Analysis 1 — Salvo composition: testing the Shahed-as-decoy hypothesis · recomputed September 2026 · n=1,161 attack nights
A widely cited claim holds that Russia uses Shaheds as decoys to saturate air defence and enable missile penetration. This analysis tests it directly.
Finding: Hypothesis not supported — 86.6% vs 88.5%
Mixed salvos combining drones with missiles (n=628 nights) are neutralised at 86.6%, against 88.5% for drone-only nights (n=450). The 1.9-point difference is within error margins. Mixed nights are larger on average (139 weapons against 125), so what little difference exists tracks volume rather than any decoy function. Recomputed on the corrected basis, (destroyed + not_reach_goal) / launched; the conclusion is unchanged from the original April 2026 run, the figures are not.
Finding: Missile-only nights are the hardest to defend — 74.4%
The lowest neutralisation rate of any salvo type (n=83 nights, 1,386 weapons), holding at 74.4% both inside and outside the EW-reporting window. Ballistic and cruise missiles remain substantially harder to stop than loitering munitions. This reverses an earlier version of this analysis, which reported 80% and described missile-only nights as the best-defended category; that figure was computed on the kinetic-only basis and was wrong. Sample is small, so treat as indicative.
Analysis 2 — Russian weapon mix evolution · recomputed September 2026 · Sep 2022–Sep 2026
Finding: Drones are now 97% of all launches
Drone-class weapons rose from 39% of launches in late 2022 to 97% in both 2025 and 2026. The Shahed-labelled share alone reads 93% for 2025 and 78% for 2026, but only because a growing number of drone rows are logged as "Unknown UAV"; counting all drone rows removes that artefact. Monthly volume peaked at 10,098 in May 2026, roughly 25 times the 2023 monthly mean of 401.
Finding: Missiles down to 3% of launches proportionally, not in absolute terms
From 61% of launches in late 2022 to 3% across 2025 and 2026. The strategic emphasis has shifted to attrition by drone mass, with missiles reserved for high-value precision targets. They remain the harder defensive problem: missile-only nights are neutralised at 74.4% against 88.5% for drone-only nights.
Analysis 3 — Geographic dispersion of Russian strikes · April 2026
Finding: Mean oblasts per attack tripled — ~3 to 8+ (mid-2025 to Mar 2026)
Wider geographic reach forces Ukraine to maintain air defence coverage across more regions simultaneously, creating resourcing pressure that volume attacks on a single axis do not.
Finding: Kharkiv most targeted — ~128 events
Kharkiv (~128) and Sumy (~115) lead due to front-line proximity. Kyiv (~96) and Dnipropetrovsk (~98) follow as primary strategic targets.
All analyses by Andro Mathewson, PhD Candidate in War Studies, King's College London. Data: Petro Ivaniuk / Ukrainian Air Force; USF Pidrakhuyka. Code: github.com/Androm2018. Working findings — not peer-reviewed.