WNBA · FINAL · MON, SEP 21, 8:00 PM

ATL @ NYL

Dream at Liberty

← All WNBA games on 2026-09-21

Spread (home)

Total

-132 / +106

Moneyline

Snapshot market context · game-quote observation time and source book unavailable · snapshot assembled Sep 21, 10:19 PM EDT

GAME BRIEF · START HERE

Start with what happened, then judge the process.

Dream at Liberty is a completed-game review. Any call below is record-only.

REVIEW

Open the 2-minute projection guide

01 · PROP DECISIONS

What the evidence says

1 of 1 prop card has eligible recent or player-opponent history. Each card explains whether that context supports, opposes, or is too thin.

02 · AVAILABILITY

What could break it

3 hard availability flags returned (Brionna Jones, Matilde Villa, Elizabeth Balogun). Player-level observation times are unavailable, so verify the source report.

03 · MARKET MOVEMENT

What changed

The movement source could not be checked. 3 hard availability flags and 0 uncertain statuses could change player roles. Unknown movement is not neutral.

MODEL UPDATES

HISTORY TEMPORARILY UNAVAILABLE

We could not load the saved forecast history right now. This does not mean the model never changed or made no call.

DECISION TIMELINE · REVISION FLOW

Loading frozen research checkpoints…

HISTORICAL FORMULA FEED · AUDIT ONLY

Results & Recorded Calls · 1 qualified

Start with the verified result and flat-1u CORE benchmark P&L. Then compare the recorded call with the evidence that can be reconstructed without using this game's outcome.

FINAL RESULT · CORE BENCHMARK

LOSS

Actual: 2 vs 3.5
Flat-1u P&L: −1.00u
CLV: +1.2¢
Minutes: 31.7

Postgame context: Conversion unlucky

SERVER RECORD · CORE FORMULA · REBOUNDS

Leonie Fiebich O 3.5 REB +128

Persisted decision price · 2.0u model stake · bet_clv:1421008

RECORD ONLY

Current meaning: This is a settled, record-only decision. It is historical evidence, not a current wager.

PROP BRIEF · START HERE

Review the recorded result—not a current play.

What the evidence saysThe context is split

Recent games · Last 5 supports; Against this opponent pushes back; Similar players vs opponent unavailable. Do not turn correlated or conflicting history into extra confidence.

Main riskWhat could break the read

Decision time is unavailable; the capture bucket must not be read as decision time.

Next stepReview the process

Compare the result, minutes, closing line, and receipt limits below.

Verified receipt & source timing

Canonical decision receipt · Receipt bet_clv:1421008 · formula consensus_devig v1.0.0 · capture bucket Sep 20, 8:00 PM EDT (not decision time) · decision time unavailable · point-in-time unverified

WHY IT WAS SURFACED · FEED PRICE

48.3% model chance vs 43.9% needed at +128 · +4.5pp probability edge · +0.1% expected profit / 1u

Model stake2.0u

Why “<52%” can still show positive value: that label describes raw hit likelihood. At plus money, a candidate can hit less than half the time and still carry positive expected value because a win pays more than 1u.

PREGAME CHECKS · AUDIT

What the receipt can—and cannot—prove

Decision receiptRecord only

+128 · sportsbook unavailable · quote observation unavailable · bet_clv:1421008

Pregame availabilityNot timestamped

No immutable pre-tip player-status receipt is connected.

Pregame roleNot verified

Pick-time minutes, lineup, and role inputs are not verified by the canonical receipt.

SUPPORTING CONTEXT · NOT EXTRA VOTES

Recent, H2H, and similar players

Read each row on its stated basis

Recent games · Last 5Supports

5 of 5 would have hit

Against today's OVER 3.5 · 4.8 average · 29.4 average minutes

Against this opponentPushes back

1 of 5 would have hit

Vs ATL at today's OVER 3.5 · 2.4 average

Similar players vs opponentUnavailable

Comparable-player sample unavailable

The matchup comparison service returned 500.

BOTTOM LINE

The context is split

Recent games · Last 5 supports; Against this opponent pushes back; Similar players vs opponent unavailable. Do not turn correlated or conflicting history into extra confidence.

WHAT THE ARCHIVE IS MISSING

Decision time is unavailable; the capture bucket must not be read as decision time.
The source sportsbook or quote observation timestamp was not preserved.
The pre-tip availability observation was not stored immutably.

See game-by-game evidenceL3 / L5 / L10 · opponent games · comparable players

L3: 3/3
L5: 5/5
L10: 6/9

Player vs opponentFinished above 3.5 in 1/5 prior player meetings vs ATL

Position/team matchupNo stat-matched position/team factor is connected; direct defender assignments are unavailable

Reconstructed from a snapshot assembled Sep 21, 10:19 PM EDT. The target game and all same/future dates are excluded. L3 is contained inside L5, which is contained inside L10; they are one sample, not three confirmations.

DEFENSE CHECK · EACH PLAYER'S OWN CLOSE

How comparable players fared vs ATL

Descriptive · not another model vote

The matchup comparison service returned 500.

Open evidence game log

HISTORICAL MARKET CHECK

Against each game's own verified close

Normal exposure only

Checking verified pre-tip closing lines…

CURRENT-LINE WHAT-IF

Prior results against this card's 3.5 line

A retrospective comparison with this formula line—not the market line offered in each old game.
Date	Opponent	Actual	Minutes	OVER 3.5
Sep 20	@ TOR	4.0	23.9	Hit
Sep 18	@ MIN	5.0	27.8	Hit
Aug 29	vs CHI	4.0	29.5	Hit
Aug 27	vs GSV	6.0	32.0	Hit
Aug 22	vs IND	5.0	33.8	Hit
Jul 3	vs MIN	3.0	36.5	Miss
Jun 30	vs LVA	8.0	33.2	Hit
Jun 28	@ GSV	0.0	31.2	Miss
Jun 25	@ SEA	2.0	30.7	Miss

Player history vs ATL

Latest 5 of 9 eligible meetings · finished above 3.5 in 1/5 · average 2.4. Context only; roles and schemes change.

Date	Actual	Minutes	OVER 3.5
Jun 11	3.0	26.6	Miss
Aug 23	2.0	24.5	Miss
Jul 13	4.0	32.7	Hit
Sep 24	2.0	21.5	Miss
Sep 22	1.0	25.4	Miss

See methodology & source audit6 pillars · 49 checks

The six items are pillars, not six yes/no questions. Each check below is evidence—not a gamified score. One hard failure can outweigh many verified inputs.

1 · Price & market value

Caution

What exact side, line, odds, and book are offered?

Leonie Fiebich OVER 3.5 +128 · sportsbook unavailable

Source · canonical grading decision receipt

Caution

When was the formula price captured?

Capture bucket Sep 20, 8:00 PM EDT (storage bucket, not decision time) · decision time unavailable · quote observation unavailable

Source · bet_clv:1421008

Verified

What hit rate does the price require?

43.9%

Source · derived from formula-feed American odds

Verified

What is the calibrated probability at this line?

48.3%

Source · canonical decision model.calibrated_hit_probability

Verified

What is the probability-point edge?

+4.5pp

Source · p_hit minus formula-price break-even

Verified

What is expected profit per unit?

+0.1%

Source · canonical decision model.claimed_ev_pct (when persisted)

N/A

Does the exact line appear in the connected snapshot?

Mutable latest-market rows are intentionally excluded from historical pre-tip proof

Unavailable

What price or line turns this into a pass?

Server-authoritative playable boundaries are not published yet

N/A

Does the broader market agree?

Mutable latest-market rows are not used to reconstruct historical consensus

Unavailable

Is the market liquid enough?

Limits and liquidity are not connected

2 · Projection quality & uncertainty

Unavailable

What is the projected median?

Unavailable

What does the outcome distribution look like?

Distribution unavailable

Unavailable

How sensitive is the pick to minutes, usage, or pace?

Sensitivity model unavailable

Caution

What do L3, L5, and L10 show at this exact line?

L3 3/3 · L5 5/5 · L10 6/9 at the formula OVER 3.5 line

Source · reconstructed hydrator game history; target/same/future dates excluded

Caution

Is the model calibrated for this league, prop, and bucket?

The pick carries calibrated p_hit; bucket-level proof is not exposed here

Unavailable

Was calibration measured out of sample?

Out-of-sample calibration evidence is not exposed per pick

Caution

How large is the comparable historical sample?

9 eligible prior games in L10 · 0 under 10 minutes · nested descriptive sample, not 9 independent confirmations

Unavailable

When did the model run and were inputs complete?

Research run timestamp unavailable

Source · Gameplay Intelligence

3 · Availability & role

Unavailable

Is the player active, questionable, limited, or returning?

No hard game-report flag; individual status is not fully resolved

Source · player-state service

Caution

When was status last verified?

Snapshot assembled Sep 21, 10:19 PM EDT; injury-source observation time is not mapped

Source · data-hydrator snapshot assembly

Unavailable

What is the expected minutes range?

Minutes projection unavailable

Unavailable

Is the player starting or restricted?

Starter and restriction status unavailable

Unavailable

Has opportunity changed?

Opportunity feed unavailable

Unavailable

Which teammate absences change the role?

No confirmed teammate-absence attribution returned

Unavailable

Were injury-shortened games identified?

The joined outcome does not include a verified abnormal-exposure flag

Unavailable

What rotation change could invalidate the projection?

No material role change returned

4 · Game environment & matchup

Unavailable

How many possessions are expected?

A game-level pace projection is not exposed on this card

Caution

What do spread and total imply?

Spread (home) 0 · Total 0 · Moneyline -132 / +106 · archive context only; not an immutable pre-tip market receipt · snapshot assembled Sep 21, 10:19 PM EDT

Source · data-hydrator assembled game-market snapshot

Unavailable

Does opponent scheme affect this stat?

No stat-matched position/team factor is connected; direct defender assignments are unavailable

Caution

What has this player done against this opponent?

Finished above 3.5 in 1/5 prior player meetings vs ATL

Source · reconstructed player-vs-team game history; target date excluded

Unavailable

Which defender or lineup matters?

Defender attribution is not exposed

Unavailable

Is rest, travel, or schedule a disadvantage?

Rest and travel context are not connected

Caution

What foul, blowout, or game-script path creates downside?

Spread is visible; player-level foul and game-script risk are not quantified

5 · Execution & portfolio risk

N/A

Is the number still inside its playable boundary?

Not applicable after tip; mutable current offers are excluded from historical execution proof

Verified

How was stake size derived?

2.0u model-sized stake

Source · canonical decision model.kelly_stake_units

Unavailable

How much correlated exposure already exists?

Player, game, and injury-thesis exposure are not connected

Unavailable

Are exposure limits respected?

Daily, league, player, and prop limits are not exposed

Caution

What event withdraws or expires the pick?

Game start and hard availability conflicts are enforced; price boundaries remain a backend gap

Caution

Is this pregame, live, or halftime analysis?

Schedule status: final · formula-feed archive context

Caution

Is there a no-chase rule?

The UI expires the card after tip and requires exact-line verification; explicit price-bound enforcement is pending

6 · Outcome evidence & accountability

Verified

Did the pick beat the comparable close?

+1.2¢ closing-line value (price CLV)

Source · joined grading settlement

Verified

Was the formula line stored immutably?

OVER 3.5 · decision key pd_0b6723aa51daf7db13fbe067 · receipt bet_clv:1421008 · formula consensus_devig v1.0.0

Source · prop-decision-envelope.v1

Verified

Was a miss variance, injury, foul trouble, or blowout?

Conversion unlucky

Source · joined outcome receipt

Verified

What are flat-1u and actual-staked ROI?

This pick's flat-1u CORE result is −1.00u. Personal staking remains separate.

Source · joined grading ledger

Unavailable

What are sample size and uncertainty interval?

Per-cohort confidence interval is not exposed on the card

Unavailable

Are probabilities calibrated by Brier or log loss?

Per-cohort calibration metrics are not exposed

Unavailable

Where is performance weakening?

League/prop/price/book/time-to-tip drift is not joined here

Unavailable

What are drawdown and losing-streak behavior?

Portfolio drawdown belongs to the ledger and is not connected to this pick

Unavailable

Is the edge persisting in new data?

Model-versioned out-of-sample persistence is not exposed per pick

DECISION LAB · MODEL RECORD + YOUR READ

Challenge what the system knew.

The original records stay frozen. Your thesis is checked against the captured evidence, then any explicit numerical assumption is tested separately—never silently fed into CORE or Gameplay.

HISTORICAL

SELECTED PROP

Leonie Fiebich

OVER 3.5 · Rebounds

NO GAMEPLAY CAPTUREThe exact CORE record has no matching immutable Gameplay player capture to compare.

RECORDED CORE LANE

What the formula record actually says

DECISION RECEIPT

FORMULA FEATURE INPUTSNOT STORED IN THIS RECEIPTThe receipt preserves pricing outputs and the action—not the lineup, minutes, usage, or matchup contribution path.

Stored pricing outputs

CAPTURED PRICE+128book not stored · time not stored

CALIBRATED SIDE48.3%break-even 43.9% · EV +0.1%

STORED ACTIONrecord onlyOVER 3.5 · stake 2.00u

Formula: consensus_devig
Version: v1.0.0
Decision time: time not stored
Input status: unverified
Point-in-time: unverified
Receipt quality: partial

Receipt integrity flagssource sportsbook unavailable · quote observed at unavailable · decision at unavailable · capture receipt unavailable

game settled

NO GAMEPLAY CAPTUREEvidence beside the call is not automatically evidence inside the call.

IMMUTABLE GAMEPLAY CAPTURE

What the research system captured

CAPTURE BUILDING

Gameplay evidence is capture building.

The CORE row remains visible, but there is no immutable research projection from which to build a numeric human scenario.

REPLAY LAB · HINDSIGHT CHALLENGE

Tell us what the model may have missed

Write your basketball thesis normally. We'll turn it into checkable claims, compare each one with the frozen Gameplay capture, and show whether a quantified counterfactual changes the make.

UNSAVED REPLAY

HINDSIGHT MODEYou can test assumptions, but this page cannot save or alter a pregame read after the authoritative tip time. Browser timestamps are a convenience, not independent proof of publication.

PRIMARY WORKFLOW

Challenge the call

FROZEN CAPTURE · DETERMINISTIC CHECK

What is your basketball read?

Your prose supplies hypotheses, never evidence. This check only recognizes explicit starter, minutes, availability, usage, matchup, pace, and rotation claims.

CORE RECORDOVER 3.548.3% side probability

GAMEPLAY MAKE—— model minutes

GAMEPLAY LEANNO SAME-LINE CAPTUREno capture at 3.5

MY MAKE—scenario unavailable

Advanced manual scenarioDirectly edit minutes, opportunity, matchup, and pace

ISOLATED-PLAYER SENSITIVITYMoving one player's minutes here does not conserve the team's 200 regulation minutes. Name whose minutes or role you think changes so the assumption stays basketball-real.

Starting-role beliefRecorded label: unknown. This label does not secretly add minutes.Expected minutes —Model: — · regulation maximum 40Opportunity / usage change +0%Use for role, touches, shot volume, rebound chances, or potential assists.Matchup adjustment +0.0Direct rebounds added to or removed from the make.Pace / game-script adjustment +0.0Direct rebounds added to or removed from the make for pace, blowout, overtime tail, or foul-risk beliefs.

My decision

Confidence in my readLineup / availability beliefWhose minutes / role moves?This note is required for a rotation-aware interpretation; the calculator itself remains player-only.

TRANSPARENT SCENARIO MATH

Gameplay baseline—
After your minutes—
After opportunity—
After matchup + pace—

WHAT FLIPS THE LINE?Unavailable

Under your current opportunity, matchup, and pace assumptions, this is the minutes threshold where the personal make equals 3.5.

A numeric personal make needs both a stored Gameplay projection and model minutes. You can still record a stance and thesis.

How this personal make is calculated

Gameplay mean × (your minutes ÷ model minutes) × (1 + your opportunity change) + your matchup adjustment + your pace/game-script adjustment. It is a sensitivity tool, not a retrained probability, and it does not claim the CORE formula uses these inputs.

DECISION LAB

Loading the immutable evidence map…

AVAILABILITY SNAPSHOT · AUDIT CONTEXT

Availability & Injuries

3 hard availability snapshot flags appear. Player-level observation time is unavailable, so this does not prove the status existed before tip; formula rows remain record-only.

ATLBrionna Jones · Injury/Illness-LeftLeg;N/AOUT

ATLMatilde Villa · Coach'sDecisionOUT

NYLElizabeth Balogun · NotWithTeamOUT

GAMEPLAY INTELLIGENCE · IMMUTABLE PRE-TIP CAPTURE

Model Outlook

BUILDING

The first frozen read has not posted for this game yet. We publish a checkpointed snapshot before tip instead of inventing a live answer from incomplete inputs.

GAMEPLAY INTELLIGENCE · IMMUTABLE PRE-TIP CAPTURE

Model Outlook

Checking the latest captured lineup, rotation, matchup and market state…

MARKET CONTEXT · NOT A PICK · THIS GAME · T−1H

Prop Watchlist

A shortlist of unusual market lines. Compare this game's checkpoint with a player's typical close, then see whether recent results against their own lines agree. Only the Formula-Qualified Recommendations section contains CORE candidate wagers.

PROP FEED UNAVAILABLE

The market observation feed could not be checked

The player-prop signal feed could not be checked. The matchup page still uses the independently available schedule, injury, and recommendation data.

The Read

This preview is built on an extremely thin data set, and that has to be the headline before anything else. The only usable inputs are the New York Liberty team profile: a 110.0 season offensive rating on 53.8% effective field goal percentage, a 15.0% turnover rate, a 23.3% offensive rebound rate and a 30.4% free-throw rate, paired with a 105.8 defensive rating. That nets out to roughly a +4.2 net rating at a pace of 81.8, which the team agent characterizes as slightly above average. On the defensive end, New York allows 86.52 points per game (rank 7), 31.4 rebounds (rank 3), 21.43 assists (rank 12) and 9.02 made three-pointers (rank 11).

The most important note is what is missing. The Atlanta Dream team agent returned an empty object, the matchup agent returned an empty object, and all twelve player agents returned empty prop_picks arrays with no stats summaries and no minutes projections. There is no schedule, home/away, rest-day or back-to-back information for either side, no last-5 or last-10 split data, no opponent shooting splits, and no head-to-head history. Without a matchup agent there is no pace or total projection, and without coach or player agents there are no minutes projections to cross-validate counting-stat props against. Any prop pick produced under these conditions would be fabricated rather than derived.

The one concrete availability note is that Elizabeth Balogun is listed as out with a NotWithTeam designation. No minutes or usage figures were provided, so the redistribution of her role cannot be estimated from this data. On the Atlanta side, nothing at all is available — no record, no ratings, no injury report. That means any claim about how the Dream will match up, how they will defend the perimeter, or how they will handle New York's frontcourt would be guesswork.

Given that, the honest read is that this game cannot be modeled with confidence. The only defensible lean is a mild one toward New York on the strength of its +4.2 season net rating and a top-10 scoring defense, but with zero opponent data that lean carries no real signal. No player props are carried forward because none were supplied, and none can be responsibly constructed. If a fuller data feed becomes available — particularly the Dream team profile, the matchup agent's pace and total projection, and individual minutes projections — this game should be re-run from scratch.

SUPPORTING RESEARCH · NOT A BETTING RECOMMENDATION

Matchup Outlook

NYL

Projected winner

-3.5

Model spread

O 158.5

Model total

40%

Outlook confidence · MODERATE

Team Series History

NYL vs ATL · last 24 meetings · NYL leads

Team results are game-environment context—not validation for an individual player prop. Player-specific opponent history lives inside each recommendation card.

14–10

NYL record

85.3–81.8

Avg score (home)

+3.5

Avg margin

35.7

Avg reb

RECENT MEETINGS

06/11NYL 104 @ ATL 90WNYL