POR @ GS
Portland Fire · Golden State Valkyries · Saturday, 19 September 2026 at 04:00
CorrectPreview saved before the gameAs of 19 Sept
What Forecast expects · Favourite
84%GSchance to win
GS · 84.4%15.6% · POR
- GS starts with a 84.4% chance to win; POR with 15.6%.
- In games: out of 10 like this one, GS would win 8 and POR 2.
Confidence
medium
- The model sees a clear favorite.
- Research-stage model: shown as what it said before tip-off, not as an advantage.
- The model does not know this game's injuries or starting five.
Why
- The model looks at only two things: each team's accumulated season strength and home court.
- Home court is worth 4.4 probability points to GS.
- GS comes in with the stronger accumulated season form: that moves 30.0 points toward GS.
What weighs most
Home court+4.4 toward GS
Season strength+30.0 toward GS
probability points
What could flip it
- The model does not read the injury report and has no players inside it: an absence would change the game without changing this number. There is no calculation saying by how much.
- What the model does not look at and this number does not capture: rest and travel, the day's injury report, the starting five and recent form.
Uncertainty
- The model gives a number, not a band: its reliability is measured by comparing forecasts with results, not by drawing a margin.
- It does not know injuries, rest, travel or the starting five: any of them can move this game without the number changing.
- For reference, the pre-tip market consensus gave GS 85.3%.
What changed since the last snapshot
Loading…
After the game
GS 84%
Expected
GS 82 – POR 66
Actual
0.02
this game's error
- Expected: GS at 84.4%.
- Actual: GS 82 – POR 66. GS won.
- The favorite delivered.
- This game's error: 0.02 (0 = nailed it, 1 = worst possible; a 50% call would score 0.25).
- The market consensus did get the side right (85.3% on the winner).
- What we learn: the market consensus was closer to the result than the model. Logged to check whether there was pre-tip information (injuries, rest) the model does not have.
Is it calibrated?
49
games measured
76%
accuracy
0.172
error (Brier)
market 0.169
12 pp
average deviation (ECE)
- Over 49 settled games with this model: it picked the right side 76% of the time and its average error (Brier) was 0.172. When it said X%, reality was on average 12 points away from that X%.
- On those same games, the market consensus had less error (0.169 versus 0.172): the model does not beat it, and we say so.
- 49 games is a small sample: these numbers still move a lot.
How it is measured and limits
Model
SportiveWorld's WNBA game model, research stage: it reads each team's accumulated season strength and home-court advantage. Shown as what it said before tip-off, not as an edge.
This is the preview saved before the game, not current data.
Sample · Measured against 49 settled games.
Injuries · Availability known before tip-off: an injury report existed and was read (it does not enter the number).
- · What you see is the preview sealed before the game, not a later recalculation.
- · 49 games is a small sample: these numbers still move a lot.
As of 19 Sept · SportiveWorld model
Verified against the source · 13/13
- ✓ expected competition and game
- ✓ home and away
- ✓ probability intact
- ✓ favourite = higher probability
- ✓ sealed and signed snapshot
- ✓ registered model
- ✓ sealed before tip-off
- ✓ consistent official result
- ✓ core verified by the server
- ✓ verified components
- ✓ numbers add up
- ✓ freshness without contradiction
- ✓ consistent track record
This is reading the data, not a bet: we work with no bookmakers and show no odds. And a game turns on an injury, a hot night or a bad call.