The 4-Step Kalshi Prop-Edge Workflow For DFS Players

Updated August 14, 2026 · 8 min read · by Jake Hari

The Quick Answer

If you already build DFS projections, you are most of the way to trading player-prop contracts on Kalshi. The workflow is four steps: start with a projection, convert it to a probability, compare that probability to the contract price, then apply a fee-and-spread test before you size anything. The last step is the one most people skip, and it is the reason a "5-point edge" so often loses money. Below we run all four steps on tonight's real Kalshi pitcher-strikeout board, show the full model-by-model panel, and mark exactly which props clear the test and which get thrown back.

Free: The Weekly PM Market Brief — the 8-model panel's graded record, the week's biggest market-vs-model gaps, and what's spiking next. One email, Sundays. Sign up in the box at the end of this article.

New to event contracts? Start with prediction markets for DFS players for the one-page primer, then come back here for the workflow.

Note on the prices below: they were fetched August 14, 2026 at 8:22 a.m. ET, hours before first pitch and on a thin market-maker book. They will move as real volume arrives near game time, which is itself part of the lesson below.

Step 1: Start With A Projection

A prop edge has to start with a number that is not the market's. As a DFS player you already generate one every day. The Stokastic Sims project every pitcher's strikeout total for tonight's slate, the same engine behind our live Kalshi MLB home run edge board, where the Sims' home run probabilities run against Kalshi prices in public.

To keep every number on this page reproducible, our worked example below uses two open projection engines instead of a single proprietary line: a plain Poisson model built from the pitcher's own 2026 game log, and our price-blind eight-model panel. The step is identical when you hold a Sims strikeout projection instead. What matters is that the projection is yours and is built before you look at the price.

Step 2: Convert The Projection To A Probability

Kalshi prices are already probabilities. A contract that trades at 51 cents is the market saying "about 51 percent." So your projection has to become a probability too, or you are comparing apples to a price.

Strikeouts are a count, so a Poisson distribution is the honest first pass. Take the pitcher's expected strikeouts for the night, call it lambda, and the distribution gives you the chance of clearing any threshold. Build lambda from three fetched inputs:

• Strikeout Rate Per Nine Innings — the pitcher's raw bat-missing ability.
• Recent Innings Per Start — a workhorse gets more chances than a five-and-dive arm, so lambda scales with expected length.
• Opponent Lineup Strikeout Rate — measured against the 22.1 percent league average. A whiff-prone lineup pushes lambda up; a contact lineup pulls it down.

That is the same logic the panel runs, only the panel weighs recent form, matchup, and workload as a group and commits to a single number per contract, price-blind.

Step 3: Compare To The Contract Price

Now line your probability up against the market. If your projection says 57 percent and the contract trades at 50, you have a 7-point raw gap. That gap is the headline every naive edge post stops at. It is not your edge.

The one habit that separates sharp prop traders from tourists: never act on the raw gap. The number that matters is what survives the ask and the fee, and it is always smaller than the headline.

Step 4: Apply The Fee-and-Spread Test

This is the step the request for this piece insisted on, because it is where naive edges die.

Two costs sit between the raw gap and real money. First, the spread: you do not buy at the midpoint, you buy at the ask, and on a thin book the ask can sit well above the mid. Second, the fee: Kalshi's taker fee is roughly the price times one minus the price times seven percent, rounded up to the cent, per contract. The formula value peaks at 1.75 cents at a 50-cent price, which rounds up to a 2-cent charge, and shrinks toward the extremes. There is no fee at settlement.

So your all-in cost to hold a YES contract is the ask plus that entry fee, and your real edge is:

real edge = your probability − (ask + fee)

A useful shortcut is the breakeven win rate, your price plus the fee: a 50-cent contract needs to hit 51.75 percent to break even, and a 60-cent contract needs 61.68 percent. If your projection does not clear the breakeven, there is nothing there no matter how big the raw gap looked.

The Worked Example: Sandy Alcantara, 6+ Strikeouts

Alcantara faces Cincinnati tonight. His 2026 log shows 7.04 strikeouts per nine over 25 starts, recent starts averaging 6.65 innings, and the Reds carry a 25.3 percent lineup strikeout rate, well above the 22.1 percent league mark. Both projection engines agree closely: the Poisson model lands at 55 percent to clear 6, and the price-blind panel blends to 56.8 percent.

The contract sits at 49 bid, 51 ask. Run the test:

• Raw gap versus the 50-cent mid: about 7 points. Looks like a play.
• All-in cost: 51 ask plus a 2-cent fee equals 53 cents.
• Real edge: 56.8 minus 53, about 4 points.

The spread and fee ate roughly 44 percent of the raw gap. What looked like a 7-point edge is a 4-point edge, and it sits on a book with almost no volume yet. That is a size-tiny-or-pass, not a hammer. The test did its job.

The Board: Tonight's Strikeout Props Through The Same Workflow

Here is the whole slate we ran, blend versus market, with the fee-and-spread test already applied. "Net edge" is your probability minus the all-in cost on the side the panel favors.

Every Seat's Number

The panel is price-blind: each model saw the same fetched card of game logs and opponent strikeout rates, never the price. Five of the panel's eight model seats priced tonight's board; the other three sat out this run. These are round-two numbers, after each model read the others' anonymized reasoning and could revise.

Every seat on this panel is graded against real market settlements — records to date: Claude Fable 82% on 460 graded calls · Claude Opus 84% on 524 graded calls · Claude Sonnet 83% on 508 graded calls · GLM 81% on 2,318 graded calls · DeepSeek 80% on 2,343 graded calls. Recomputed daily; the full scoreboard is public.

These are model estimates, not predictions of fact and not financial advice.

More live boards from the same panel: Kalshi MLB home run edge board — Elly De La Cruz 1+ HR 17¢ · Top Kalshi MLB picks today — Cardinals to beat the Cubs 37¢ · Kalshi prices into projections, the same math run the other direction. Prices fetched August 14, 2026.

Hottest Prediction Markets Right Now

• 2028 Democratic presidential nominee · $177M traded
• 2027 Pro Football Champion · $59M traded
• 2028 U.S. Presidential Election winner? · $58M traded
• 2028 Republican presidential nominee · $56M traded
• Pro Baseball Champion · $51M traded
• More tech layoffs in 20​26 than in 2025? · $31M traded

Every market above links to our full AI model verdict; browse them all on the Stokastic prediction markets hub, and see how every settled call actually scored on the full graded scoreboard.

Reading The Edges

Three contracts clear the test cleanly. Yoshinobu Yamamoto over 7 strikeouts is the strongest fade: the market prices it a coin flip at 53 cents, but the panel and the Poisson model both land near 31 to 38 percent, because his recent workload has trended shorter and Milwaukee is a middling strikeout lineup. Buying NO costs 47 cents, or 49 all-in after the fee, against a true chance closer to 69 percent, the biggest gap on the board. Gerrit Cole over 6 and Gavin Williams over 7 are the YES-side edges, both elite strikeout arms the thin early market has not fully priced.

Two contracts get thrown back, and the reasons matter more than the plays. Chris Sale over 7 prices at 51 to 52 cents, and the panel blends to 50 percent. After the spread and fee there is nothing on either side. When your projection agrees with the market, the correct action is to pass, not to invent a lean.

The Trap: George Kirby, 5+ Strikeouts

Kirby is the honest cautionary case. The two projection engines disagree. The season-long Poisson model, using his 7.64 strikeouts per nine, likes YES at about 54 percent. The panel, weighting three straight short, low-strikeout starts (3, 3, and 2 strikeouts, none past seven innings), lands at 43 percent and likes NO. When your own methods split like that, the split is not an edge, it is uncertainty. The discipline is to size down hard or pass, not to pick the number that agrees with the market you wish were wrong.

Dark Horses The Panel Won't Dismiss

Beyond the headline edges, a couple of lower-conviction contracts drew a real number worth watching. Alcantara over 6 is the marginal 4-point case above: a live but small edge that only exists because Cincinnati whiffs so much, and one that vanishes if he gets pulled early. And on the deep end of Williams, at least one seat priced his 8-plus strikeout contract high enough that his tail is live if San Diego's contact rate slips; his last six starts include four double-digit strikeout nights. Neither is a core play, but both are the kind of secondary contract a full board hides if you only read the top line.

Where The Panel Changed Its Mind

The revision round moved real numbers.

DeepSeek came in the highest across the board in round one and pulled toward the group after reading the others: its Alcantara number dropped from 68 to 60, its Yamamoto from 42 to 35, and its Williams from 81 to 78, each time citing the same workload and matchup points the other seats had raised. Sonnet moved the other way on Sale, up from 45 to 52, after peers pointed out his elite strikeout rate against a free-swinging Arizona lineup. — DeepSeek and Sonnet

The number that did not move is the tell. On Yamamoto, every seat held near 30 percent through both rounds. There was no reasoning that pried the panel off the fade, which is why it is the board's most confident call.

What Would Change The Panel's Mind

Every number here is a projection, and projections move on events. Watch for these before first pitch:

• A Late Scratch Or An Opener. If a listed starter is pulled or the club announces a bullpen game, every strikeout contract for that arm resets. This is the single biggest source of a number going stale.
• A Lineup Card That Sits The Whiffers. Cincinnati resting a strikeout-prone bat drops Alcantara's lambda and erases the marginal edge; Milwaukee stacking contact hitters strengthens the Yamamoto fade.
• Weather And A Shrinking Pitch Count. A rain-shortened game or a pitcher on a tight leash after a long prior outing caps the innings, which caps the strikeouts, which pushes every "over" down.
• Volume Arriving On The Book. These prices are market-maker quotes with almost no trades behind them. When real money comes in near game time and the spread tightens, re-run the fee-and-spread test on the live ask before acting.

Settlement And Timeline

Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook, and these strikeout contracts settle at 100 cents or zero on the real box score. Fees, spreads, and thin liquidity are real costs; state and platform availability vary. Every number in this piece gets graded in public once the market settles, on the full graded scoreboard. For the reverse of this exercise, reading a market price back into an implied projection, see Kalshi implied probability: prices into projections.

The Bottom Line

The projection is the easy part for a DFS player, and the price comparison is arithmetic. The workflow lives or dies on the fourth step. Tonight the same test that flagged Yamamoto, Cole, and Williams as real edges also told us to pass on Sale, throw back Kirby, and size Alcantara down to almost nothing. That is not the test failing to find plays. That is the test working, because most props are priced about right, and the money is made by acting only on the few that are not, at a size the spread and fee still leave room for.