Fantasy Football Trade Calculator

How to read the number

Every asset is scored on a 0 to 10,000 scale, where 10,000 is the most valuable asset in the pool under your settings. The number is not a ranking. It measures how much better a player is than the man you would have to start instead of him, which is why your settings move it so much: in an 8-team league the waiver wire is stacked and the alternative is good, so nobody looks special. In a 16-team superflex league the alternative is grim and everything you own is precious.

A gap under 5% is noise and the tool says so. Under 10%, it is close enough that your own roster should break the tie. Anything past 30% is the kind of offer you send once and do not mention again.

The formula, in full

Most calculators call their model a secret sauce. KeepTradeCut uses that exact phrase. Here is ours, in the order it runs:

raw       = fantasy points, scored from component stats
            under YOUR scoring (PPR, TE premium)

surplus   = raw − replacement(position | your settings)

base      = max(0, surplus) ÷ best surplus in the pool × 10,000

value     = base × ageFactor(age, position)

              where, over the seasons ahead,
              ageFactor ∝ Σ  0.85^t × produces(age+t) × plays(age…age+t)
                                      ↑ if he plays    ↑ chance he still is

side      = Σ  value(i) × λ^(i−1)     assets sorted best first

raw is computed in your browser, not baked in at build time, and that distinction is the difference between a settings panel that works and one that is decoration. Receptions are worth whatever your league says they are worth, a tight-end premium is added on top, and every downstream number moves with it. The rest of the scoring is fixed at the near-universal standard — four points for a passing touchdown, one per 25 passing yards, one per 10 rushing or receiving yards, six for a rushing or receiving score — because effectively every league agrees on those, and receptions are a setting precisely because no two leagues do.

Replacement is the last player at that position who is rostered in a league your size, not the last one who starts. That distinction matters for a trade tool: the real question is what you could get for free instead, and the answer sits behind everyone's bench. The flex is not assumed — the league's starting lineups are simulated, locking in the top starters at each position and then letting the best players left fill the flex spots. Full PPR pushes receivers into the flex and standard scoring pulls running backs back in, with no hand-tuned constant anywhere. Replacement is then capped at a common baseline for every flex-eligible position, because any of them can fill a flex slot, so none of them can be held to a tougher bar than the last rostered flex player. What survives the cap is real scarcity: tight end, where the last rostered tight end genuinely is worse than the last rostered flex player, and where the premium is therefore earned rather than invented.

ageFactor compares expected production at the player's next-season age against his age when he produced the stats we measured. In redraft it looks one year ahead and barely moves anything. Switch to dynasty and it averages the next six seasons at a 15% annual discount, which is where a 23-year-old and a 31-year-old stop being worth the same. It is two things multiplied — what he produces if he plays, and the chance he is still playing at all — and both are derived from the data rather than asserted. See below, because it is the part most likely to surprise you.

Superflex is not a multiplier

This is the setting people get wrong, and it is the single most common reason a fantasy football calculator gets called broken. Turn superflex on here and quarterback values roughly double. Nothing is multiplied to make that happen.

In a 1QB league your team starts one quarterback. Twelve teams start twelve of them, and once you allow for the backups people roster, the man you could pick up for free is somewhere around the 23rd-best quarterback in the league — who is still a genuine NFL starter putting up genuine points. So even an elite quarterback's surplus is thin, because you are not replacing his points, you are replacing the gap between him and someone you can have for nothing. That is the whole reason the best quarterback in football is not the most valuable asset in a 1QB dynasty league despite scoring more fantasy points than anyone alive.

Flip superflex on and the league starts roughly two quarterbacks per team. Now replacement is somewhere around the 45th-best quarterback — a backup who barely plays and scores accordingly. The gap between an elite quarterback and that man is enormous, so his surplus explodes. The slot count changed; the maths did the rest. That is why this tool will not let you treat superflex as a cosmetic toggle, and why the verdict names the actual replacement rank it used rather than a rounder number that reads well.

Why three good players do not equal one great one

This is the part every other calculator gets wrong, and it is the reason people screenshot them to complain. Add the players on each side together and three decent starters will out-value a superstar every time. Managers know that is false, which is why they stop trusting the tool.

The fix is the last line of the formula. Assets on a side are sorted best first and each one after the first is discounted by λ for each place down the list, so in a 12-team league the second arrives at 75% of its standalone value, the third at 56%, the fourth at 42%. Nobody joins your roster for free. The second player in a deal has to displace whoever was in that slot, and the fourth is riding your bench.

λ moves with league size, because the man being displaced moves with league size. In an 8-team league there is real talent on waivers and the depth you receive is nearly free to replace, so consolidating into one star is strongly favoured and λ drops to 0.66. In a 16-team league replacement is dire and depth genuinely counts, so λ rises to 0.85. These sit below the equivalents in our fantasy baseball trade analyzer on purpose: a fantasy football team starts about nine players and carries a short bench, while a baseball team starts twenty-three. Football rosters run out of room faster, so football punishes quantity harder. Copying the same constant across both sports would have been a copied number rather than a modelled one.

The age curve, and the running-back cliff

Football has no single aging curve, and a tool that uses one will misprice every dynasty trade you put through it. A 28-year-old running back and a 28-year-old quarterback are not remotely the same asset. So the curve here is per position, and it is derived from nflverse production by the delta method: compare the same player in consecutive seasons, average the year-over-year change at each age, then chain those changes together. Comparing different players at different ages instead would measure survivorship — the only 34-year-olds still playing are the good ones, which makes aging look free.

What comes out is the story every dynasty manager already believes, confirmed without being assumed. Running backs peak around 24 and fall off a cliff: by 30 the curve has them at roughly 65% of peak, by 31 under half. Receivers peak around 25 to 26 and decline gently from 28. Tight ends develop late and hold on longest. Quarterbacks are almost flat from 23 to 33 — about 7% of variation across a decade — which is why an ageing quarterback in superflex stays expensive here while an ageing running back does not. None of those numbers were typed in. They fell out of 15 seasons of play.

The half of aging that curve cannot see

There is a hole in the method above, and it is worth showing you rather than hiding. Comparing a player against himself a year later only ever describes men who were still playing. It is production conditional on being in the league, and it is structurally blind to the single most important fact about a 38-year-old: that he is about to retire.

Left alone, that blindness is not academic. It priced a 38-year-old quarterback at an age factor of exactly 1.000 — five more seasons at 74% of peak, at age 43 — and floated him to sixth on the superflex board. So a second curve is derived from the same data: the chance that a player who is a real contributor at this age is still a real contributor next season. Dynasty multiplies the two together and compounds the survival across the horizon, because to produce in year three he has to have survived years one and two as well. That single term is what finally separates a 24-year-old from a 34-year-old, and it is why an ageing quarterback stays expensive here while an ageing running back falls off the board.

Past the oldest age at which anyone has ever played the position, survival is zero — not "the last value, held forever", which is what a clamp silently asserts and which is how the 38-year-old became immortal in the first place. That cutoff is not a number we chose. The data supplies it, and it lands in a different place for every position: running backs are credited to 37 and quarterbacks all the way to 45 — because across 15 seasons, a quarterback played there and no running back did.

How rookie picks are priced

In dynasty, picks are half the trades, so they are real assets here rather than an afterthought — and their value is derived from what picks have actually returned, not from a chart someone eyeballed.

Every skill player drafted between 2012 and 2022 — 866 of them — is scored on what he produced in his first 3 NFL seasons, under your scoring settings, against your replacement levels. The ones who never played carry zeros, and that is the point: only 57% of them were still producing three years later, and a curve built from the players who hit would price 1.01 like a guaranteed star. Each draft class is then ordered the way a manager would have ordered it on the night, using only what he knew — position and NFL draft capital, never what the player went on to do. Ordering by what actually happened would be hindsight, and it would value 1.01 as "the best player in the class", which is not what 1.01 buys. The value of pick k is what the player at board slot k actually returned, averaged across 11 classes, smoothed across neighbouring slots and forced to never rise.

Because the rookies are scored under your settings, the pick curve responds to them. Turn superflex on and every quarterback in every historical class becomes more valuable, quarterbacks climb the draft board, and early picks — which is where the quarterbacks are — get more expensive. That is the real superflex rookie market, and it falls out of the model rather than being pasted on top of it.

Where the values come from, and where they do not

These values are our own, computed from nflverse production data. They are not KeepTradeCut values, not Dynatyze values, and not derived from FantasyPros consensus rankings. Where this tool disagrees with the crowd, it is disagreeing on purpose — the formula that produced the number is published below in full.

That claim is worth spelling out, because this category is more incestuous than it looks. DynastyProcess, which ranks second for the head term, publishes its formula openly: its values are the FantasyPros expert consensus ranking put through an exponential decay. Dynasty Daddy offers FantasyPros as a source. A good deal of the apparently independent competition is the same consensus ranking wearing a different hat. There is nothing wrong with that as analysis, but it means those tools mostly agree with each other by construction rather than by corroboration.

We took the harder path on purpose, and it has a cost you should know about: an independent model disagrees with the consensus, and disagreement reads as error until you can see why. That is exactly why the formula is on this page. If a number here looks wrong to you, the reasoning underneath the verdict will name the assumption doing the work — the replacement rank, the age curve, the λ. Then you can decide whether you disagree with the model or with reality. That is the entire point of publishing it, and it is the one thing a secret sauce cannot offer.

What this tool does not know

It runs on 2025 and prior production, not on 2026 projections. It has no injury data, so a running back who finished the year hurt still looks like his stat line. It has no depth charts, so a receiver who just lost his job still looks employed. Kickers and defences are absent because nobody trades them and pretending to value them would be padding.

The biggest gap is rookies. Players from the 2026 draft class have not taken an NFL snap, so there is nothing factual to value them on and they are not in the pool — 2027 rookie picks are tradeable here, but this year's rookies are not yet. That is a real limitation and it is worth knowing before you trust a verdict on a rebuilding roster. Every value is also drawn from a pool of skill players on 2026 rosters, so a player who is currently unsigned will not appear.

None of that makes the verdict useless. It makes it a starting point for the argument rather than the end of it, and a tool that tells you where it is blind is worth more than one that does not.

Settled the argument? Copy the link under any verdict and paste it into your league chat. It carries the whole trade and your settings, so the other manager sees exactly what you saw instead of a screenshot. Then go and lose a different argument in 20-0, where you draft nine players and chase a perfect NFL season, playoffs included.

Frequently asked questions

How accurate are trade analyzers?

Accurate enough to rank players sensibly, not accurate enough to trust on close calls. Every analyzer is built on past production plus assumptions about the future, and this one is explicit about both: it runs on 2025 and prior records and knows nothing about injuries, holdouts or depth-chart moves since. Treat any gap under roughly 10% as a coin flip. The gap is printed with every verdict so you can see which side of that line you are on.

Should I trust a fantasy trade calculator?

Trust it to catch the badly one-sided offers, which it does reliably, and to settle the arguments where both managers are guessing. Do not trust it to know your roster. It cannot see that you are already three deep at running back, that your league trades picks like currency, or that you are 1-5 and should be selling. The useful habit is to read the reasoning rather than the number: if you disagree with a verdict here, the disagreement is almost always in an assumption the page has already shown you.

Is there an AI to help with fantasy football?

For trade maths, you do not want one. General-purpose chatbots invent statistics and cannot check their own arithmetic, which is the one thing that matters when you are pricing a deal. This calculator applies a fixed, published formula to real nflverse records, so the same trade always returns the same answer and you can audit every step of it. Use an AI for waiver-wire chatter and start/sit opinions, not valuation.

Where the data comes from

Player names, birth dates, rosters, draft slots and game production all come from nflverse, under CC BY 4.0. The values are our own model, described above in full. Ages and experience are calculated fresh on every page load rather than stored, so they never drift. The pool covers 587 skill players on 2026 rosters with NFL production, and the pick curve rests on 866 drafted players across 11 classes.

Player data from nflverse (nflverse-data), used under the Creative Commons Attribution 4.0 International licence. nflverse does not endorse this tool.

SportsMonkie is an independent project and is not affiliated with, endorsed by, or sponsored by the National Football League or any NFL team. Player names and factual statistics are reported for informational purposes. No player images or team marks are used.