How Our Trade Values Work

Summary

Every value on this site is computed from public game statistics with a fixed formula and the league settings you choose. In short, we transform raw stats into trade values: score each game for your league, compare the player with a replacement-level starter, adjust for age in dynasty, and put everything on one 0–10,000 scale. The steps are below, followed by the data source, what we don't use, the update schedule and known limits.

Model Steps

1. Score every game for your league. We take game statistics from the 2026 season so far and the full 2025 season, and score each game with your settings: Standard, Half PPR or PPR, plus any TE premium.
2. Average points per game. We blend this season's points per game with last season's. Last season counts as if it were 8 extra games, about half a season (fewer if the player played fewer than 8 games last season), so after n games this season its weight is 8 ÷ (8 + n): half after 8 games, about a third by the end of the season. Players with fewer than 4 games in total are also pulled toward the median of their position so that a short hot or cold stretch doesn't swing their value too far.
3. Measure value over replacement. We subtract the points per game of a replacement-level starter at the same position. The replacement level depends on your league size and on how many players start at each position, which is where 1QB and superflex differ. The result is value over replacement (VOR). Instead of cutting VOR off at zero, we smooth it near the line: well above replacement it equals the plain difference, and players at or below replacement keep a small positive value, so a quarterback just under the line in 1QB is worth a little rather than nothing.
4. Dynasty only: add up five seasons. In dynasty, a player's value is this season's VOR in full plus the next four seasons. Each future season is projected from today's VOR along an age curve for the player's position and counted at 0.8 of the season before it, because later seasons are less certain. On the age curve, players below their peak age keep improving by up to 15% (10% for tight ends) until the peak; after the peak output shrinks by the same share every year: about 14% a year for running backs after 26, 11% for wide receivers after 28 and tight ends after 29, and 10% for quarterbacks after 32. A veteran still producing gets full credit for this season; a young player gets more credit for the seasons ahead. Redraft uses this season's VOR only.
5. Value rookies and picks. This year's rookies get a floor from their NFL draft position and are worth the higher of that floor and the value of their own games. For rookie draft picks, the average 1st-round pick is set to the average current value of the top players from each of the last three rookie classes, one per team in your league. Each slot's value then follows a smooth curve by overall pick number that halves every half round (every 6 picks in a 12-team league), so 3rd- and 4th-round picks stay above zero. Slots are averaged into Early, Mid and Late. Picks for the year after next are multiplied by 0.9.
6. Put everything on one scale. All players and picks are normalized to 0–10,000.
7. Discount depth in multi-player trades. On each side, the most valuable piece keeps its full value. Every other piece counts as value × (value ÷ top piece in the whole trade)^0.5, so a pile of bench players can't outweigh a starter.
8. Give a verdict. The gap is the difference between side totals divided by the larger total. A gap of 5% or less is a Fair trade, 5% to 15% is a Slight edge, and more than 15% means one side wins. The gap is shown to one decimal place, on the same side of the 5% and 15% lines as the verdict.

These steps power the fantasy football trade analyzer (redraft), the dynasty trade calculator, the superflex trade calculator and the dynasty trade value chart.

Data Source and Attribution

Player statistics come from nflverse (nflverse-data repository), made available under the CC BY 4.0 license. Draft data (draft year, round and pick) also comes via nflverse, originally from Pro Football Reference. We modify the data: we transform raw stats into trade values using the steps above. nflverse provides the data as-is, and nflverse is not responsible for our values.

Data We Don't Use

Values do not use data from FantasyPros, KeepTradeCut or FantasyCalc. There are no expert rankings, crowd votes or trade records in the model; it uses game statistics, your settings, player age and draft position only.

Update Schedule

In the offseason, values are refreshed at least once a month, plus one update after the NFL Draft at the end of April so that new rookies enter the data. During the season, values change only when we rebuild the data and publish a new version of the site; there is no fixed weekly schedule, so check the date shown. Each tool and the value chart show the date the values were last updated and the last week of stats included.

Known Limits

• No market signal. The model has no trade-market or crowd input, so its numbers will differ from crowd-sourced values, most of all for young players with little production.
• Games already played. Redraft values reflect games already played. Injuries and changes in role, such as losing a starting job, are not included.
• Rookies wait for the draft. New rookies appear after the update that follows the NFL Draft.

Treat every verdict as one reference point, not a final answer.

Change Log

• October 17, 2026: method page published.