There's a lot of strong dynasty analysis out there, especially when compared to five or ten years ago. But most of it is so dang practical-- Player X is undervalued, Player Y's workload is troubling, the market at this position is irrational, and take this specific action to win your league. Dynasty, in Theory is meant as a corrective, offering insights and takeaways into the strategic and structural nature of the game that might not lead to an immediate benefit but which should help us become better players over time.

The Best Team Usually Loses

One of my favorite topics to write about is how unlikely any given team is to win a championship. Last year, I shared data from over a thousand leagues showing that the higher-seeded team won about 58% of matchups. I started this year by building a model that calculated title odds based on various assumptions to test the impact of "win-now" trades.

I've used a half-dozen other approaches over the years, and they all tend to converge on the same point: there's about a 1-in-3 chance that the best team will win the title in any given year. Which means there's a 2-in-3 chance it goes to a team other than the best one.

Usually after sharing these sobering figures, I end with something to the effect of "hopefully you managed to reach your league's championship; if so, may the best team win". But one of our core values at Dynasty, in Theory is "question the obvious stuff". Or, as I put it in the third column, more than a decade ago:

It is because of this phenomenon that one of my personal aphorisms is “There are few things as dangerous as a plausible narrative.” Plausibility is a drug that lulls us into complacency. It causes us to accept without verifying. Some of the greatest fantasy malpractice will be committed in the name of “common sense”. Instead, I would encourage everyone to develop a little bit of uncommon sense. We must start questioning things that sound intuitively right with the same tenacity we typically reserve for questioning things that sound intuitively wrong. We must engage in a daily struggle against the temptation to shut down our critical faculties in the face of reasonable statements; and make no mistake, it is a daily struggle. When our thinking is naturally lazy, we must begin to think unnaturally.

With that admonition at the top of my mind, I want to unpack that cliche. Do we really want the best team to win?

A Misleading Reflex

My knee-jerk response is "of course" for the same reason all of yours' is-- because we are all disproportionately likely to believe we are the best team. (If not, we're at least more likely to believe we're the best manager.)

Most managers have a league that means more to them than their others. For me, that's my oldest dynasty league. I started it in 2007 with a group of guys I'd been playing fantasy with since 2002. We're wrapping up our 18th season. I won three of the first seven years, but since then, things have been fairly frustrating.

Over the previous ten years, my average all-play winning percentage was 70.1%. I made the playoffs nine times, earned a first-round bye eight times, made six championship games... and won one title. I thought this year might break the slump, but in the semifinals, I set an all-time scoring record... for a team that lost. (I had the 5th-highest score of any team all season but just happened to face the 3rd-highest score.)

Making it more frustrating, there's another manager who has only finished in the Top 3 in either points or all-play record three times (only one of them a first-place finish). He won the title all three years. Another manager only finished in the Top 3 in points or all-play record twice (neither a first-place finish). He won the title both years, plus another year where he was just 5th in points and 6th in all-play. Another manager ranked 2nd-to-last in scoring but made the playoffs thanks to incredible schedule luck and won the whole thing.

Selfishly, I feel like some of those titles should have been mine. Either some of my all-time great teams should have been rewarded for their dominance, or else some of my solid-but-unspectacular teams should have had some lucky runs of their own. It doesn't seem fair.

But the point of fantasy football isn't to be fair; it's to be fun.

Imagining Some Alternatives

We can easily imagine a world where the "best" team was no more likely to win than any other. Every week's matchup is essentially a coin flip with more steps. Nothing you do makes your team more or less likely to win, so there's no point doing anything. In such a world, fantasy football would die out extraordinarily quickly.

We can just as easily imagine a world where the best team always won. Every year, one team would go undefeated. Opposing managers wouldn't even bother setting their lineups because it wouldn't matter, anyway. Dynasty leagues would likely only last for a couple of years before everyone got sick of one manager winning all the time and quit. 90% of fantasy football players would find the hobby brought nothing but frustration.

But what might not be immediately obvious is this world probably sucks for the best manager, too.

If you don't fear the loss, you won't value the win.

— Adam Harstad (@AdamHarstad) December 24, 2024

Fantasy football is exciting because it is uncertain. You hang on every play precisely because any one of them could be the deciding factor. You are invested because the game is in doubt. Taking away that doubt hurts the favorite as much as the underdog.

So it's good for the best team to win more often than not, but also good for games to genuinely be in doubt. But there's one more fly in the ointment.

Mauboussin Ruins Everything

When we think of activities dependent on luck or skill, we can envision everything existing on a spectrum between a coin flip (all luck) and chess (all skill). The "best" coin-flippers in the world will "win" about 50% of the time. By contrast, the best chess players in the world will... also win about 50% of the time. Why? Because the best chess players in the world predominantly play games against the best chess players in the world.

This is known as the Mauboussin paradox of skill, after researcher Michael Mauboussin. The paradox of skill says that in any activity, the closer two participants are in skill, the larger a role luck will play in determining the outcome of any match between them.

Even if we wanted to engineer a world where the best fantasy managers won 80% of the time, what would likely happen is the weaker managers would drop out until the better managers were all in leagues with each other, at which point the paradox of skill would take over and drive their odds of winning closer to 50% again.

So What is the Optimal Chance of Winning?

I don't know-- this isn't a question that is subject to empirical analysis. In many respects, it's not a question with an objective answer. If I were the Czar of Fantasy Football, though, I probably would have set the target mark for the "better team wins" rate much higher than it currently is.

It's a good thing, then, that I'm not the Czar of Fantasy Football; there's every reason to believe that if the current 58% win rate isn't optimal, it's at least fairly close.

I started playing fantasy football more than 20 years ago. At the time, the hobby was viewed as something of a joke, the province of nerds who loved spreadsheets more than sports. Today, fantasy is a multi-billion-dollar industry that has eaten football whole; we're in a world where starting running backs host an award-winning weekly fantasy football podcast and every touchdown pass to a 2nd-string tight end is followed by the announcers discussing the fantasy implications. The Fantasy Sports Gaming Association-- the industry trade group-- estimates that one out of every five American adults plays fantasy sports.

The "typical" fantasy football league is the backdrop of this expansion. "The better team wins 58% of the time" is the context in which America fell in love with the hobby. It's hard to imagine the hobby could have experienced more explosive growth if it had simply used a different default setting.

Instead, the typical league spread like wildfire because the typical league is fun! Yes, it's also frustrating, but there's a fairly strong circumstantial case that this current status quo strikes the ideal balance between rewarding skill and preserving excitement.

What Do We Need Titles For, Anyway?

At the end of the day, I don't need championships to prove whether I'm good at fantasy football or not. (The "eight byes and six title appearances in eleven years" figure is a compelling testament, no matter how many of those titles I actually won.)

And a single league is a terrible measure of talent, anyway. If you have an edge, playing in more leagues will display it much more clearly. If you are playing for money, larger samples give you a better chance at approximating your true return on investment.

As for my home league, my leaguemates are some of my oldest friends. Because they are my friends, I-- of course-- want to destroy their teams and crush their spirits. But I also feel happy that they are able to experience success and always have the hope for more. (Friendship is paradoxical.) I sincerely hope they will continue playing fantasy football with me for another 20 years.

So should the best team win? I want to reject the framing here. Whether the best team wins or loses is of secondary importance-- the best team should have credible reason to fear that any given week might end its season. The worst teams should credibly believe they might be the one to end it. What happens after that is of secondary importance.

Postscript: Let's Say You Want More Than 58%

Fantasy football is not a one-size-fits-all industry; every league should be designed to best serve the desires of the members of that league. Maybe 58% is optimal overall, but let's say your league wants to reward quality a little more strongly. What tweaks could you make to do so?

I've seen many propose switching to best ball scoring to benefit the better teams. In best ball, you don't set a weekly lineup but instead automatically get the scores of your highest-scoring players.

Is best ball fun? For some people, yes! (I would note that setting a lineup is a ritual that forces managers to interact with the league every week; if you take away this requirement, some managers might find their engagement declining. Consider the managers before making a switch.)

Does best ball reward the best team? I don't think so. I don't have the dataset to test, but I'd wager the top seed advance rate in best ball leagues is fairly similar to leagues with preset lineups. We all remember the instances where we made a decision that cost us a win; we rarely notice all of the other times when our opponent made a decision that gave us a win.

(Insofar as setting a lineup is a skill, we should expect more-skilled managers to fare better in preset lineup leagues than best ball leagues, overall. Though there's a distinction here between "rewarding the best managers" and "rewarding the best teams".)

But I can think of three tweaks that would make an appreciable impact on higher-seed advance rates.

1. Pool play: instead of seeding teams and having them go head-to-head, put four active playoff teams into a pool and have the top two scorers advance.
2. Two-week playoff matchups: instead of teams winning or losing based on a single week, teams would add their scores together over two weeks, and the higher total would advance.
3. Adding a team's season-to-date point-per-game average to its total. The lower seed can still score an upset, but the higher seed starts out at an advantage to reward it for its season-long success.

Combine all three tweaks, and you'll go a long way toward ensuring the better team wins. Whether that's a good thing or not, I'll leave for you to decide.

In the meantime, to all my readers who are still alive for a championship this weekend, I wish you good luck. May the better team win if you're the better team; may the worse team win otherwise, and may you sweat it out until the final kneel down either way.