Most sports analysis sites grade players. They publish a Top 100 in the off-season, a midseason re-rank, a year-end All-Pro list. The player is the unit. A player is a thing — they get one number, one tier, one ranking. It works because the format has been around since the 1950s, when Pro Football Weekly and Baseball America were doing it on newsprint.
FanVerdicts grades players too. There is a Performance grade on every player page in the league. But the unit we are built around is not the player. It is the transaction.
A transaction is a signing, a trade, an extension, a release, a draft pick. Every one of those events gets its own page on the site, with its own four letter grades. A player who has been in the league ten years has ten or twenty transaction pages — one for the rookie deal, one for the second contract, one for each in-season trade, one for each extension. Each of those pages is graded independently.
This article is about why we built the site around the transaction instead of the player. It is a longer answer than it sounds.
The same player is different deals
Saquon Barkley was a great pick at #2 overall in 2018, and a deeply mediocre value on the final year of his Giants contract, and an excellent value on his Eagles deal in 2024. Three different transactions involving the same player. Three different grades on three different pages.
If you only grade the player, you have to pick one of those answers. You usually pick the most recent one, because that is what the audience cares about today, but in doing so you have erased the other two. The 2018 grade gets memory-holed. The 2023 grade gets memory-holed. The story collapses to "Saquon Barkley is good," when the actual story is "the same player has been a great deal and a bad deal at different points, depending on the contract structure and the team context."
Grading at the transaction level keeps every one of those answers on the record. You can pull up the player's profile on FanVerdicts and see all of their transaction grades laid out in chronological order. The third deal does not overwrite the first.
This matters most for the Contract Value Index, because contracts compound. A player's fourth NFL deal is being graded partly against the trajectory their first three deals established. We literally feed prior contract performance into the formula as an input. That input would not exist if we only graded the player.
GMs make transactions, not players
The other reason transactions are the right unit is that they are the unit of GM accountability.
A GM does not draft a player and then own them forever. A GM signs a deal with a player, on specific terms, at a specific point in the player's career arc, with a specific cap structure. The deal is the GM's decision. The player's subsequent career is partly the deal and partly everything else — coaching, scheme, injuries, locker-room context, luck.
When we publish GM grades on the site, every grade is built from the transactions that GM made during their tenure. Not from how the players they drafted are performing today. From the actual decisions — signings, trades, extensions, picks — taken at the moment the GM made them, graded against the information that was available at the time.
This is closer to how front offices internally evaluate themselves. A GM who drafted a future Hall of Famer at #1 overall does not get to claim the same credit as a GM who took the same Hall of Famer at #50 — the second pick was a much better decision even if the resulting player is identical. Grading at the transaction level captures that.
Sentiment moves on the transaction's clock
Fan and media reaction to a player and fan and media reaction to a specific deal are two different things. They move on different clocks. A trade happens, the sentiment around the deal spikes hard for two weeks, then stabilizes. The player's broader sentiment grade follows a slower trajectory tied to their on-field play.
If we only had a player-level Sentiment grade, that two-week spike would either get drowned out (because the long-running player sentiment is dominated by the bigger sample) or it would dominate unfairly (if we weighted recency heavily enough to capture it). Neither answer is right. The right answer is to grade the transaction's sentiment on the transaction page, with its own window, while the player's overall sentiment lives on the player page with its own window.
Same for the Fan Verdict. Fans voting on the trade are voting on the trade. Fans voting on the player are voting on the player. Pooling them into a single number erases the most interesting signal — when the fan base disagrees with itself between "we like this player" and "we hate this deal."
The cost of getting this wrong
The default in sports media is to talk about a deal for one news cycle and then never grade it again. The deal happens, hot takes get posted, the next deal happens, the cycle moves on. Two years later when the contract turns out to have been a disaster, nobody pulls up the original takes and checks them. The takes were graded once, by tone, on the day they were posted, and they expired into the ether.
We have a different commitment. Every transaction page has a permanent URL. The grades update as the formula updates and as new outcomes come in, but the page does not move and the grade history is preserved. You can pull up a 2019 free-agent signing today and see how the original grade aged. If the grade was wrong, you can see that. If the formula was wrong, you can see how the formula was changed and when.
That is the standard a grading system should be held to. You do not get to publish a take, never revisit it, and call yourself accountable.
The four grades follow from the unit
The four-grade structure on FanVerdicts — Performance, Sentiment, Contract Value Index, Fan Verdict — is the right grade structure precisely because the unit is the transaction. Performance is the player. Sentiment is the public reaction to this specific deal. Contract Value Index is the contract structure of this specific deal. Fan Verdict is the fan vote on this specific deal.
Two of those grades — Sentiment and the Contract Value Index — only really make sense at the transaction level. A "player Sentiment" grade is a moving average of every transaction the player has been part of. A "player Contract Value Index" is a weighted average of every contract they have been on. The transaction-level grades are the primary objects. The player-level grades are derived.
That is the inversion that matters. Most sites treat the player as primary and grade transactions as commentary. FanVerdicts treats the transaction as primary and grades the player as the running aggregate.
The transaction is the unit because the transaction is the decision. Grading decisions is what an accountable system is supposed to do.