Public scorecard · settled record

How the desk actually did.

Resolved paper-portfolio picks, published after settlement and delayed by at least 24 hours so they can't be copy-traded. Live picks and open positions do not appear here.

Research status

Current recommendations paused

Research validation is in progress. Current recommendations and subscriber research updates are paused while we review the captured evidence. No restart date has been set.

The bounded diagnostic pilot ended on September 26, 2026. Its results validate data capture, not forecasting accuracy or trading returns.

Historical paper results are presented separately when available. They are not live account holdings, executed trades, or evidence of a validated trading edge.

View historical paper results

The record

Settlement hit rate

36.3%

78 of 215 settled · covers 44% of 485 realized positions

Primary resolver-grade settlement cohort

Managed return

−20.497

total from $1 on each position · settlements + paper exits · after modelled fees

Realized positions

485

246 settled · 239 paper exits

Hit rate counts official settlements only, so early paper exits never inflate it; managed return and position count include those risk-managed exits. Published after the embargo — past results don't guarantee future outcomes; research only, not financial advice.

What counts here

Hit rate uses official settlements only — early paper exits never count toward it, so it describes part of the book rather than all of it. Managed return and the realized-position count do include those executable paper exits. Both returns are net of modelled trading fees — settlements model the entry fee, early exits model entry and exit. Proxy, archive, and legacy rows stay broken out below, never blended into the headline.

Settlement rules

Headline metrics include only trades whose market is in the resolver-aligned primary cohort (aligned + source_kind in {resolver_truth, manual_verified} + source_status in {parsed, manual_verified}). Trades sourced from station proxies, reanalysis archives, mixed/unknown legacy, or markets without any temperature_actuals coverage are listed in `diagnostic_cohorts` and `trades` with their truth_mode labelled — never blended into the headline. Paper exits are shadow-portfolio risk management events, not official YES/NO settlements, so they are excluded from the settlement headline and settlement accuracy metrics (they appear only in the separate managed record).

Full breakdown

Every return below is for one dollar staked on each position, so the figures stay comparable no matter the size. Both lines are net of modelled trading fees.

Realized positions

485

246 settled · 239 paper exits

485 of 485 positions priced

Realized return

−20.497

−0.042 average across 485 priced positions

From settlements

−116.837

official settlements · after modelled entry fee

From paper exits

+96.339

early exits · after modelled entry and exit fees

Hit rate covers the primary resolver-grade settlement cohort — 215 of 485 realized positions. Managed return covers the broader eligible settlement population (all embargoed official settlements, every cohort): 246 settlements, plus every paper exit.

Managed record combines official YES/NO settlements with executable paper exits — early risk-management exits taken at the market's executable price before settlement. Returns are flat-stake: the cumulative result of staking $1 on each realized position, so no position sizing is revealed. Both components model V2 taker fees: settlements model the entry fee only (resolution is not a trade), paper exits model entry plus exit fees (an early close is a taker round trip). These are modeled replays, not realized fees, and exclude slippage. Paper exits are not market settlements and never count toward the settlement hit rate. Settlement returns follow the same embargo as the headline; paper exits appear once recorded, as aggregate counts and return only — no per-position rows. A position with no expressible return is counted in the realized total and reported separately as return-unavailable; it is never summed as zero and never enters an average denominator.

How fees are modelled

Settlement returns model ONE V2 taker fee — the entry buy. A contract held to official resolution has no exit trade, so no exit fee applies.

Paper-exit returns model TWO V2 taker fees — the entry buy and the exit sell — because closing early is a taker round trip. Both legs use Polymarket's share- and price-dependent fee formula evaluated CONTINUOUSLY at the configured weather fee rate. Polymarket rounds the fee it actually charges to 5 decimal places per trade; this replay applies no trade-level venue rounding, so a modeled figure can differ from the venue's charge in the fifth decimal per leg.

Accounting v2. Settlements model an entry taker fee; paper exits model entry plus exit taker fees. Both are MODELED replays of the continuous analytical fee formula over recorded entry and exit prices at the published modeled_taker_fee_rate — they are not proof of the fees actually charged. Polymarket rounds the charged fee to 5 decimal places per trade and this replay applies no trade-level venue rounding (venue_rounding_applied is false), so figures are not venue-reconciled; they also model no slippage, partial fills, or price impact. Hit rate is settlement-only and is computed on the primary resolver-grade cohort; managed return covers the broader eligible settlement population plus paper exits, so the two figures have different denominators — see hit_rate_coverage. Positions whose return cannot be expressed are reported as a separate count and are excluded from every sum and average rather than treated as zero; when return_coverage_complete is false the published returns are priced-position subtotals, not totals. Under v1, paper exits carried no modeled fee, so rolling figures published before this version may be restated under the versioned convention.

Modelled fee rate: 5.00% per taker leg
Venue rounding: not applied — venue rounds to 5 decimals
Accounting: scorecard-accounting-v2

Resolved trades · latest 12 of 246

Each row's badge names how it settled: Verified = the market's official resolution source (the only rows that count in the headline); Provisional = settled another way, or not yet confirmed. Hover a badge for the exact method.

Sep 25TaipeiProvisional
Will the highest temperature in Taipei be 33°C on September 25?
Direction
NO @ 0.760
Result
lost · YES
Return / stake
−1.012
High temp
—
Sep 24Tel AvivProvisional
Will the highest temperature in Tel Aviv be 31°C on September 24?
Direction
NO @ 0.670
Result
lost · YES
Return / stake
−1.016
High temp
—
Sep 24TaipeiVerified
Will the highest temperature in Taipei be 33°C on September 24?
Direction
NO @ 0.660
Result
won · NO
Return / stake
+0.498
High temp
89.6°Fweather_underground
Sep 24TaipeiVerified
Will the highest temperature in Taipei be 32°C on September 24?
Direction
NO @ 0.639
Result
lost · YES
Return / stake
−1.018
High temp
89.6°Fweather_underground
Sep 23IstanbulProvisional
Will the highest temperature in Istanbul be 18°C on September 23?
Direction
NO @ 0.777
Result
won · NO
Return / stake
+0.276
High temp
—
Sep 23IstanbulProvisional
Will the highest temperature in Istanbul be 17°C on September 23?
Direction
NO @ 0.520
Result
lost · YES
Return / stake
−1.024
High temp
—
Sep 23TaipeiVerified
Will the highest temperature in Taipei be 33°C on September 23?
Direction
NO @ 0.750
Result
won · NO
Return / stake
+0.321
High temp
89.6°Fweather_underground
Sep 23TaipeiVerified
Will the highest temperature in Taipei be 32°C on September 23?
Direction
NO @ 0.510
Result
lost · YES
Return / stake
−1.024
High temp
89.6°Fweather_underground
Sep 22Tel AvivVerified
Will the highest temperature in Tel Aviv be 30°C on September 22?
Direction
NO @ 0.579
Result
lost · YES
Return / stake
−1.021
High temp
86.0°Fweather_underground
Sep 21IstanbulProvisional
Will the highest temperature in Istanbul be 26°C on September 21?
Direction
NO @ 0.790
Result
lost · YES
Return / stake
−1.010
High temp
—
Sep 21TaipeiVerified
Will the highest temperature in Taipei be 32°C on September 21?
Direction
NO @ 0.580
Result
lost · YES
Return / stake
−1.021
High temp
89.6°Fweather_underground
Sep 20IstanbulProvisional
Will the highest temperature in Istanbul be 24°C on September 20?
Direction
NO @ 0.500
Result
lost · YES
Return / stake
−1.025
High temp
—

Show all 246 resolved trades →

Sep 20TaipeiVerified
Will the highest temperature in Taipei be 33°C on September 20?
Direction
NO @ 0.780
Result
lost · YES
Return / stake
−1.011
High temp
91.4°Fweather_underground
Sep 19Tel AvivVerified
Will the highest temperature in Tel Aviv be 30°C on September 19?
Direction
NO @ 0.540
Result
lost · YES
Return / stake
−1.023
High temp
86.0°Fweather_underground

Other settlement sources

These rows are shown for transparency but are not counted in the headline record. They use settlement sources such as nearby stations, weather archives, legacy data, or rows that are not fully labelled yet.

Uncovered31 trades
Hit
32.3%
Avg / $1
−0.497
Cumulative
−15.406

Current recommendations paused

Research validation is in progress. Current recommendations and subscriber research updates are paused while we review the captured evidence. No restart date has been set.

The bounded diagnostic pilot ended on September 26, 2026. Its results validate data capture, not forecasting accuracy or trading returns.

Historical paper results are presented separately when available. They are not live account holdings, executed trades, or evidence of a validated trading edge.

View historical paper results

The record

Settlement hit rate

36.3%

78 of 215 settled · covers 44% of 485 realized positions

Primary resolver-grade settlement cohort

Managed return

−20.497

total from $1 on each position · settlements + paper exits · after modelled fees

Realized positions

485

246 settled · 239 paper exits

Full breakdown

Every return below is for one dollar staked on each position, so the figures stay comparable no matter the size. Both lines are net of modelled trading fees.

Realized positions

485

246 settled · 239 paper exits

485 of 485 positions priced

Realized return

−20.497

−0.042 average across 485 priced positions

From settlements

−116.837

official settlements · after modelled entry fee

From paper exits

+96.339

early exits · after modelled entry and exit fees

How fees are modelled

Settlement returns model ONE V2 taker fee — the entry buy. A contract held to official resolution has no exit trade, so no exit fee applies.

Modelled fee rate: 5.00% per taker leg
Venue rounding: not applied — venue rounds to 5 decimals
Accounting: scorecard-accounting-v2

Resolved trades · latest 12 of 246

Sep 25TaipeiProvisional

Will the highest temperature in Taipei be 33°C on September 25?

Direction: NO @ 0.760
Result: lost · YES
Return / stake: −1.012
High temp: —

Sep 24Tel AvivProvisional

Will the highest temperature in Tel Aviv be 31°C on September 24?

Direction: NO @ 0.670
Result: lost · YES
Return / stake: −1.016
High temp: —

Sep 24TaipeiVerified

Will the highest temperature in Taipei be 33°C on September 24?

Direction: NO @ 0.660
Result: won · NO
Return / stake: +0.498
High temp: 89.6°Fweather_underground

Sep 24TaipeiVerified

Will the highest temperature in Taipei be 32°C on September 24?

Direction: NO @ 0.639
Result: lost · YES
Return / stake: −1.018
High temp: 89.6°Fweather_underground

Sep 23IstanbulProvisional

Will the highest temperature in Istanbul be 18°C on September 23?

Direction: NO @ 0.777
Result: won · NO
Return / stake: +0.276
High temp: —

Sep 23IstanbulProvisional

Will the highest temperature in Istanbul be 17°C on September 23?

Direction: NO @ 0.520
Result: lost · YES
Return / stake: −1.024
High temp: —

Sep 23TaipeiVerified

Will the highest temperature in Taipei be 33°C on September 23?

Direction: NO @ 0.750
Result: won · NO
Return / stake: +0.321
High temp: 89.6°Fweather_underground

Sep 23TaipeiVerified

Will the highest temperature in Taipei be 32°C on September 23?

Direction: NO @ 0.510
Result: lost · YES
Return / stake: −1.024
High temp: 89.6°Fweather_underground

Sep 22Tel AvivVerified

Will the highest temperature in Tel Aviv be 30°C on September 22?

Direction: NO @ 0.579
Result: lost · YES
Return / stake: −1.021
High temp: 86.0°Fweather_underground

Sep 21IstanbulProvisional

Will the highest temperature in Istanbul be 26°C on September 21?

Direction: NO @ 0.790
Result: lost · YES
Return / stake: −1.010
High temp: —

Sep 21TaipeiVerified

Will the highest temperature in Taipei be 32°C on September 21?

Direction: NO @ 0.580
Result: lost · YES
Return / stake: −1.021
High temp: 89.6°Fweather_underground

Sep 20IstanbulProvisional

Will the highest temperature in Istanbul be 24°C on September 20?

Direction: NO @ 0.500
Result: lost · YES
Return / stake: −1.025
High temp: —

Show all 246 resolved trades →

Sep 20TaipeiVerified
Will the highest temperature in Taipei be 33°C on September 20?
Direction
NO @ 0.780
Result
lost · YES
Return / stake
−1.011
High temp
91.4°Fweather_underground
Sep 19Tel AvivVerified
Will the highest temperature in Tel Aviv be 30°C on September 19?
Direction
NO @ 0.540
Result
lost · YES
Return / stake
−1.023
High temp
86.0°Fweather_underground