Simulation Betting 101: Find Value in NFL Props and Parlays

Seahawks-Broncos kicks off Thursday night in Denver, and as far as Stokastic is concerned, the game has already been played 10,000 times. Every snap, every third-down scramble, every fade that sails out the back of the end zone sits on a server in 10,000 final scores and 10,000 full box scores, like the world's nerdiest DVR.

That's the engine behind Ask the Sims, the search bar at the top of the Kalshi Parlay Simulator. Type any question about any game on the NFL slate, from "do the Broncos win?" to a six-leg monster with a quarterback rushing prop jammed in the middle, and the Sims count every game where it came true, then set that number next to the live price at Kalshi, Polymarket, DraftKings or FanDuel.

A tool is only as good as the person holding it, though. So before we start firing questions at it, let's cover three things: what simulation-based betting actually is, how it differs from the way most of the market sets its prices, and how to use the Sims to find the spots where a price is wrong.

What Is Simulation-Based Betting?

A simulation model answers betting questions by playing the game out. Stokastic's Sims start with the players on the field (snap shares, target shares, how each offense calls plays on early downs, how each defense holds up against the run, etc.) and plays the game out snap-by-snap. Run it once and you get one believable version of Thursday night. Run it 10,000 times and you get the whole range of Thursday nights, from the 13-7 slogfest to the 42-38 shootout.

From there, every probability is a raw count of how often the thing does/does not happen. For instance, Jaxon Smith-Njigba found the end zone in 5,279 of the 10,000 games, so his anytime touchdown is 52.8%, or -112 in sportsbook terms. No adjectives, no "I like him here," but a raw tally with potentially actionable information.

The part that matters most for betting: because every sim produces a full box score, the Sims know which things happen together. In the games where Sam Darnold throws multiple touchdowns, Smith-Njigba is far likelier to be on the receiving end of one. A book pricing those two props on separate pages can't see that relationship, while the Sims see it in every row.

Top-Down vs. Bottom-Up: Two Ways to Price a Football Game

Most of the betting market works top-down. A sportsbook opens a spread and a total, sharp bettors push them around, and the number settles where the money stops moving. Player props get built downstream of that: take the team total implied by the spread and total, split it among the players by their usual shares, shade it for the public's love of overs and post a line. Prediction markets like Kalshi and Polymarket run on different plumbing but land in a similar place, since their traders and market makers anchor to the books and to each other.

That process is excellent on the big numbers. Spreads, totals and moneylines have thousands of people with real money arguing over every half-point, and the price that survives the argument is usually pretty good. It gets lazier the farther you travel from the middle. Alt lines, 2+ touchdown props, quarterback rushing props and especially combos of several legs see a fraction of the money, so their prices come from formulas more than from fights.

The Sims go bottom-up. They start from the players and the play calls, play the game out and count what happened. So the useful question becomes: where do the two approaches disagree, and who's more likely to be right about that specific spot?

Chargers-Chiefs is a clean example. Ask the Sims for the best Patrick Mahomes props on the FanDuel tab and nearly every passing-yards line comes back "fair value." The Sims have him over 224.5 yards at -123, FanDuel charges -130, and nobody's going to be retiring betting on that gap over the long haul. The market is sharp on a franchise quarterback's passing yards because the whole country bets them. Then look at the top of the list: Mahomes anytime touchdown, which the Sims price at +292 (25.5%) and FanDuel hangs at +429 (18.9%). That's more than six percentage points of daylight on a prop that rarely gets the market's full attention, and it's exactly the kind of corner where a bottom-up model earns its keep.

The Sims do have blind spots, and the tool lists them right on the page: no penalties or timeouts, passing volume that runs a little light and more ties than the real NFL produces. Lean on the Sims hardest in the corners where the market is lazy, and give the market more respect on the spots the Sims are less certain about.

How to Use Correlation As a Weapon

When a sportsbook prices a parlay, the starting point is multiplication. Packers win (50%) times Jordan Love 2+ passing touchdowns (61.4%) gets you 30.7%, which is the right price only if those two things had nothing to do with each other.

They obviously do. Ask the Sims how often the Packers win and Love throws 2+ touchdowns on Sunday night, and the answer is 35.9%, 3,586 of the 10,000 games, or +179. The tool labels it "Hit together more," with the 30.7¢ unrelated price sitting right underneath. That five-point gap is the whole reason same-game parlays exist, and it's also why books pile extra margin on them. The Sims show you the true correlated number so you can judge whether the price on your screen sits above or below it. (Small side note from the same answer: the Sims make Green Bay a coin flip at home while Polymarket charges 53.7¢, so the Packers moneyline on its own carries no value.)

Correlation also cuts the other way, and it gets sneaky. Ask who leads the game in receiving yards if the Packers win, and the top two names both play for Dallas: CeeDee Lamb at 81.8 yards a game and George Pickens at 76.0, ahead of every Packer. A team that's losing late throws to catch up, so in the games Green Bay wins, Dak Prescott is slinging it for four quarters. "Packers win + Christian Watson over" points in one direction. Tack a Lamb under onto a Packers win, and the Sims say you're betting against the very game script you just picked.

How to Use Ask the Sims: Five Questions, Simple to Complex

Here's a walkthrough built on Thursday night in Denver, starting with a one-prop question and ending with a six-leg ticket. Prices refresh constantly, and every question you've asked gets recounted against the newest Sims and prices each time you come back. Once the game kicks off, the Sims keep updating live, which comes in handy for anyone watching Thursday night trying to add to their position, hedge their position, or sweat their position directly on screen.

All screenshots from Ask the Sims, Tuesday, October 13, 2026.

#1: "How often does Jaxon Smith-Njigba score a touchdown?"

Start simple: one player, one prop. The answer card gives you three numbers. The Sims have Smith-Njigba scoring in 52.8% of games (-112), Kalshi sells it at 47.7¢ (+110), and the gap is +5.1¢ of edge per contract with the fee already included, since every Kalshi price in the Simulator is fee-in. The card also hands you a walk-away number, telling you the bet is worth buying at 52.8¢ or less.

Keep scrolling, because every answer doubles as a little research memo. You get the alt lines (2+ touchdowns at 15.2% in the Sims, +558, against 12.7¢ at Kalshi, +687), the props that pair with it (Smith-Njigba scoring and Darnold throwing 2+ touchdowns hit together 44% more often than their prices imply) and the other side of the bet, where Smith-Njigba not scoring costs 56.7¢ for something the Sims see 47.2% of the time.

#2: Same question, different book

Flip the tab at the top to DraftKings and ask the exact same question. The Sims number stays put at 52.8%, since the game hasn't changed. The prices do. DraftKings hangs the anytime touchdown at +110, same as Kalshi, but the 2+ touchdown ladder is a different story: DraftKings pays +747, Kalshi pays +687 and the Sims make it +558. Same opinion, same player, about $60 more on a $100 ticket at one shop than the other. Four tabs sit at the top of the page for a reason, and checking all four takes about 10 seconds.

#3: "Best Bo Nix props"

Any "Best ___ props" question returns the lines where the Sims beat the price by the most, filtered to bets that hit at least 20% of the time. For Nix, the top of the list tells a story: YES on 10+ rushing yards (68.7% in the Sims against 63.6¢ at Kalshi), NO on 35+ pass attempts (55.5% against 51.8¢) and NO on 4+ rush attempts (47.7% against 44.7¢).

Read those together and the Sims are describing a specific quarterback, one who picks up his rushing yards on a couple of chunk scrambles rather than a steady diet of designed runs, and who doesn't need to throw 35 times to get through a game. That's a handicap you could spend an hour building from film and game logs, delivered in about one second. (Kalshi won't let pass attempts or rush attempts into a combo, so those two are singles only.)

#4: "Best 3-leg parlay if the Seahawks win"

Now we add an opinion. Tell the Sims what you think happens, and they filter down to the games where it did. The top combo here is Seahawks win + George Holani 10+ rushing yards + RJ Harvey under 25 rushing yards, and the card shows two different prices for it.

"If right" is 59.8¢: in the games Seattle wins, this ticket cashes 59.8% of the time. "Every game" is 35.0¢, which is what the ticket is worth across all 10,000 sims, including the ones where Denver wins and your parlay dies on leg one. Kalshi doesn't know your take, so it prices on every game, with an expected quote of 31.0¢ (+223). Across all 10,000 games, no opinion required, the ticket still grades out a few cents above that quote. The +27.3¢ edge on the card is what your read on Seattle adds if it's right, and that "if" is doing a lot of work.

#5: "Best 6-leg parlay"

Time to go big. Ask for the best six-leg parlay, and the top combo is gloriously boring: Nix 10+ rushing yards, RJ Harvey doesn't score, Tory Horton doesn't score, Emanuel Wilson 40+ rushing yards, Elijah Arroyo doesn't score, under 51.5 total points. The Sims hit it 35.4% of the time against an expected Kalshi quote of 25.1¢ (+299), good for +9.1¢ of edge per contract, fee in.

That's the lesson in one ticket. The best long parlays tend to be six likely outcomes that all lean the same way, in this case toward a game with a modest score where secondary pass-catchers don't find the end zone. The second combo on the list is the swing-for-the-fences version, Smith-Njigba scoring plus five more legs, and it hits 18.1% of the time against an expected 12.0¢ (+736).

Five Ways to Find Value With the Sims

Start in the corners. The market is sharpest on spreads, totals and specific prop markets such as a star quarterback's passing yards. We've found it to be less efficient in markets such as anytime touchdowns, alt lines, quarterback rushing props and combos, sometimes because less money has argued the price into shape and other times because the Sims are seeing the game through a different lens.

Shop every tab. The same Sims number gets compared against Kalshi, Polymarket, DraftKings and FanDuel. A 2+ touchdown prop at +747 and the same prop at +687 are different bets, and the Simulator shows you both.

Let correlation do the lifting, and keep the legs few. A ticket built from legs that move together is worth more than its parts. Every extra leg adds variance, and the Simulator itself notes that fewer legs give steadier prices.

Know what "if right" is assuming. A conditional price is only as good as the condition. Check the "every game" number before you buy, because that's the one Kalshi is pricing.

Check the live quote and respect the "rough" note. "Expected" is the Sims' best guess at what Kalshi's market makers will quote, and the live price can differ. If you narrow the score or play-calling sliders so far that only a handful of games fit, the tool flags the answer as rough, so widen it back out.

Final Thoughts

For years, betting a football game meant picking a side, stacking a parlay on a hunch and hoping the book hadn't already priced your idea in. Ask the Sims flips that order. You bring the question, whether it's one touchdown prop or a six-leg ticket built on the game script you believe in, and 10,000 games come back with the answer and a fair price to hold up against Kalshi, Polymarket, DraftKings and FanDuel in about one second. Every bet starts with a number you can trust and ends with a price you can check, so your gut still gets a vote, just with 10,000 games of evidence sitting next to it.

Be a better bettor. Ask the Sims.

— Lindy

More on this: Cowboys vs. Packers Prediction & Props: Pickens TD at +160 · Seahawks vs. Broncos Prediction & Props: Smith-Njigba TD +110 · Ask The Sims Wins: Winston's 1.7% Ticket Hit · Bills vs. Rams Props: Buffalo by 4+ Pays +240 · Eagles vs. Jaguars Props: Lawrence Hasn't Completed 20 Yet