NFL Prices vs Final Scores: 70-Cent Favorites Won 75%
We checked Friday's NFL exchange prices against the final scores. On one small slate, the prices came close to what happened.
Published 2026-11-02 · Data as of 2026-11-02 · Market & data intelligence · Educational, not advice.
Delta Arc checked 88 resolved NFL lines, priced on Friday, October 30, against ESPN final scores. Moneyline favorites priced at 70 cents on average won 9 of 12 games. In each price range, results landed within a few points of the average price. The biggest upset was Dallas at 77 cents. Late price moves were small. This is one weekend and a small sample, so it proves very little.
A prediction market price is a claim about the future that someone was willing to pay for. After the games are played, you can check those claims. This post does that. It takes the NFL prices posted on Friday, October 30, and compares them with what happened on the field.
First, a reminder of how these markets work. On an exchange like Kalshi or Polymarket, you are not trading against a house. Every contract has a buyer and a seller. The exchange matches them, and may charge fees. The price is roughly where buyers and sellers meet. A sportsbook works differently: it sets its own number and takes the other side of every bet. An exchange price reflects what the crowd thinks, so you can grade it against results.
What was scored
Delta Arc saved 103 NFL prices at 13:51 UTC on October 30, per Delta Arc's saved price snapshots. It then scored them against final scores from ESPN's NFL scoreboard, checked at 14:00 UTC on November 2. At that point 12 games were final, 88 lines had resolved, and 2 games were still to play.
The 88 lines cover several kinds of markets on the same games, including moneylines and totals. These are the market's own prices checked against results. Nothing here is a pick record, and none of it recommends a trade.
Did the higher-priced side win as often as priced?
Start with the simplest question. Per Delta Arc's calculation, Friday's moneyline favorites won 9 of 12 games, or 75 percent. The market had priced them at 70 cents on average, which is about a 70 percent implied probability.
That is close. A 70-cent contract is the market saying an outcome happens about seven times in ten. Over twelve games, it happened nine times. Twelve games is far too few to tell whether that small gap means anything.
The calibration buckets
A better test is calibration. Calibration asks whether a 30-cent outcome happens about 30 percent of the time, and a 75-cent outcome about 75 percent of the time. Delta Arc grouped all 88 resolved lines by their Friday price:
- Priced under 40 cents: happened 3 of 11 times (27 percent), against an average price of 29 cents.
- Priced 40 to 60 cents: happened 31 of 64 times (48 percent), against an average price of 52 cents.
- Priced over 60 cents: happened 10 of 13 times (77 percent), against an average price of 75 cents.
In every group, the share that happened came within a few points of the average price. That is close to what well-calibrated prices would look like, but it is too small a sample to say. Notice where most lines sat: 64 of the 88 were priced between 40 and 60 cents, so most of the board was close to even. The low and high groups each hold only about a dozen lines, so one result can move their percentages by several points.
The biggest miss
Calibration is about averages, and averages include misses. By Friday price, the biggest upset was Dallas to win at home against Arizona. That contract was priced at 77.0 cents, and Dallas lost 33-30. A 77-cent price still gives the other side a bit under a one-in-four chance. When that happens once, the price has not failed. It is behaving the way a 77-cent price should.
Same outcome, different venues
The same game trades on more than one exchange, so one outcome can have slightly different prices at the same moment. Per Delta Arc's calculation, across the 44 winning outcomes, the typical (median) price at the cheapest venue was 1.3 cents lower per $1 payout than at the most expensive venue.
That describes how the market is built. It is not a recommendation. Venues can have different traders, fees and amounts of trading, so their prices drift a little apart.
How much prices moved before kickoff
Prices keep moving between Friday and the game. Delta Arc took a second snapshot at 16:25 UTC on Sunday, November 1. Across the 66 Kalshi lines quoted both times, the median line moved 1.0 cent.
Most lines barely moved. Seven moved 3 cents or more. In 3 of those 7, the outcome the price moved toward happened. That is too few lines to say whether late moves carried useful information.
The biggest move was the Over 40.5 in Kansas City at Denver. It went from 52.0 cents on Friday to 56.0 cents on Sunday, and the over hit. That was the largest move in the sample, and one example says nothing about why it moved.
What one weekend can and cannot say
This tally shows prices that tracked results fairly closely on one slate of games. It does not show that the market is always right, that any price was off, or that any pattern will repeat. Calibration only means something over many weekends.
Each additional slate would add to the sample. Over a season, the groups would either keep lining up with their prices or start to drift.
Sources: Delta Arc saved price snapshots (Delta Arc's calculation: Friday venue prices vs ESPN final scores); ESPN NFL scoreboard.
Delta Arc Prediction Markets is where Delta Arc tracks Kalshi and Polymarket prices. Get in touch.