NBA
Hawks vs Magic Prediction, Odds & Pick (Oct 21, 2026)
Model vs marketThe numbers
| Team | Model | mid | Ask | Edge vs mid | Edge vs ask | DraftKings no-vig |
|---|---|---|---|---|---|---|
| Atlanta Hawks | 45.1% | 45.5¢ | 46¢ | -0.4¢ | -0.9¢ | 45.7% |
| Orlando Magic best side | 54.9% | 54.5¢ | 55¢ | +0.4¢ | -0.1¢ | 54.3% |
Snapshot: Oct 21, 2026, 21:20 UTC+8 (last pregame data pull). Edge = model probability minus price, in cents per $1 share.
SetupMatchup and context
A regular-season game puts the Hawks at Kia Center against the Magic, scheduled for 7:00 PM EDT (07:00 on Oct 22, UTC+8).
Records: Hawks 0-0, Magic 0-0.
Model vs marketWhere the model and the market disagree
The model has the Magic at 54.9%. Polymarket priced them at 54.5¢ at the mid, with the best ask at 55¢. That's +0.4¢ of edge at the mid and -0.1¢ at the ask, and the ask is the number that matters because it's what you would actually pay.
For the Hawks it's the mirror image: 45.1% on the model vs 45.5¢ on the market.
The market had roughly $236K of liquidity when we pulled the book (Oct 21, 2026, 21:20 UTC+8). The bar for an NBA lean is 2.5¢ at the ask.
MatchupThe ratings behind the number
The model runs on 2026-27 team ratings. In that data, the Hawks went 0-0 with a +2.0 net rating (114.0 offense, 112.0 defense) and a 0-0 last-10 mark; the Magic went 0-0 with a +0.5 net rating (112.8 offense, 112.3 defense) and a 0-0 last-10 mark.
Put together, it projects the Magic by 2.7 points, and that includes 3.5 points of home court.
Big caveat: the NBA model has no idea who suits up. It's ratings in, probability out.
BooksWhat the books say
DraftKings had ORL -2.5 with the total at 234.5. Strip the vig and that's 54.3% for the Magic. ESPN's matchup predictor had the Magic at 52.7% and the Hawks at 47.3%.
The model and the books are basically in the same place on this one.
AvailabilityAvailability report
For the Hawks, ESPN had Henri Veesaar (knee, out) and Mouhamed Gueye (foot, out).
For the Magic, ESPN had Izaiyah Nelson (ankle, out), Noah Penda (hamstring, out), Anthony Black (ankle, out), and Jonathan Isaac (knee, out).
VerdictThe call
No lean. The best the model could find was -0.1¢ on the Magic at the ask, short of the 2.5¢ needed. Passing on thin edges is part of the process, not a dodge.
MoreRelated pages
SourcesWhere these numbers come from
- Polymarket market: order book (bid, ask, mid) and liquidity
- ESPN game page: schedule, venue, injuries, DraftKings line
- FOOFIGHTER model: last-season team ratings, not lineup-aware
- Original model log post on X: published 2026-10-21