Welcome to the next installment of the Get Better At Dynasty series. In Part III, we talked about profiling your league mates’ rosters. In this part, we’re going to discuss what we think about when we think about drafting. The dynasty rookie draft is comfortably the most fun part about being a dynasty manager, but despite the thousands of hours dynasty punditry has spent putting out content about it, drafting can still feel like a mystery.
How can we know before clicking “Draft” that Michael Mayer is destined to flounder while Brock Bowers will put up a TE1 overall season on the same team? While I have as much of a chance of answering that question as I do of discovering magnetic monopoles, I can at least organize your thoughts so that the hours you spend thinking about the draft will get you ahead of your league mates.
Expected Value vs Upside Variance
How do we really know when to make a pick vs when to trade it for a real player? When dynasty pundits discuss this topic, they often probe around the answer without getting to the center of it. Without burying the lede, the real way to think about drafting is according to these two rules:
1. When you trade a draft pick for a player, you are selling upside variance in exchange for expected value
2. When you make a draft pick, you are forfeiting expected value in exchange for upside variance
If you take away nothing else from this article, remember these two points, and you will find yourself thinking much more strategically about what to do with your assets. Every piece of wisdom you have ever heard about drafting can be reframed into this paradigm. Without getting too technical, a player’s expected value can be loosely defined as how many fantasy points they can be expected to score on average. A player’s upside variance can be thought of as their ceiling outcomes, along with the (low) probability they actually hit their ceiling. We can make two assumptions about expected value and upside variance that hold for most cases:
1. Draft picks have lower expected value than most established players
2. Draft picks have higher upside variance than the tier of players an owner would trade one-for-one for that pick
Let’s say you’re considering trading a high draft pick for an established WR2 type who you don’t think has a WR1 ceiling. When you’re on the clock, you’re choosing between making this trade or picking Luther Burden with the 1.05. What you’re really thinking about is which of these probability distributions you want. (Let’s agree not to quibble about the exact percentages I chose. If I could get this exactly right, I’d be in Vegas or on a yacht.)
As is depicted by these tables, you’re choosing between a player with a higher average outcome vs a player with more uncertainty. The uncertainty means you’re getting a higher chance at lucking into a massive ceiling, but also a much higher chance of getting nothing.
Reframing Established Dynasty Wisdom
If you’ve been playing dynasty for any extended amount of time, you’ve heard lots and lots of punditry about how dynasty rookie draft picks are overvalued and how you should trade them for established football players. This leads us to ask an important question. Given this wisdom is frequently correct, how and why do circumstances arise such that we should still be making draft picks?
First, let’s acknowledge why this advice often holds up, namely that dynasty owners routinely overestimate how certain they can be that a prospect will become an established contributor. And in many cases, they mistake upside variance for expected value. Any pundit or analyst who has ever looked at hit rates for prospects has found they’re somewhat depressing. Objective analysis of dynasty rookie draft pick value needs to begin by admitting that most draft picks will bust.
Even in the top five picks, you’re relatively safe from getting a bum, but you should be prepared not to end up with the game-changer you were hoping for. Now that we’ve acknowled the wisdom in this logic, we arrive at three reasons to still make draft picks.
1. No owner has enough trade chips to pay retail price for every starter on a contending roster
2. There are diminishing returns to adding mid-tier expected value to a roster
3. Our discussion so far has ignored nuances related to timing
The Retail Price of a Stud
If your goal in dynasty fantasy football was to maximize your average finish in the standings every single season, you would essentially never make a draft pick. Like this, you’d probably finish in the top 40% of your league every year. However, in fantasy your goal is to win a championship. And when you look at the rosters of your leagues’ champions, you’ll realize you cannot possibly acquire as many studs as they have just by paying retail for every single one.
Let’s take a look at the rosters of the champions in a couple of my dynasty leagues (12 team superflex, TE premium), and list out how many mid or late 2025 firsts it would cost to acquire their core players according to KTC.
From these two rosters, it’s pretty clear that...
1. You cannot make these rosters in a single startup draft
2. You cannot make these rosters from scratch just by paying retail price for all of their studs. It would take you 13-18 first round picks.
Let’s pretend you’re a rebuilder with a ton of draft capital and only one or two studs on your roster. You have no shot at trading all your draft capital away and acquiring a team as strong as these ones. We’ve now established two points:
1. League champions have a disproportionate share of Untouchable and Stud players on their rosters
2. You cannot acquire a disproportionate share of these players by paying retail price for all of them
This brings us to the reason we make dynasty rookie draft picks. Draft picks are your best path to acquiring a stud player before they cost as much as studs do. Every single player you draft costs you a first rounder or less, but every year some of them become worth 3+ first rounders after their rookie seasons. Assuming a sharp league, whenever you trade away a draft pick for a player, you are acquiring a player who is worth one draft pick. But the owners who stayed put and drafted Brock Bowers or Jayden Daniels ended up paying a single draft pick for a player who is worth 3+ base first rounders.
Generally, the best contenders own a disproportionate share of the expected value in their league, while the best rebuilders own a disproportionate share of the upside variance in their league, which they acquire by giving away expected value and getting back Strong Stashes or draft picks. This logic leads us to my preferred formula for rebuilding teams.
1. Hold on to any Untouchable players who aren’t imminently aging out of those tiers in the next two years.
2. Hold on to any Stud players who might organically develop into Untouchable players.
3. Add as much upside variance to your team as possible by making lots of draft picks all over the draft. Zealously clear cloggers off your roster in order to have space for more stashes and picks.
4. Find chances to trade back for future draft capital or flip key starters / rentals for value-insulated assets in order to have ammo for future trades.
5. Once you hit big on a couple draft picks, keep those guys and use your other assets to pay retail price for another one or two elite established players.
Yes, most of your draft picks will bust or become mediocre. That’s natural and hopefully you can re-roll them. But the key here is that if you take enough dart throws, a couple of those picks will turn into Untouchable / Stud players, and you will have paid less than retail price for them. And then once that’s happened, you can use the future draft capital and insulated assets you’ve acquired to pay retail price for one or two more. Just like that, you have a disproportionate share of the game-changers in your league. A key point here is that I’m not arguing never to pay retail price for a difference-maker, but rather that for rebuilders to turn into a contender, they have to smash a couple draft picks first and then can afford to pay retail for elite players with the rest of their assets.
Diminishing Returns of Expected Value
Let’s say you’re managing a fringe contender and you’re considering acquiring a 28 year-old RB who is guaranteed to score like a high-end RB2. How much are you willing to offer for that player in the following three scenarios:
1. You have an RB1 but no startable RB2
2. You already have five RB2s on your roster
3. You could use an RB2 upgrade for depth, but you have an abysmal WR room
My guess would be that for (a), you’d answer your late first or a 2nd rounder plus a little extra. And then for (b) and (c), you’d probably never make an offer. Here we see that expected value doesn’t exist in a vacuum. In your first example, you’re selling upside variance for expected value, and you’re willing to pay full price because 100% of that expected value slides right into your starting lineup. In the second example however, that expected value does not slot into your starting lineup, so you’d rather spend the pick elsewhere or keep it and buy upside variance.
We see that there are diminishing returns to expected value because the number of slots in your starting lineup is capped. There comes a time when a certain amount of expected value, if it’s mid-level enough, doesn’t appeal to you more than rolling the dice and hopefully getting a difference maker.
In the third example, the expected value of that player didn’t change, but there is now an “opportunity cost” to acquiring that player. In economics, any time you make a spending decision, that decision has both an “explicit cost”, i.e. how much you actually spent to acquire something, and an “opportunity cost”, i.e, what alternatives you can no longer pursue with that money (or draft pick) because you spent it elsewhere. In this example, you’d much rather use that late first or early second to find a way to address your WR room.
To further probe the implications of opportunity cost, let’s consider a hypothetical trade where a fringy contender packages 2025 and 2026 late (assumed) firsts for a high-end WR2 type. This move is inarguably a win from the perspective of expected value. Run a bunch of simulations where you make two late first round draft picks in 2025 and 2026, and most of the time neither player turns into a high end WR2 (this area of the draft purportedly has hit rates around 20%, depending on whose analysis you read). However, many of you probably feel uneasy about making this sort of deal, and you aren’t wrong. So how do we put this hesitation into words?
This hypothetical trade costs you flexibility and opportunity cost, without turning a fringy contender into a real contender. Those are draft picks that are crucially important to fringe contenders because they’re ammo in packages that turn into difference makers, or to flexibly add high-performing Key Starters in-season. It makes the fringe contender more brittle (no backup plan to address injury/underperformance, more sense of urgency to make impulsive win now trades), without adding a player in the Untouchable or Studs tier, while the more established contenders maintain a lion’s share of these players. No, two late picks aren’t enough by themselves to add a game changer, but there’s no tiering up without those pieces. While two late picks on the surface don’t carry crazy value, losing them both for one player closes innumerable paths that lead to championship-level pieces.
Clearly, expected value wins aren’t enough when opportunity cost is taken into account. Suffice to say, dynasty owners are rarely in a position to make decisions based solely on the expected value of a draft pick vs the player they could acquire for that pick. There are diminishing returns to mid-level expected value, particularly for contenders looking to get over the hump. And past a certain point, the opportunity cost of adding expected value exceeds the benefits.
Timing Nuances
Our discussion of dynasty rookie draft pick outcomes so far has focused mostly on the logic behind making vs trading a pick for teams looking to win in the upcoming season. However, there is a major element of timing that throws a wrench into draft pick strategy even further.
The most dynamic part of dynasty is that a fantasy point in the current season matters a different amount team-by-team. For somebody who has already traded their 2025 first, every fantasy point in the current year can feel life-or-death, and it might be justifiable to sell out to win a championship. However for rebuilders, wins in the current season can be worthless, or even carry a negative value, since fantasy points can hurt your draft position. In this case, age and timing becomes important in the discussion of whether to trade, make, or acquire a pick.
For the rest of this discussion, let’s assume we’re talking about a rebuilder who tore down their roster and expects not to compete for two years. Once we’ve made that assumption, we need to acknowledge that the expected value on your roster on average will decay over time, while upside variance realizes into expected value over time (the danger here is when the upside variance realizes into less expected value than you hoped). This assumption leads to the following axiom about timing.
Teams that want to win in the future should prioritize acquiring upside variance and insulated assets in the present, and wait to acquire expected value until just before it’s needed. The logic of “just trade a draft pick for a real player” breaks down if a rebuilder wants 2027 fantasy points more than they want 2025 fantasy points. While you can always try to acquire a young player who can be projected not to drop off by 2027, in reality there are so many variables related to player value (not just aging, but also situations getting worse, regression, contract disputes, and injury), that acquiring a player for retail price in spring 2025 and expecting equal fantasy points in winter 2027 on average will water down your 2027 scoring. (At risk of writing a vacuous truth, if you acquire a young player not just hoping that they maintain their fantasy scoring, but because you have reason to believe they’ll improve, you’re buying upside variance not just expected value.)
All this is to say, when deciding whether to trade a draft pick for a real player, teams that want to prioritize a rebuild will value every marginal increase in upside variance much more than they’ll value every marginal increase in expected value, especially because there is no sense of urgency to acquire expected value until later when they’re ready to pivot to competing.
How to Think About Downside Variance (Risk)
You may have noticed that so far I have used the term “upside variance” instead of just “variance”, and the distinction arises because variance has two tails, a downside tail and an upside tail. By “tail” I mean an outcome that is less likely, but possible, and more extreme than the average outcome, good or bad. In the table where I list 2027 WR Fantasy Finishes and their probabilities, the upside tail would be the “High End WR1” and “WR1” finishes, and the downside tail is the “Bum” and “Out of the League” categories.
The reason that so far I have only talked about expected value and upside variance, and not downside variance, is that established players have higher expected value than draft picks primarily because their downside tails are much less probable while a draft pick that you would offer one-for-one for that player is much more boom or bust. So to this point, it saved me a lot of time to talk about only expected value vs upside, since we’re all implicitly familiar with the way downside variance is baked into expected value.
Now that we’re familiar with the definition of “downside variance”, we need to acknowledge that downside variance turns into lost expected value along different timelines for picks vs players. When you acquire a player, that player’s value will rise and fall fluidly every minute of every day, most days barely at all but some days a lot, much like a stock. When you hold a draft pick, that pick’s value will increase steadily until at least the NFL draft, and often through to your dynasty rookie draft when it turns into a rookie player.
This demonstrates that draft picks have “value insulation”. It’s essentially impossible for them to lose value until they turn into a real player, at which point all hell can break loose. Once you hold a rookie player, their value is likely going to be way, way different after their rookie season, either higher or lower, maybe even worthless. Before you make a draft pick, that pick’s value is more insulated than any player. Once it becomes a player, all bets are off, and their value insulation usually depends on the player’s draft capital, how clear their runway to playing time is, and how long a leash they get from their coaches.
Risk of Trading Away a Player
When you trade a player for a draft pick, you are incurring two risks.
1. Expected value loss - The risk that the draft pick flops and you lose your player’s fantasy points in exchange for very few, or even no, future fantasy points.
2. Upside variance loss - The risk that the player you traded away takes a leap and you miss out on it.
When you think about your player’s ceiling, and whether it’s maxed out, you are making a judgement on how much of upside variance loss you think you’re risking. When you trade away a key starter or an older player, you’re primarily exposing yourself to expected value loss. In order to make a trade with another owner, you must value one or both of these risks differently from the other owner. We now end up with a few dynamics:
1. All owners worry about upside variance loss at all times
2. Only teams trying to compete worry about expected value loss, but rebuilers should still try to get fairly compensated for expected value loss, since it’s much easier to measure. Don’t let a contender bully you when he argues that the expected value is worthless on your team. You’re not here to do anybody favors.
3. When you trade a player for draft picks, you should be turning expected value loss into a greater quantity of upside variance.
4. When you tier down, you should acquire much more upside variance than the expected value loss you’re giving up, since not all of it will turn it into fantasy points.
5. You should always prioritize acquiring upside variance when you ask for throw-ins.
6. When you tier up, you should try as hard as possible to minimize the upside variance loss, since if enough of that upside variance turns into expected value, your tier-up was pointless. Instead you should try to offer expected value that has diminishing returns on your roster.
7. Your throw-ins, to the greatest extent possible, should avoid adding upside variance loss. Sometimes you can’t avoid it though.
When you trade away assets, the following assets tend to carry significant upside variance loss (see Part 1 of this series if you’re unfamiliar with the player categories):
1. Draft picks
2. Any young players who are not strong re-rolls or cloggers
3. Strong stashes of any age
4. Your Studs tier
5. Players stuck in bad offensive schemes
6. Players currently behind inferior or old players on the depth chart
7. Handcuffs not currently getting starter workloads
You generally want to avoid offering these types of assets as throw-ins as much as possible. And it’s likewise not too helpful to turn the listed player types into draft picks, since you’re turning upside variance into upside variance. Your focus on throw-ins should be expected value loss that was not making it into your starting lineup. As an example, Tucker Kraft is a great player to acquire as a throw-in, and a bad player to give up as a throw-in, because losing him carries relatively little expected value loss but pretty major upside variance loss.
Risk of Trading Away a Draft Pick
On the other hand, when you trade a pick for a player, your risk is that you paid retail price for an asset, and lost a shot at paying less than retail price for a better player. In this sense, sellers of draft picks lose upside variance but jettison downside variance. When trading away a pick for a player, you expect to lose more upside variance than the expected value you’re getting back, but you want that difference to be as little as possible.
When offering throw-ins in your trade packages, you should try as hard as possible to give up players you consider to have low upside variance, such as young cloggers and strong re-rolls, and to keep your draft picks. However, realistically this can be hard, since everybody wants picks, but not everybody wants your specific clogger.
Liquidity Risk
The last risk we need to consider is liquidity, i.e. how easy or difficult it is to find an acceptable market for a given asset. Because draft picks are insulated and don’t take up a roster spot, they’re extremely liquid. Players, on the other hand, can be extremely illiquid, since you have no way of knowing who in your league will actually be interested in acquiring your specific player. “Liquidity risk” refers to the inability to get fair value for a player, or the need to give up more than fair value to acquire a player. Liquidity risk mostly boils down to two factors.
1. Personal preference, i.e. maybe you wanna flip a specific WR you bought low on, but the only WR-needy contender in your league has that player on his do-not-draft list
2. The endowment effect, i.e. somebody requires more than fair value to sell you an asset, and might not be willing to offer fair value to acquire that exact same asset.
Fairly obviously, any time you sell a draft pick for a player, you take on liquidity risk. Any time you give up a player for a draft pick, you reduce liquidity risk. You should not care much about incurring liquidity risk if that player slots into your starting lineup, either now or in the future. But, you should care a ton about liquidity risk if you’re trying to flip a player, or acquiring a player who doesn’t really fit into your roster just vaguely because “he’s a buy low”.
For this reason, I almost never advocate for trying to flip players. Even if you acquire a player and his value seems to go up, it’s still possible that
1. You had to overpay to pry that player away from his owner
2. You’ll get lowballed when trying to capitalize on that player later
Players with higher liquidity risk include...
1. Older players at any position
2. Running backs
3. Low-end TE2/WR2/RB2 type fringy performers who don’t slide into every starting lineup and whom rebuilders don’t want
Players with lower liquidity risk include
1. Rookies, especially if they have a long leash from their coaching staff
2. QBs in Superflex
3. Young WRs
It’s probably not worth nixing a trade deal just because of the liquidity risk you’re taking on, but you should at least be self-aware about whether your motivation to make a trade might become pointless due to liquidity risk.
Takeaways
A really common pattern I see among dynasty owners is day trading the value on their roster, making superficial changes in the hope of value gains but without committing to a direction or plan for the roster. Players turn into picks, which turn into players, which turn into…you get the idea. To help us break out of this cycle, we’ve built up a better understanding of how to think about the dynasty rookie draft, and how to think about what we’re gaining and giving up when we make a pick or trade a pick. When we’re having trouble deciding between an unknown vs an established player, what is that unsaid something that makes us uneasy about either side? Without realizing it, we implicitly answer this question the wrong way in our draft day decisions.
We make the mistake of treating the draft like it’s gambling, where we can choose something safe or we can get excited about the possibility of something better. However, we should be treating the draft as a path to our goals. Depending on our goal for our roster, sometimes the draft is the best path, but many times it isn’t. Before you roll into the offseason, consider your position along this chart.
Feel free to ignore or use this chart as you see fit, but the key point here is that the benefit of drafting is more complex than “if you’re drafting, you’re gambling.” There is a time when your roster needs more upside variance and a time when your roster needs to get rid of uncertainty and prioritize production. Focus your trading and draft strategy based on where you stand, and you’ll build efficient rosters that rebuild quickly and stay on top consistently.
That’s all for today. Stay tuned for future installments of this series, when we leave the theoretical and dive into the practical, examining how this new perspective on drafting can help us think about common pick trading strategies and scouting. Until next time, stay hungry and stay sharp.