Kalshi vs. Polymarket vs. Your Sportsbook: Same Bet, Different Price
By JJ Roelant · Published September 23, 2026
Shohei Ohtani went without a home run in 102 of his 130 games this season. We bet $100 on his Under 0.5 Home Runs every game, at every place that offered it.
Bettorly isn't affiliated with Kalshi, Polymarket, theScore or any sportsbook.
Kalshi and theScore: same 129 games, same 101–28 record. $452 vs. $52. The only difference was the price.
- Exchange results are after each exchange’s current trading fee, buying 30 minutes before first pitch. Before fees: Kalshi +$685, Polymarket US +$187.
- Polymarket US only listed MLB player props from mid-May. Its fee was lower during the season (0.05 through June, 0.06 from July, 0.0695 from Sept. 17). At the fees it charged then, it made about +$61.
- DraftKings and FanDuel never offered the Under on a player home run: 0 of 136,000+ home run prices we collected from them.
- Ohtani’s last game was Sept. 7 (injured list).
Where the vig sits
We compared every MLB player prop where we had a sportsbook price and a Kalshi price within 3 hours of each other, July 24 to Sept. 16. That’s four books (DraftKings, theScore, Hard Rock, BetMGM) and about 500 hitters. “Vig” is how far the book’s price sat above Kalshi’s middle price.
| Market | Usually the favorite | Vig on the Over | Vig on the Under |
|---|---|---|---|
| Hits (0.5) | Over, ~60% | 3.9 pts | 2.9 pts |
| Total bases (1.5) | Under, ~62% | 4.4 pts | 2.4 pts |
| RBIs (0.5) | Under, ~70% | 4.2 pts | 2.8 pts |
| Home runs (0.5), theScore | Under, ~88% | 3.2 pts | 3.5 pts |
- Books charge more on the Over, whether the Over is the favorite (hits) or the long shot (total bases, RBIs).
- Home runs come out about even. DraftKings and FanDuel don’t list the Under there at all.
- Same in the NFL. In 39,511 NFL player prop prices from our app research, Unders were priced at 54.3% and hit 53.9%. Overs were priced at 52.5% and hit only 46.1%.
Example: Kyle Schwarber, July 25 vs. Yankees. Hits, where the Over is the favorite.
| Over 0.5 hits | Under 0.5 hits | |
|---|---|---|
| DraftKings price | −191 | +143 |
| What it implies | 65.6% | 41.2% |
| Kalshi's middle price (61¢ / 62¢) | 61.5% | 38.5% |
| DraftKings' vig | 4.1 pts | 2.7 pts |
| Winning $100 pays at DraftKings | $52 | $143 |
| Winning $100 pays on Kalshi, after fee | $59 | $152 |
Kalshi’s quote was within 10 minutes of DraftKings’. Schwarber got a hit.
How much more Kalshi pays
Same 125,000 sides (every Over and Under from the table above). For each one we compared what a winning $100 pays at the book with what it pays on Kalshi after buying at the asking price and paying the fee.
| How likely | $100 win pays at book | On Kalshi after fee | Kalshi pays more | Kalshi better on |
|---|---|---|---|---|
| Under 20% (long shot) | $626 | $809 | +26% | 98% |
| 20–40% | $184 | $207 | +12% | 94% |
| 40–60% (coin flip) | $91 | $98 | +7% | 86% |
| 60–80% | $42 | $44 | +6% | 73% |
| Over 80% (favorite) | $10.07 | $12.57 | +30% | 94% |
- The biggest gains are at the extremes. Kalshi’s fee shrinks on lopsided prices. The books’ vig doesn’t.
- On a heavy favorite, each win pays so little that a few cents is a big share of it.
- Most of the extreme rows are home run props at theScore (about 88%), because that’s where the lopsided prices are.
By market and side:
| Bet | Book vig | Kalshi cost | Kalshi pays more | Kalshi better on |
|---|---|---|---|---|
| Hits Over | 3.9 pts | 2.4 pts | +10% | 93% |
| Hits Under | 2.9 pts | 2.4 pts | +6% | 83% |
| TB Over | 4.4 pts | 2.4 pts | +13% | 98% |
| TB Under | 2.4 pts | 2.4 pts | +3% | 65% |
| RBIs Over | 4.2 pts | 2.5 pts | +13% | 95% |
| RBIs Under | 2.8 pts | 2.4 pts | +5% | 67% |
| HR Over* | 3.2 pts | 1.3 pts | +26% | 98% |
| HR Under* | 3.5 pts | 1.3 pts | +35% | 98% |
- Overs gain more than Unders, because that’s where the books put the vig.
- Where the book is already close to fair (total bases Unders), the exchange barely helps.
- *Home runs (HR) are theScore only. TB = total bases.
- “Kalshi pays more” is the typical (median) bet. “Kalshi cost” is the fee plus buying at the ask instead of the middle price.
Why the same record paid differently
Back to Ohtani. A favorite wins often but pays little, so what matters is how many wins cover one loss.
| theScore | Kalshi | |
|---|---|---|
| Typical Under price | −370 | 74¢ (about −285) |
| Winning $100 pays | $27 | $33 after fee |
| Wins needed to cover 1 loss | 3.7 | 3.0 |
| Break-even win rate | 79% | 75% |
- Ohtani went 3.6 wins per loss (101–28), or 78%.
- theScore needed 3.7 → about break-even. Kalshi needed 3.0 → profit.
- Kalshi’s Under was cheaper than theScore’s in 128 of 129 games. Polymarket US was cheaper than theScore in 87 of 89.
Vig vs. exchange fees
| Sportsbook | Exchange (Kalshi, Polymarket US) | |
|---|---|---|
| Who sets the price | The book | Other traders |
| How it charges | Vig: a margin built into both sides | A trading fee, separate from the price |
| Can you see it? | Only by adding both sides (they total more than 100%) | Yes, it's a published formula |
Exchange fee = rate × contracts × price × (1 − price)
- Kalshi: rate 0.07, rounded up to the next cent once per order. Posting your own price and waiting is usually free.
- Polymarket US: rate 0.0695 (since Sept. 17). Posting your own price and waiting earns a rebate (rate 0.0125). More than $250,000 of volume the previous month gets 10–50% of fees back.
- Buying right away means paying the asking price, usually 1¢ above the bid, so about half a cent over the middle. That isn’t a fee or vig. It’s what the other trader asks.
| Contract price | Kalshi fee per contract | Polymarket US fee per contract |
|---|---|---|
| 50¢ (coin flip, biggest fee) | 1.75¢ | 1.74¢ |
| 75¢ | 1.31¢ | 1.30¢ |
| 90¢ | 0.63¢ | 0.63¢ |
Example: $100 on a 75¢ contract = about 133 contracts × 1.31¢ = $1.75 in fees on Kalshi.
Ohtani’s Under, cost above a fair price (the exchange’s middle price), on the 89 games all three listed:
| Vig | Half-cent to buy now | Fee | Total | |
|---|---|---|---|---|
| theScore | 4.6 pts | — | — | 4.6 pts |
| Kalshi | — | 0.6 pts | 1.3 pts | 1.9 pts |
| Polymarket US | — | 0.8 pts | 1.3 pts | 2.1 pts |
A winning $100 on the Under paid $27 at theScore, $33 on Kalshi and $33 on Polymarket US, after fees.
Reading the vig from the odds
Mike Trout, Aug. 1 vs. Brewers. theScore: Under 0.5 Home Runs −500, Over +325. The two sides imply 83.3% + 23.5% = 106.9%. The extra 6.9% is the vig.
Taking it back out is called de-vigging. The two common methods disagree:
| Method | Fair Under | Fair Over |
|---|---|---|
| Take the vig out evenly | 78.0% (−354) | 22.0% (+354) |
| Power method: take more off the long shot | 81.1% (−428) | 18.9% (+428) |
| Kalshi's middle price (79¢ / 80¢, within 20 min) | 79.5% (−388) | 20.5% (+388) |
Kalshi landed right between the two. theScore’s vig that day: 3.8 points on the Under, 3.0 on the Over. Trout didn’t homer.
Which fair price was right? Across all 4,313 home run props, and on the other markets:
| Predicted | Actually hit | |
|---|---|---|
| Home run Under, vig taken out evenly | 85.7% | 87.7% |
| Home run Under, power method | 89.8% | |
| Home run Under, Kalshi | 88.0% | |
| Hits Over, Kalshi | 60.2% | 60.0% |
| Total bases Over, Kalshi | 37.8% | 37.2% |
| RBIs Over, Kalshi | 29.8% | 30.4% |
- Both de-vig formulas missed by about 2 points, in opposite directions.
- Kalshi was within 0.6 points on every market.
- If you de-vig a book’s odds yourself, you’re guessing how it split the vig. An exchange price takes the guess out.
This isn’t a reason to bet any Over or Under. It’s a reason to check the price in more than one place.
Before your next bet
- How often has this hit? That’s the hit rate.
- What does the price need? Turn the odds into a break-even rate. −370 needs about 79%. −285 needs about 74%.
- Is there a cheaper price? Compare the book with an exchange’s middle price. The gap is what you’re paying.
The MLB props pages show how often each market has gone Over or Under. The parlay analyzer prices a full set of legs.
Bettorly is an analytics tool — we don’t take bets and don’t sell picks. Historical frequencies never guarantee a future result. 21+. Gambling problem? Call or text 1-800-GAMBLER.
Try the parlay analyzer →