Cowboys-Giants Betting Recap: Two Drive-4 Props at +1600 and +370

New York beat Dallas 28-20 on Sunday night, September 13, 2026, at MetLife Stadium. The Giants led at every break and never trailed. The number that mattered on our side was +1600 — a single Giants drive in the third quarter produced two winning receiver props, one at 16-to-1 and one at +370, both placed live while the game was still 14-7.

Sunday was the biggest live betting day of our season so far. FanDuel's own activity statement closes it at $126,888 wagered against $143,019.78 returned, a profit of $16,131.78. At $100 a unit that is +161 units across the slate.

This article is about one drive of it.

What Happened

The Giants scored first and led wire to wire. Dallas never got closer than a possession in the second half. It was a clean, unremarkable divisional game in every way except the one that paid us.

The Drive That Paid Twice

Our two tickets are both on Giants Drive 4, placed live in the third quarter with New York ahead 14-7.

Sixteen-to-one and nearly four-to-one, on the same possession, from a combined $176 of risk.

The two bets are not independent guesses. They are one read expressed twice. The Giants were going to move this drive through the air and through their secondary targets rather than their primary ones, and there were two ways to get paid for that. Both landed.

That is the difference between betting a game and betting a drive. The game markets on this one were boring. Drive-level markets were not, because nobody has priced them properly yet.

Why Drive Props Are Priced So Loosely

Pregame markets are the most efficient consumer markets in American sports. Millions of dollars and every professional syndicate push a spread or a total toward its true probability days before kickoff. Winning against a closing line consistently is genuinely difficult, which is the honest reason most football betting picks sold before kickoff do not clear the vig over a long sample.

In-game drive markets are the opposite. A book has to price hundreds of them per game, in seconds, with no time for the market to correct anything and no trader reviewing each number. The model leans on generic distributions: how often does a receiver at this depth chart position catch a pass on a given drive, how often does this offense throw on second and short.

What the model cannot see is the thing you can see by watching — that a defense has committed to taking away one target, or that a quarterback has spent the last two drives looking to a specific spot. When you can see the shape of a drive before it happens, a +1600 on two receptions is not a lottery ticket. It is a price the model set without the information you have.

That is the whole thesis behind NFL picks built around live action rather than pregame sides, and it is why six sportsbooks have limited this account.

What This Game Teaches About Football Handicapping

The useful lesson from Sunday is not "bet longshot props." It is that where you bet matters more than who you bet.

Serious football handicapping is a search for markets where your information is better than the price. Against a Sunday night spread, your information is almost certainly worse than the market's — that number has absorbed injury reports, weather, sharp money and public money for six days. Against a drive-level prop generated four seconds ago by a model that has never watched this game, your information can genuinely be better, because you are watching and it is not.

This also reframes what a good win rate looks like. A +1600 winner does not need to hit often to be profitable. It needs to hit more than once in seventeen. If your read on drive shape is even slightly better than the model's generic distribution, that is a large edge on a price that big. Volume matters more than accuracy at those numbers, which is exactly backwards from how most people evaluate sports handicappers.

The practical version: stop asking whether a team will win. Start asking whether the number in front of you is better than the number the situation deserves. Those are different skills, and only the second one pays. It applies the same way to NFL picks against the spread, where the question is never who wins but whether the number is wrong.

The last thing worth saying is about proof. The reason this article shows two screenshots rather than a claimed record is that screenshots can be checked and records cannot. Any sports handicapping service can produce a flattering percentage by choosing which twelve weeks to count. Dated slips with the price and the payout on them are harder to manufacture, which is why we publish those and not a season-long number.

What This Doesn't Prove

Two bets. One drive. That is not evidence of a system.

A +1600 winner is, by definition, something that usually loses. If Likely finishes that drive with one reception instead of two, this article does not exist and the underlying read was identical. Variance over a sample this small is enormous, and anyone showing you a 16-to-1 winner as proof of skill is selling you a story.

The day's full number is the more honest figure: $126,888 wagered, $143,019.78 returned. That is a large sample of bets across an entire Sunday, and it still only describes one Sunday. Past results do not guarantee future performance.

Where to See the Rest

Every documented ticket from Sunday, and from every day this season, is posted with the book, the price and the payout on our results page. Saturday's college slips are being added shortly. This week's live card is on best bets today.

Frequently Asked Questions

What was the final score of Cowboys vs Giants on September 13, 2026?

New York beat Dallas 28-20 at MetLife Stadium on Sunday night. The Giants led at every quarter break — 7-0, 14-7, 21-14 and 28-20 — and never trailed. Combined scoring was 48 points.

What is a drive prop in live betting?

A bet on what happens during one specific offensive possession rather than the whole game. Examples include whether a named receiver catches a pass on that drive, how many yards a quarterback throws for on it, or whether the drive ends in a field goal attempt. They are priced by automated models in seconds during the broadcast.

Why are live drive props priced loosely?

Because a sportsbook has to generate hundreds of them per game in real time, with no market correction and no human review. The model uses generic distributions and cannot see situational information that a person watching the game can see. That gap between the model's generic price and the actual situation is the edge.

Are football handicappers worth paying for?

Some are, most are not. The test is not a claimed win rate, because any service can build one by choosing a flattering window. The test is whether they publish individual dated tickets showing the real price they got, whether they show losing bets alongside winners, and whether they explain the reasoning. A service that will not show you its losses is showing you marketing.

How much did The Best Bet on Sports win on September 13, 2026?

FanDuel's activity statement for that day shows $126,888.00 wagered and $143,019.78 returned, a profit of $16,131.78, which is roughly 161 units at $100 a unit. The account balance shown on that statement falls because $22,240 was withdrawn the same day. Withdrawals are money moved to a bank, not losses.

How fast do you have to act on a live bet?

Usually under a minute, often much less. Both of these tickets went in during a single third-quarter possession and settled minutes later. That is why alerts go out by Email and Discord simultaneously rather than in a daily email.

How do I get these picks as they happen?

Packages start at $39.95 a month and are listed on the packages page. Every alert names the side, the price at dispatch and the unit size, delivered by Email and Discord. If you want to see how an alert actually arrives before paying, you can reserve a free live pick first.