How close are we to World War 3?
Think of this as a weather report for global conflict risk. Every day we read real-world tension between the world’s major powers and blend it with the historical pattern of when big wars have happened before. It’s an educated, transparent estimate — not a prediction.
0 = calm · 100 = imminent
The Doomsday Dial
Severe76.2 / 100
Multiple major powers are in open conflict with high nuclear or sanctions tension.
updated 9/14/2026, 5:50:36 PM
LowGuardedElevatedSevere
What’s driving today’s score
Six ingredients go into the number above. None of them alone decides it.
Conflict vs. cooperation in the news
54 / 100
How hostile or friendly recent real-world events between major powers have actually been.
434 severe events (Goldstein ≤ -7) across tracked dyads in this window. Dyads are combined weighted by sqrt(event count), so low-volume pairings can't swing the score.
Mood of world news coverage
57 / 100
How negative or positive the global news coverage of these tensions has been lately.
8 of 14 dyads have a usable own-media reading this run; see each flashpoint below.
Historical timing risk
+3 of 8
Big wars have often clustered around turning points in long economic cycles — this measures how close we are to one.
K6 (AI / biotech / green-energy wave (emerging, contested)) has no established peak/trough yet — early-wave periods carry structural uncertainty rather than a specific turning-point signal.
Unchecked-power risk
+5 of 6
How little the governments in today's flashpoints are constrained by voters, courts, or a free press — not left vs. right, just how free a hand each government has to act alone.
Least institutionally-constrained side of each tracked dyad, averaged: 81/100. Highest: PRK (90/100), VEN (86/100), YEM (84/100).
Nuclear tension
+10 of 10
How elevated nuclear threats and posturing currently are.
3.6% of sampled headlines across tracked-country outlets mentioned nuclear-related terms in the current feed sample (heuristic keyword scan, not NLP).
Trade & sanctions tension
+3 of 5
How intense today's trade wars and sanctions are.
1% of sampled headlines across tracked-country outlets mentioned tariffs/sanctions/trade-war terms in the current feed sample (heuristic keyword scan, not NLP).
Who has the material power
Background context, not part of the score above — who can actually afford to fight.
🇺🇸 United States 37.6%🇨🇳 China 11.8%🇷🇺 Russia 5.6%🇮🇳 India 3.3%🌐 Rest of world 41.7%
Share of global military spending, 2024 (World Bank). Global military spending rose 11.2% from 2023 to 2024 ($2385B → $2652B).
Who’s most likely to start it
Ranked by each country’s single most dangerous active flashpoint right now.
Where we are in the long economic cycle
Historically, big wars have often broken out around turning points like this one.
Schematic phase position (illustrative interpolation, not measured economic output) — red marker = now, currently in K6
K1 · Industrial Revolution wave (1780–1849)
Steam power, mechanized textiles, canals, early iron
Napoleonic Wars (1803–1815) · Revolutions of 1848
K2 · Railway / steel wave (1849–1896)
Railways, steel, telegraph
Franco-Prussian War (1870–71) · Long Depression (1873–96)
K3 · Electricity / chemicals wave (1896–1945)
Electricity, heavy chemicals, internal combustion, steel-hulled navies
World War I (1914–1918) · World War II (1939–1945)
K4 · Automobile / petrochemical wave (1945–1991)
Mass automobiles, aviation, petrochemicals, early electronics
Cold War proxy wars (Korea, Vietnam) — no direct great-power war
K5 · Information technology wave (1991–2020)
Computing, telecom, internet, globalized finance
No direct great-power war; proxy tension (Georgia 2008, Crimea 2014)
K6 · AI / biotech / green-energy wave (emerging, contested) (2020–?)
Artificial intelligence, biotechnology, renewable energy, advanced materials
Russia–Ukraine war (2022–), Gaza war (2023–), rising US–China tension — dating debated among economists
Today’s flashpoints
The major-power relationships we track, and how tense each one is right now.
🇮🇱 Israel – 🇮🇷 Iran
85 news events in the last 24h
20 severe
Tense
Own-media framing: The Times of Israel: no matching coverage · IRNA: -4.17
🇷🇺 Russia – 🇺🇦 Ukraine
780 news events in the last 24h
236 severe
Tense
Own-media framing: TASS: -1.02 · Ukrainska Pravda: no matching coverage
🇮🇱 Israel – 🇾🇪 Yemen
4 news events in the last 24h · limited data, treat with caution
Tense
Own-media framing: The Times of Israel: no matching coverage · Yemen (no accessible outlet)
🇰🇵 North Korea – 🇰🇷 South Korea
86 news events in the last 24h
20 severe
Tense
Own-media framing: North Korea (no accessible outlet) · Yonhap: -2.5
🇸🇦 Saudi Arabia – 🇮🇷 Iran
175 news events in the last 24h
21 severe
Tense
Own-media framing: Saudi Arabia (no accessible outlet) · IRNA: 0
🇺🇸 United States – 🇮🇷 Iran
358 news events in the last 24h
70 severe
Tense
Own-media framing: CNN: no matching coverage · IRNA: -2.5
🇰🇵 North Korea – 🇺🇸 United States
22 news events in the last 24h
7 severe
Tense
Own-media framing: North Korea (no accessible outlet) · CNN: no matching coverage
🇮🇳 India – 🇵🇰 Pakistan
12 news events in the last 24h
4 severe
Tense
Own-media framing: The Hindu: no matching coverage · Associated Press of Pakistan: no matching coverage
🇮🇱 Israel – 🇱🇧 Lebanon
147 news events in the last 24h
19 severe
Cooperative
Own-media framing: The Times of Israel: no matching coverage · Lebanon (no accessible outlet)
🇨🇳 China – 🇹🇼 Taiwan
49 news events in the last 24h
Cooperative
Own-media framing: Global Times: -1 · Taipei Times: +0.42
🇺🇸 United States – 🇨🇳 China
497 news events in the last 24h
22 severe
Cooperative
Own-media framing: CNN: 0 (thin sample) · Global Times: -2
🇺🇸 United States – 🇷🇺 Russia
146 news events in the last 24h
13 severe
Cooperative
Own-media framing: CNN: no matching coverage · TASS: +1.15
🇺🇸 United States – 🇻🇪 Venezuela
36 news events in the last 24h
Cooperative
Own-media framing: CNN: no matching coverage · Venezuela (no accessible outlet)
🇨🇳 China – 🇮🇳 India
123 news events in the last 24h
2 severe
Cooperative
Own-media framing: Global Times: 0 (thin sample) · The Hindu: no matching coverage