WR-heavy, contingency-based draft approach. It uses the power of the antifragile to benefit from the chaos and stress of a fantasy season and thus build a dominant roster. In essence, the objective is to create a squad with overwhelming firepower at receiver and to win the race to fill the Flex.
What formats are favorable for Zero RB?
Zero RB is a strategy to attack PPR formats which allow you to start 4 or more WRs. This includes formats where one or more of the WRs will start in the Flex.1
What’s in a name?
Occasionally readers will reasonably object to the name because drafters are not going to literally select zero running backs. The idea behind the Zero RB moniker is to communicate the idea that drafters should select zero RBs during the high leverage portion of the draft where participants can load up on positions that both score more points and have a stronger relationship between ADP and final points.
Zero RB is occasionally labeled as “good marketing,” and while I certainly wouldn’t object if readers find the name compelling, the goal is to help differentiate the approach from other WR-heavy strategies that have very different structures, objectives, and results.
What is the definition of a Zero RB lineup?
Zero RB drafts can take many shapes, but the bare minimum to qualify would be a draft that selects at least 4 WRs or TEs before selecting a RB. Such a lineup would be considered a weak version of Zero RB.
In Part 2 of the original article, Zero RB and Antifragile Rosters in Real Events, the lineup I discussed had only a single RB in the first 8 rounds. My goal in every draft is to land six of the eventual Top 15 WRs. Although I have found it occasionally possible to accomplish this, the goal itself is important even if you fail to achieve it.
I’m frequently asked: “Where are you going to start all of those receivers?”
If we assume that one of our six WRs will underperform and one will struggle with injury, then we still have four elite starters – the bare minimum for dominating a league. If we have better luck, then having five or more WR1/2 hybrids puts us in good position to pull away during the bye weeks.
What is the “antifragile” and how does it relate to fantasy football?
Antifragility is a concept developed by Nassim Nicholas Taleb and defined in the following way:
“Some things benefit from shocks; they thrive and grow when exposed to volatility, randomness, disorder, and stressors and love adventure, risk, and uncertainty. Yet, in spite of the ubiquity of the phenomenon, there is no word for the exact opposite of fragile. Let us call it antifragile. Antifragility is beyond resilience or robustness. The resilient resists shocks and stays the same; the antifragile gets better.”
The main issue in fantasy football is how to deal with a RB position that no longer scores the way it once did. Fantasy owners have three basic choices for addressing the RB problem.
|Fantasy Strategy||Value-based drafting||Triple RB or RBx5||Zero RB|
Injuries are one common way in which Zero RB squads benefit from shocks, but they aren’t the only one. Depth chart changes, the emergence of backups, and the increasing value of receiving backs all provide other means of generating RB production.
The original article referenced the Myth of Value-Based Drafting. What does that mean, and what is contingency-based drafting?
Value-based drafting is based on the premise of defining our expectations of a player through projections and then valuing players based on their value over replacement (or over average starter). This is a seemingly straightforward and powerful approach, but it fails to hold up to real conditions. As humans, we always overrated our ability to predict the future. Player markets like those represented in drafts and auctions are notoriously difficult to beat while simultaneously having surprisingly limited predictive value themselves.
Moreover, when we create our preseason projections, we’re not usually very responsive to what history tells us about those projections. Over the last 5 years, RB scoring has been increasingly difficult to to predict using ADP and the shape of that scoring has shifted such that late round RBs are now more viable than ever.2
Contingency-based drafting, by contrast, is an attempt to embrace this uncertainty. Instead of representing players with a single number, we attempt to order our board with an understanding of the range of outcomes embedded in a player’s profile. We do this at RotoViz by applying multiple screens for player value: Similarity Scores, Projection Machine results, other algorithmic projections. We can then target players based on their range of outcomes and risk profile as it relates to ADP.
Beyond this respect for the wide variety of potential outcomes and our limited ability to choose the correct one beforehand, we must also be aware of the structural advantages to be gained by embracing contingencies over false certainties. Oft-cited and foundational strategic approaches like J.J. Zachariason’s Late Round QB and Steve Gallo’s Zero QB Theorem both fit naturally under this umbrella.
What do the numbers say?
In fantasy football, Zero RB and contingency-based drafting are a natural fit. Of the four relevant fantasy positions, ADP has the strongest relationship with WR scoring. Just as importantly, the scoring implied by ADP is higher in terms of total points and points above replacement. You can also expect to receive a full extra game played when you draft a WR over a RB.
We can break this down to a series of simple and true statements that highlight the necessity for loading up at WR.
- WRs score more points.
- ADP is more predictive for WR points.
- Most PPR leagues require more WR starters.
- Most PPR leagues have one or more Flex positions.
- In the high leverage rounds, WRs have more value than RBs.
How does Zero RB compare to other WR-early strategies?
Sometimes the easiest way to understand the objectives of an approach is to contrast that approach with other purportedly similar strategies. The most well-known adjacent strategy is Matt Waldman’s Upside-Down Drafting. Much like Zero RB, his choice of appellation is adept at conveying the strategy’s approach and objectives.
Upside-Down is a balanced approach where WRs are selected in the rounds usually earmarked for RBs and vice versa. The recommended draft start is WR-WR and then a mix of QB, TE, and RB in Rounds 3 through 6.
I recently participated in a draft that helps illustrate the similarities and differences between Zero RB, Upside Down, and more traditional approaches. These three teams were drafted from draft slots of 2, 3, and 4.
|Upside Down||Zero RB||Traditional|
|Odell Beckham||Julio Jones||Le’Veon Bell|
|Alshon Jeffery||Amari Cooper||Jamaal Charles|
|Eddie Lacy||Demaryius Thomas||Brandon Marshall|
|Cam Newton||Tyler Eifert||Doug Baldwin|
|Thomas Rawls||DeVante Parker||Allen Hurns|
|Greg Olsen||Stefon Diggs||Torrey Smith|
On first glance we see that the Zero RB squad has 5 WRs in the first six rounds, the traditional roster has 4, and Upside Down only 2. While RB-RB teams do not always then shift to hammering WRs, it highlights the differences between Zero RB and Upside Down when they do. In this instance the Zero RB squad and traditional squad have more in common than do the Zero RB and Upside Down squads.3
The contrast between Zero RB and Upside Down, especially as it relates to antifragility and contingency-based drafting, helps explain why Mr. Waldman did not recommend his approach for PPR leagues prior to the 2013 season.4
Zero RB, the Marketing Angle, and the Reach of Explanations
In the introduction I alluded to the “marketing criticism” of Zero RB. I prefer to see this as a compliment. We know that the stocks of companies with easy to pronounce names outperform those with more difficult ones. Beyond that, we can also surmise a narrative element to the transmission of ideas through a culture. Any time an idea has a pithy name that simultaneously embeds a large quantity of information, we should expect the transmission to be faster and more seamless.
However, beyond the marketing element, it’s my belief that the theory has thrived primarily because it works, and that this is the biggest contrast between it and similar ideas.
Quantum computing pioneer David Deutsch refers to this as the reach of explanations.
It is your conjecture, your original creation. Yet because it is a good explanation – hard to vary – it is not yours to modify. It has an autonomous meaning and autonomous domain of applicability. You cannot define its predictions to a region of your choosing. Whether you like it or not, it makes predictions about places both known to you and unknown to you, predictions that you have thought of and ones that you have not thought of.
This reach of explanations is the ability of some of them to solve problems beyond those they were created to solve. . . . The reach of an explanation is not something the creator of the explanation can use to obtain or justify it. We find out about it only after we have the explanation – sometimes long after.
Thus the reach of an explanation is neither an additional assumption nor a detachable one. It is determined by the content of the explanation itself. The better an explanation is, the more rigidly its reach is determined – because the harder it is to vary an explanation, the harder it is in particular to construct a variant with a different reach, whether larger or smaller, that is still an explanation.
This view of explanations itself has plenty of reach and myriad ramifications for pursuits like fantasy football.5 But the takeaway is simple: the idea of antifragility has reach in its ability to explain why certain approaches to fantasy football are successful and others are not.
If Zero RB has caused a paradigm shift, I believe it is due to the quality of its explanations.
Zero RB and the 2013 NFFC Primetime Championship
A few months after the original article appeared, I had the great good fortune to win the 2013 NFFC Primetime Championship. This became a small part of the Zero RB narrative during those halcyon days of summer 2014.6 For those who find “process over results” to be the tired excuse of losers, it offered some Bill Parcells-style proof of concept. In the long term we are very much what our record says we are.
It is perhaps ironic then that the team in question was not Zero RB. It was led by first round pick Jamaal Charles, the heir to Barry Sanders and the greatest RB in the league today. Selecting Charles was itself a contingency-based pick.
While I’ve been lucky enough to have one or more Top 15 finisher in a Main Event in three of the last four seasons, 2013 was a once-in-lifetime season by any standard. That year nine Banana Stand teams made the NFFC playoffs and race for the Grand Prize.
Seven of those squads were Zero RB, but many of those squads featured Calvin Johnson. Most of my championship hopes were dashed in a Philadelphia snowstorm Week 14. They were rescued by Charles’ transcendent 5-touchdown performance the following weekend. Bolstered by a WR corps that included Josh Gordon and Alshon Jeffery, the two remaining RB-early squads took home first and second place.7
Zero RB is one of the best ways to capitalize on the ideas of antifragility and contingency-based drafting, but it’s not the only way. It’s possible to draft RB-early teams that incorporate similar aspects, and I intend to write a RB-oriented look at contingency-based drafting later this summer. The idea behind that piece will not be to undermine Zero RB but to explore all of the various ways we can turn humility-based strategies to our advantage.
Zero RB versus Best Player Available
One of the most salient criticisms of Zero RB is that it forces you to ignore value at other positions and that it creates exploitable opportunities for RB-early drafters.
I hope it’s been evident that I’m very committed to Zero RB but simultaneously willing to look at it from a wide variety of perspectives and even to take RBs early in certain situations. It’s always important to be cognizant of player value and to be flexible during drafts.
The one caveat I would add is simply another warning about RB value. Traditional methods of value calculation have overvalued RBs to the point where we’re often tempted to see a sliding RB as “a value,” when it would take a much bigger slide for that player to be a good risk-adjusted bet. ADP may be a decent indicator of value within a position,8 but it’s a poor indicator of value between positions.
It’s easy to believe that the current pro-WR mentality is a case of recency bias, especially since it is true that 2015 was an outlier season as it relates to the shape of RB scoring. Awareness of this may actually obscure rather than elucidate WR value. FD wrote about this prior to 2014, showing that ADP didn’t track actual player value and offering this visual.
PPR SCORING – POSITIONAL ADP VS. VALUE
While ADP is shifting to reflect growing awareness, scoring trends are also accelerating in this direction.
I’m a big believer in Best Player Available, but WRs are the best players available in many situations where convention would disagree.
Thanks again to everyone who contributed to the Fantasyland episode, including Steve Palazzolo, Brian Burke, and Brian Owens. I also really appreciate the tremendous work by host Peter Overzet and the great work by Fantasy Douche, Matthew Freedman, and Patrick Kerrane on the production side. I hope everyone enjoys.
- It’s possible to use Zero RB or similar WR-heavy approaches in other formats, but in those formats Zero RB is merely a contrarian strategy and not a dominant strategy. (back)
- I asked RotoDoc to take a second pass at my ADP numbers since I don’t possess a completely unbiased eye. He fit a second order polynomial to both PPG and Points based on overall ADP, and looked at R-squared, Root Mean Square Error, and calculated variance by RMSE/Mean of Response. He will have some more detailed thoughts on this in the future, as there’s a lot happening beyond the obvious. Here are the simple results.
ADP is taken from myfantasyleague using the settings “real drafts” and “post-August 25 ADP.” (back)
- The point isn’t that Zero RB and traditional approaches will necessarily share a common interest in Rounds 3-6, but that Zero RB and Upside Down are structurally separated in these rounds. While RB-early tactics can be implemented with contingency-based elements, Upside Down requires you to hit on your TE, QB, and middle round RB picks. (back)
- From his 2013 Upside-Down article:
I believe the strategy works best in non-PPR scoring formats where there isn’t a third RB option allowed as a flex for starting lineups. If you play only in PPR leagues or RB-flex-friendly lineups, going “Upside Down” can help you win big, but the risks are higher because the turnover in wide receiver rankings is higher in this league type and the supply and demand dynamic is minimized. The information I’m providing today is only for non-PPR formats only. I will not be writing a PPR or flex-RB strategy later this summer or in the near future. (back)
- For example, it’s easy to vary the explanations for scouting-based ‘misses,’ which makes it structurally difficult for scouting to improve. (back)
- The summer of 2014 was just two years ago, but it seems a lifetime in the history of fantasy strategies. (back)
- In the podcast, I mentioned that the winning team had Rob Gronkowski and Julio Jones, both of whom were injured, but I had their rounds flipped. Jones was the 2nd round pick and Gronk the third. Aside: The aforementioned squad with six of the Top 15 WRs was the best of my Zero RB playoff teams and finished 11th. (back)
- and even there it’s quite limited (back)