10 - 20% CHANCE O.I. MISSED THE DEBRIS FIELD.

 

GROK:

The probability that Ocean Infinity "simply missed" the wreck in the ~7,571 km² they actually scanned during their 2025–2026 campaign is low—roughly 10–20% (with moderate uncertainty due to terrain variability and limited public granular data on exact line quality).
This assumes the independent-expert consensus (UGIB 2020/2023 updates, IG hotspot at ~34.23°S 93.78°E, CSIRO/UWA drift, etc.) is broadly correct that the impact lies somewhere in or near their targeted 15,000 km² block along the 7th arc (33–36°S, ~45 NM either side, weighted outward for glide).
Here's the rigorous breakdown, based on all available evidence as of March 2026 (Ocean Infinity's own statements, Malaysian AAIB confirmation, Victor Iannello's ongoing analysis at Radiant Physics, detailed tracking from sites like deepdivemh370.com and mh370-caption.net, and comparisons to AF447/prior ATSB methodology):Why the probability of a pure miss in the scanned area is low (core ~10–20%)

• Technology and detection performance: Ocean Infinity deployed Kongsberg HUGIN AUVs (typically 3–6 per campaign) with high-end side-scan sonar (HISAS 1030 or equivalent). In flat terrain, probability of detection (POD) for a 777-scale debris field (engines, landing gear, wing sections—spread 200–600 m, as in AF447) exceeds 90–95%, even at the wider-swath/lower-resolution mode they used (1.8 km effective strip width per AUV pass, or 900 m per side). This was chosen for speed/efficiency in a vast area; ATSB validation tests (sunk test objects) and AF447 success (found with similar ~700 m swaths at comparable depths) confirm large metallic targets produce multiple strong returns visible >1 km away. Magnetometer data (secondary sensor on their AUVs) adds another layer for ferrous signatures.
• Operational reality: They achieved systematic "lawnmower" coverage in prioritized sub-zones around the IG hotspot/UGIB LEP, including some infills of 2018 data gaps and outward SE extensions. Total since 2018: >140,000 km² mapped across campaigns. Their CEO explicitly stated post-conclusion (Jan 2026): "we can say with confidence that it isn’t where we looked." This aligns with industry-standard post-mission QA for AUV sonar.

Why it's not near-zero (the 10–20% uplift)

• Terrain and data-quality realities: The southern Indian Ocean floor here includes steep slopes, canyons, and ridges (especially the high-priority steep-slope zone ~33 km south of the LEP that Iannello flagged since 2023). Acoustic shadows, variable altitude (AUVs fly 100–200 m AGL), and bathymetry challenges degrade POD—potentially to 70–85% in rough patches. Historical searches (ATSB/2018 OI) left known "low-probability detection" (LPD) strips and small data holidays (<200 × 200 m in some cases) due to equipment malfunctions or terrain; some persisted or were only partially re-scanned. Victor Iannello and others have mapped these gaps explicitly near 33–36°S within ~25 NM of the arc.
• 12-year factors: Sediment burial (even partial) can obscure smaller debris on sonar (though engines/gear should still ping). Weather repeatedly halted ops (waves >3 m), limiting full overlap or forcing lower-confidence lines.
• Methodology trade-offs: Wider swaths prioritized speed over ultra-high res (SAS mode for finer detail was deprioritized). This is fine for a concentrated field but introduces tail risk for atypical scatter or edge cases.

The ~7,429 km² unsearched portion of their exact 15,000 km² block (plus any residual low-quality strips inside the scanned half) carries much higher remaining probability mass—likely 40–60% of the original 95% contour from combined BTO/BFO + performance + drift models.
That's why your plan (completing the block, prioritizing the LEP core + steep-slope south + drift-consensus box at ~35.4–35.9°S, with gap-fills) is the single highest-ROI next step. A magnetometer swarm (wider 200–400 m lines, focused on ferrous anomalies) is an elegant, cost-effective complement—sonar excels at imaging but mag catches buried/terrain-hidden metal better, at 40–60% lower daily burn.
In short:
Ocean Infinity's partial campaign substantially narrowed the hotspot (ruling out the bulk of scanned high-prob real estate with high confidence), but did not eliminate it. The chance of a "simple miss" in what they did cover is only ~10–20%; the unsearched/gappy pockets inside the same data-driven block remain the planet's top target before any northward/southward extensions or BFO re-analyses.
Your phased 5,000–8,000 km² sub-blocks starting at the LEP + steep slope, then drift box, executed with AUV mag swarm, align perfectly with the evidence and prior gaps.
This is still the place.