# Weekly Model Watch #4

**WEEKLY · MODEL WATCH** · September 2, 2026 · MarketMania Research · Weekly series

Window: **Aug 24-30, 2026 UTC** · Cutoff: **Mon Aug 31, 2026, 16:00 UTC** · Methodology **v1.1 (2026-08-10)**, hash `e66c7e8c864a2233`

PDF: https://marketmania.ai/research/reports/model-watch-2026-08-24.pdf · Open data (JSON): https://marketmania.ai/research/reports/model-watch-2026-08-24.json

> Research question: *"who led the week, which tickers were readable, who saw the turn first, and does a model agreeing with itself across timeframes/horizons make its call stronger?"*

## Research Snapshot

| RESEARCH SNAPSHOT |  |  |  |
|---|---|---|---|
| **4,576** directional calls scored | **7** models (grok line spliced) | of **9,301** mature forecasts | window **Aug 24-30, 2026 UTC** |
| hit = direction rule: tp1/tp2->hit, sl->miss, expiry->sign of gross pnl | FH gate **1h / 4h / 1d** | base field hit **42.9%** | cutoff **Mon Aug 31, 2026, 16:00 UTC** |
| market: BTC net **-0.07%** (prior +23.58%) | ann vol **30.9%** (was 71.1%) | TOP5 volume **$19.72B**, -24% w/w | pairwise corr **0.85** (was 0.81) |

## Key finding

> **KEY FINDING · OBSERVATION (one weekly window)**
>
> ### Field accuracy went 54.0% -> 42.9% (-11.1pp) with 0 of 7 models improving, and no weekly title was awarded, for the fourth issue running.

## TL;DR

- **OBSERVATION -- No weekly title again.** deepseek-v4-pro leads descriptively at 44.6% [41.0%, 48.2%]; the gap to #2 (gemini-3.1-pro, 44.2%) is 0.3pp. 0 of 7 models improved week over week; issue #3's leader claude-opus-5 is now 6 of 7.
- **No ticker cleared 51%.** SOL is the week's most readable coin (45.8%, n=1,025) and XRP the hardest at 40.3%; 0 of the 5 assets cleared 51%. Pair of the week: qwen-3.8-max x SOL 48.7% (n=150); worst: claude-opus-5 x XRP 34.6% (n=104).
- **3 of 4 qualifying turns found a caller.** BTC short -2.49%, 8 caller(s), first gpt-5.6-sol; ETH long +2.31%, 13 caller(s), first gpt-5.6-sol; ETH short -2.80%, 0 caller(s), first NOBODY; ETH short -2.52%, 1 caller(s), first deepseek-v4-pro.
- **Self-agreement read as a penalty in every cell.** Cross-TF/FH lifts ran -7.6pp to -1.9pp (issue #3: -0.6 to +4.6pp). On trend days, 4h TF-agreement hit 41.9% (n=248) against 36.0% (n=225) on flat days.

### Week-over-week chart (see PDF for the grouped bar chart)

Directional hit-rate (%) by model, week over week, ranked by this week's rate. Short names: opus-5 = claude-opus-5, fable-5 = claude-fable-5, gemini-3.1 = gemini-3.1-pro, gpt-5.6 = gpt-5.6-sol, qwen-3.8 = qwen-3.8-max, deepseek = deepseek-v4-pro. Field base 54.0% -> 42.9%. The grok column compares the spliced 4.5+4.6 line against issue #3's pure grok-4.5. "Was" values are the numbers published in issue #3; the leaderboard's W/w column is computed on unrounded rates and can differ by 0.1pp from the difference of the rounded columns.

## Why it matters

MarketMania scores 7 model lines against the same market, hour after hour. Weekly Model Watch asks four practical questions about that week of calls: who led the week, which tickers were readable, who saw the turn first, and does a model agreeing with itself across timeframes/horizons make its call stronger? Each question gets its own block below, with its own table, its own N, and its own honesty caveats. Issue #4 is the first issue whose grok row is a lineage splice, so read that row as a line, not a single model build.

## How to read this

Every block below is scored on the same pool: matured directional calls inside the 1h/4h/1d FH gate, hit by the methodology v1.1 direction rule. The weekly title is awarded only on a >=5pp gap AND non-overlapping 95% Wilson CIs with N>=10 per side -- a descriptive lead is not a title. Reversals use the 4h grid with a >=2.0% counter-move sustained 8h, BTC/ETH only in v1. Cross-TF/FH blocks ask whether a model agreeing with itself does any better, and are split trend vs flat wherever the week supplies both sides. The grok row is the lineage splice grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22).

## Leaderboard of the week

| Model | N | Cov. | Hit rate | 95% CI | W/w pp |
|---|---|---|---|---|---|
| deepseek-v4-pro | 729 | 54.8% | 44.6% | 41.0%-48.2% | -7.0 |
| gemini-3.1-pro | 755 | 56.9% | 44.2% | 40.7%-47.8% | -9.9 |
| claude-fable-5 | 651 | 49.2% | 43.0% | 39.3%-46.8% | -13.5 |
| gpt-5.6-sol | 729 | 54.8% | 42.7% | 39.1%-46.3% | -11.0 |
| qwen-3.8-max | 672 | 50.5% | 42.6% | 38.9%-46.3% | -9.7 |
| claude-opus-5 | 561 | 42.2% | 42.4% | 38.4%-46.6% | -16.7 |
| grok-4.6 (spliced) | 479 | 36.0% | 39.9% | 35.6%-44.3% | -11.5 |

No weekly title is awarded this issue -- for the fourth time. Title rule: **>=5pp gap AND non-overlapping 95% CI, N>=10** per side. deepseek-v4-pro's lead over gemini-3.1-pro is 0.3pp and their CIs overlap, so neither condition is met. grok-4.6 (spliced) = grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22).

## Ticker of the week

| Symbol | N | Hit rate | 95% CI | Issue #3 |
|---|---|---|---|---|
| SOL | 1,025 | 45.8% | 42.7%-48.8% | 54.3% |
| BTC | 940 | 43.7% | 40.6%-46.9% | 57.5% |
| ETH | 827 | 43.3% | 40.0%-46.7% | 52.8% |
| BNB | 895 | 41.2% | 38.0%-44.5% | 54.0% |
| XRP | 889 | 40.3% | 37.1%-43.5% | 51.2% |

SOL was the field's most readable ticker (45.8%) and XRP the hardest (40.3%); 0 of 5 assets cleared 51% this week (issue #3: 5 of 5). Issue #3's best ticker was BTC and its hardest XRP -- this issue's ordering is SOL > BTC > ETH > BNB > XRP.

## Pair of the week

Top 6 of the 12 model x ticker pairs tracked this week (best hit-rate), then the 5 worst in their own table -- the pair list is split rather than paged so neither table breaks across a page. Pairs shown require N of 8 or more calls.

| Model (top 6) | Symbol | N | Hit rate |
|---|---|---|---|
| qwen-3.8-max | SOL | 150 | 48.7% |
| gemini-3.1-pro | ETH | 130 | 47.7% |
| claude-opus-5 | SOL | 131 | 47.3% |
| deepseek-v4-pro | SOL | 154 | 46.1% |
| gemini-3.1-pro | XRP | 154 | 46.1% |
| gpt-5.6-sol | SOL | 157 | 45.9% |

| Model (5 worst) | Symbol | N | Hit rate |
|---|---|---|---|
| grok-4.6 (spliced) | ETH | 81 | 38.3% |
| claude-fable-5 | XRP | 116 | 37.1% |
| qwen-3.8-max | XRP | 135 | 36.3% |
| grok-4.6 (spliced) | BNB | 100 | 35.0% |
| claude-opus-5 | XRP | 104 | 34.6% |

qwen-3.8-max x SOL (48.7%, n=150) is the week's best pair and claude-opus-5 x XRP (34.6%, n=104) the weakest. Pair history accumulates across issues -- read this as the fourth data point, not a ranking.

## Who saw the reversal first

Reversal definition: 4h grid; trend = sign of prior 24h; counter-move of 2.0% or more (BTC/ETH) sustained 8h. A model 'calls' the reversal if it has a matured directional hit call in the new direction, FH 4h or 1d, in the slot window [T-12h, T+FH].

| Symbol | Time (UTC) | New side | Move (8h) | Callers | First caller |
|---|---|---|---|---|---|
| BTC | Fri Aug 28, 12:00 | Short | -2.49% | 8 | gpt-5.6-sol |
| ETH | Wed Aug 26, 16:00 | Long | +2.31% | 13 | gpt-5.6-sol |
| ETH | Fri Aug 28, 12:00 | Short | -2.80% | 0 | NOBODY |
| ETH | Sun Aug 30, 16:00 | Short | -2.52% | 1 | deepseek-v4-pro |

The BTC short turn drew 8 caller(s), first gpt-5.6-sol on a 4h/1h call at Aug 28, 04:01 with confidence 62; the ETH long turn drew 13 caller(s), first gpt-5.6-sol on a 1d/4h call at Aug 27, 00:01 with confidence 69; the ETH short turn drew NOBODY; the ETH short turn drew 1 caller(s), first deepseek-v4-pro on a 4h/4h call at Aug 30, 04:01 with confidence 56. Scope in v1 remains BTC/ETH only; the event count across four issues is still far too small to characterise turn-detection skill either way.

## Cross-TF and cross-FH confirmation: does agreeing with yourself help?

| Block | FH | vs. | Agree n | Agree hit | Dis n | Dis hit | Base hit | Lift |
|---|---|---|---|---|---|---|---|---|
| cross-TF | 4h | TF 4h vs 1h | 473 | 39.1% | 18 | 11.1% | 41.0% (712) | -1.9 |
| cross-TF | 1d | TF 1d vs 4h | 105 | 49.5% | 10 | 100.0% | 55.9% (143) | -6.4 |
| cross-FH | 4h | by FH 1h | 429 | 38.9% | 25 | 24.0% | 41.0% (712) | -2.1 |
| cross-FH | 1d | by FH 4h | 89 | 48.3% | 10 | 90.0% | 55.9% (143) | -7.6 |

* = N<10 (insufficient); insufficient cells are never used to rank models. Mixed pairs (a model with only one side of the comparison present): 218 / 28 / 258 / 44. Pooled across all four cell families the lifts run -7.6pp to -1.9pp this week (issue #3: -0.6 to +4.6pp).

**Regime control.** Methodology calls for a trend-vs-flat split on this table. This window has 3 trend days of 7: 08-24 +1.61% (trend), 08-25 -0.58%, 08-26 +0.63%, 08-27 +1.53% (trend), 08-28 -2.91% (trend), 08-29 +0.48%, 08-30 -0.70%. On trend days, 4h TF-agreement hit 41.9% (n=248) against 36.0% (n=225) on flat days.

## Side-mix and the week's failure

| Model | N | Long | Short | Sideways | L:S |
|---|---|---|---|---|---|
| claude-fable-5 | 1,323 | 34.8% | 14.4% | 50.8% | 2.41 |
| claude-opus-5 | 1,330 | 29.2% | 13.0% | 57.8% | 2.24 |
| deepseek-v4-pro | 1,330 | 32.3% | 22.5% | 45.2% | 1.44 |
| gemini-3.1-pro | 1,328 | 35.2% | 21.6% | 43.1% | 1.63 |
| gpt-5.6-sol | 1,330 | 30.3% | 24.5% | 45.2% | 1.24 |
| grok-4.6 (spliced) | 1,330 | 18.3% | 17.7% | 64.0% | 1.03 |
| qwen-3.8-max | 1,330 | 31.6% | 18.9% | 49.5% | 1.67 |

L:S = each model's own long-share divided by its short-share; the field average of the per-model ratios is 1.66 this week. 'Sideways' was 50.8% of mature forecasts (4,725 of 9,301). Consensus skew: 539 symbol/FH/TF cells had 5 or more models on the same side and hit 40.0% (n=3,371) against the 42.9% field base.

| Model | Symbol | Side | Conf | FH / TF | Slot (UTC) | Exit | Net PnL |
|---|---|---|---|---|---|---|---|
| gemini-3.1-pro | SOL | Long | 85 | 4h / 4h | Aug 24, 16:01 | EXPIRY | -$0.28 |

**Fail of the week.** The week's single highest-confidence individual miss, exit reason expiry. Listed as a single card, never as a model ranking.

## Calibration bridge

deepseek-v4-pro is both this week's hit-rate leader (44.6%) and the best-calibrated line (Brier 0.2760, gap +15.9pp). Per-model ok-rate this window ran 99.52% to 100.00%. The full confidence-bucket analysis lives in Weekly Calibration #4.

## Market check: the tape cooled off

> **MARKET CHECK · OBSERVATION (two adjacent calendar weeks)**
>
> OBSERVATION — mean |1d move| 4.28% -> 2.35% (-45% rel), BTC realized vol 58.7% -> 37.9% (ann., hourly); field directional accuracy 54.0% -> 42.9% (-11.1pp), 0 of 7 models improved, sim win-rate up for 0 of 7, field sim PnL +$889.37 -> -$473.59 (adjacent calendar weeks Aug 17-23 vs Aug 24-30; descriptive, one pair of weeks, not a claim).
>
> Robustness: raw price-sign accuracy 53.5% -> 44.2% — the same direction as the trade-based hit rule.

| Measure | Aug 17-23 (week A) | Aug 24-30 (week B) | Change |
|---|---|---|---|
| Mean \|1d move\| (5 assets) | 4.28% | 2.35% | -45% rel |
| BTC realized vol (ann., hourly) | 58.7% | 37.9% | -20.8pp |
| Field directional accuracy | 54.0% | 42.9% | -11.1pp |
| Models improving hit-rate | -- | 0 of 7 | -- |
| Raw price-sign accuracy | 53.5% | 44.2% | -9.3pp |
| Field sim win-rate | 46.3% | 37.0% | -9.2pp |
| Field sim net PnL | +$889.37 | -$473.59 | -- |

Week A is exactly the issue-#3 window, so this slice and every week-over-week column in this report compare the same two back-to-back weeks. Descriptive, one pair of weeks; the hit rule is methodology v1.1 (trade-based) and the raw price-sign row is the robustness check (hourly closes at :00 against slots at :01). BTC realized vol here is the hourly-return series from the alive slice; the snapshot row's figure is the audited daily-candle estimate -- different estimators, both reported as measured. Lineage splice ACTIVE: grok-4.5 rows are aggregated into grok-4.6 (flip 2026-08-24T09:22:00+00:00, inside week B). Week A (2026-08-17..2026-08-23) is pure grok-4.5 and reproduces the published issue #3; week B mixes the 4.5-era (Aug 24 00:00-09:22) with 4.6.

## Practical implications

- Do not read this week's leaderboard as a skill ranking: no title was awarded and the top CIs overlap.
- Treat per-ticker readability as weekly weather. Four issues, and the best ticker has changed hands every time (SOL -> BNB -> BTC -> SOL).
- Self-agreement lifts run -7.6pp to -1.9pp with small disagree cells; carry the regime split into the monthly test, do not act on it.

## Limitations

- Single week (Aug 24-30, 2026 UTC). Descriptive for this window only; no claim about next week or any model's underlying skill. This window ran 3 trend days of 7 (issue #3: 5 of 7); every comparison with issue #3 is a comparison across regimes as well as across weeks.
- Observations inside one window are not independent, and the 95% Wilson CIs shown throughout are descriptive, not inferential.
- The reversal detector remains v1 and BTC/ETH-only, on price series reconstructed from trade entry prices. 8 events across four issues cannot characterise detector performance either way.
- Cross-TF and cross-FH disagreement cells are small (n=18, 10, 25, 10); cells under N=10 are insufficient and never used to rank models. Week-over-week deltas are computed on unrounded rates and may differ by 0.1pp from the difference of the rounded columns.
- Model lines in this window: claude-fable-5, claude-opus-5, deepseek-v4-pro, gemini-3.1-pro, gpt-5.6-sol, grok-4.6 (spliced), qwen-3.8-max. Series density and lineage notes (wave 4): (1) rolling-1w forecasts emit daily since Aug 22, 2026 and rolling-1M daily since Aug 23, 2026, so inside this window (Aug 24-30) daily coverage is FULL for the first time in the series: the 1w series has slots on 7 of 7 days and the 1M series on 7 of 7 days; both sit outside this report's FH gate (1h/4h/1d) and appear only in the exclusions counter (1w 488, 1M 490). (2) The grok line flipped 4.5 -> 4.6 at Aug 24, 2026 09:22 UTC, INSIDE this window: every grok row in this issue is the lineage splice "grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22)" (110 rows from the 4.5 era, 1,360 from 4.6); issue #3 was pure grok-4.5, so the grok week-over-week row joins a spliced line to a single-model line. (3) qwen-3.8-max succeeded qwen-3.7-max on Aug 5, 2026 (fully before this window).
- Market-state row is computed on a single-exchange (binance) daily candle series, as in issue #3 after the audit that found the raw table mixes two exchanges; issue #3 values are as published.
- Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Aug 31 16:00 UTC) -- the definition pinned in issue #3, which printed 33,809 at the Aug 24 16:00 cutoff. The pack reproduces that pin at the previous cutoff (control OK), so this issue's 38,385 is an additive step under one definition; counter deltas against issue #2 and earlier remain definitional.

## Counters & lineage

| Counter | Value | Counter | Value |
|---|---|---|---|
| Forecasts total | 10,290 | Mature (scored pool) | 9,301 |
| OK in gate | 9,301 | -- of them directional | 4,576 |
| Out of gate (1w / 1M) | 488 / 490 | -- of them sideways | 4,725 |
| Invalid | 11 | Pending (next issue) | 0 |
| Late closes | 0 | Uptime, all grids | 266 / 266 slots |
| Source file | weekly_metrics_2026-08-24.json | Report cutoff | Mon Aug 31, 2026, 16:00 UTC |
| Generated at (pipeline) | 2026-09-02 06:14 UTC | Methodology | v1.1 (2026-08-10) · hash e66c7e8c864a2233 |

**THIS REPORT: 4,576 scored observations.** MARKETMANIA RESEARCH TO DATE (as of Aug 31, 2026 cutoff): 38,385 directional forecasts resolved since Jul 11 · 7 models tracked (current line-up; earlier versions folded into their successors' lineage) · 5 assets · 5 horizons · hourly · 17 published reports.

> **Issue #4.** Weekly Model Watch is a living series. Issue #3's open questions -- does the leaderboard reshuffle again, does the reversal detector find callers on a non-ETH turn, and does the every-ticker-above-51% floor survive? -- read this issue as: deepseek-v4-pro at the top (44.6%), 3 of 4 qualifying turns found a caller, and 0 of 5 tickers above 51%. The grok row is a lineage splice this issue (grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22)) because the 4.5 -> 4.6 flip happened inside the window. Engine 1.1 has powered the sandbox since Aug 18, i.e. before this window; no cross-engine PnL comparisons are claimed.

## What we're testing next

- Next issue: does any model hold the top rank two issues running, and does the title rule ever fire?
- Does the reversal detector find callers outside ETH?
- Monthly series: cross-TF/FH lifts with regime control across mixed weeks (trend vs. flat), block bootstrap.

## Related research

| Report | Direct PDF link |
|---|---|
| Consensus Watch #4 | https://marketmania.ai/research/reports/consensus-watch-2026-08-24.pdf |
| Weekly Calibration #4 | https://marketmania.ai/research/reports/weekly-calibration-2026-08-24.pdf |
| Weekly Model Watch #3 | https://marketmania.ai/research/reports/model-watch-2026-08-17.pdf |

The three weekly reports publish together as one issue each week; each links straight to the others' PDF and to its own previous issue. Direct links are the posting rule from wave 2 on.

## Cite this report

```bibtex
@misc{mm_modelwatch_2026w35,
  title  = {Weekly Model Watch #4: weekly leaderboard, ticker/pair reads and cross-confirmation, Aug 24-30 2026},
  author = {{MarketMania Research}},
  year   = {2026}, month = {September}, day = {2},
  url    = {https://marketmania.ai/research/reports/model-watch-2026-08-24.pdf},
  note   = {Methodology v1.1 (2026-08-10), hash e66c7e8c864a2233; window Aug 24-30, 2026 UTC; source weekly_metrics_2026-08-24.json}
}
```