DECLASSIFIED   DOSSIER M-2026-K0NG-002 REV 2.1  ·  PUBLIC RELEASE  ·  CC BY-NC 4.0
FIELD SUMMARY k = 27 N = 1,184 RESULT: DECISIVE

Godzilla wins.
The data is not ambiguous. A field summary of the 2026 meta-analysis

Across twenty-seven independent studies and 1,184 modeled engagements, the pooled standardized combat advantage of Gojira titanus over Megaprimatus kongensis is approximately five times the conventional threshold for a large effect. This is the executive summary.

Continue to findings Listen to the official anthem

Primary Outcome

Pooled Standardized Mean Difference (Hedges' g)

2.43
95% CI [2.21, 2.65]  ·  z = 21.7  ·  p < 0.001

Approximately five times the conventional threshold for a "large" effect (Cohen, 1988). The 95% prediction interval — the range of effects expected in a future comparable encounter — is [1.31, 3.55]. Entirely above zero.

Translated: no plausible future engagement between these two organisms is expected to reverse the direction of the advantage.

Yes, that's an effect size you can see from space.

Specimens Under Comparison

Subject A vs. Subject B

The two organisms under comparison occupy radically asymmetric morphological and bioenergetic niches. The asymmetry is not subtle.

Subject A  ·  FILE TH-17

Gojira titanus

Common name: Godzilla

Standing height108 – 119.8 m
Estimated mass~99,634 t
StanceBipedal
Primary armamentTrans-pharyngeal directed-energy emitter
Defensive systemPhotonic crystalline dorsal plates
Locomotion (aquatic)Native; sustained
ClassificationAPEX

Refs: Honda (1954); Serizawa & Brooks (2019); Tanaka (2024).

Subject B  ·  FILE GH-21

Megaprimatus kongensis

Common name: Kong

Standing height31.7 – 102 m
Estimated mass~158 t
StanceBipedal (facultative)
Primary armamentBlunt force; improvised tool use
Defensive systemIntegumentary (see Fig. 3)
Locomotion (aquatic)Not observed
ClassificationALPHA

Refs: Cooper & Schoedsack (1933); Lind & Ilene (2023).
Note: alpha ≠ apex.

The Evidence

Twenty-seven studies. One direction.

Every included study produced an effect size in favor of G. titanus. The pooled estimate (diamond, bottom) is the random-effects average. Bars left of the dashed line would favor M. kongensis. There are none.

-1 0 1 2 3 4 Standardized Mean Difference (Hedges' g) ← favors M. kongensis favors G. titanus → Study SMD [95% CI] Tanaka et al. (2014) 2.41 [1.92, 2.90] Honda & Ifukube (2016) 1.88 [1.42, 2.34] Wu & Mazzello (2018) 3.02 [2.51, 3.53] Serizawa Lab (2019) 2.77 [2.30, 3.24] Brooks et al. (2019) 1.95 [1.48, 2.42] Monarch Working Group (2021) 2.89 [2.41, 3.37] Apex Cybernetics (2021)* 1.04 [0.42, 1.66] Chen & Andrews (2022) 3.18 [2.68, 3.68] Hollow Earth Survey (2023) 1.72 [1.21, 2.23] Mark Russell Sim. Lab (2024) 2.71 [2.21, 3.21] Pooled (k = 27) 2.43 [2.21, 2.65]

Figure 1. Ten representative studies (of k = 27 included). Square area is proportional to study weight. *Apex Cybernetics (2021) flagged for Critical risk of bias — corporate-sponsored simulation with documented commercial interest in Kong's survival. The pooled estimate is unchanged when this study is excluded.   Full meta-analysis (6 pp, peer-reviewed) →

Mechanism

Four reasons it isn't close.

The pooled advantage is not the product of any single trait. It is the product of a layered asymmetry in armament, defense, energy economy, and habitat — each of which would be decisive on its own.

Mechanism 01 / Subject A

Atomic breath restructures engagement geometry.

Kong's grappling skill, tool use, and reach advantages all require closing distance. The trans-pharyngeal directed-energy emitter denies the close-range engagement window entirely. The fight is decided in the geometry, not the contact.

Documented yield: 0.8 – 1.4 Mt TNT-eq.  ·  Effective range: > 2,500 m

Mechanism 02 / Subject A

Photonic dorsal plates store and re-emit incident energy.

The dorsal plates are not decorative armor. They are a crystalline biocomposite that absorbs incident kinetic and thermal energy and channels it back through the trans-pharyngeal emitter. Hitting Godzilla harder makes Godzilla hit back harder. This is not a metaphor.

Component: crystalline biocomposite  ·  Confirmed re-emission pathway: yes

Mechanism 03 / Subject B

The comparable defensive structure in M. kongensis.

Presented here for completeness, in the same engineering register and to the same standard, is the corresponding defensive schematic for Subject B. The reader is invited to draw their own conclusions.

Stratum corneum: ≤ 2 mm  ·  Photonic re-emission: none observed  ·  Dermal armor: absent

Mechanism 04 / Subject A

The amphibious envelope.

Of the 27 studies in the meta-analysis, six took place primarily in aquatic terrain. All six favored G. titanus decisively. Subgroup pooled effect: g = 3.41 — the largest in the entire dataset. Kong does not swim. Kong sinks.

Subgroup k: 6  ·  Pooled g (aquatic): 3.41 [2.95, 3.87]

Anticipated Objections

A pre-emptive cross-examination.

// Common defenses raised by counsel for Subject B. Addressed in order.

"But Kong won in the 2021 film."
Narrative framing artifact. See Discussion §4.2 of the full paper: cinematic outcomes in studio-marketed ensemble productions are confounded by commercial considerations favoring shared protagonism. When kinematic subscales are extracted from the same engagement and z-standardized within study, the dyadic kinematic dominance still favors Subject A. → Meta-analysis §4.2, p. 5
"Kong has heart. Kong has spirit."
Spirit was not in the pre-registered inclusion criteria (PROSPERO CRD42025-K0NG-002). We accept that the meta-analytic literature on heart and spirit is currently underdeveloped, and we leave the construction of a validated instrument to future researchers. → PROSPERO CRD42025-K0NG-002, §3.1
"The axe. You're forgetting the axe."
We are not forgetting the axe. The axe is priced in. Tool-mediated engagements occur disproportionately in Hollow Earth terrain, which we report as a separate subgroup (k = 3, pooled g = 1.42). Even with the axe, with full Hollow Earth radiogenic amplification, in the most Kong-favorable terrain class we identified, the effect still favors Godzilla by 1.42 standard deviations. → Meta-analysis Fig. 3 (subgroup analysis)
"It's all hypothetical anyway."
So is everything in physics before the experiment is run. The hypothetical-zoological literature is no less rigorous than any other observational science for which controlled experimentation is logistically or ethically infeasible. We registered the protocol, declared the priors, weighted the studies by inverse variance, and ran the sensitivity analyses. The pooled estimate is robust to every deletion we tried. → Meta-analysis §3.3, leave-one-out analysis
"What about Mechagodzilla, Ghidorah, [other titan]?"
Third-party combatant presence was a pre-specified moderator. Effect after controlling for terrain: β = 0.18, p = 0.21. Not significant. The presence of a third titan does not meaningfully change the dyadic estimate between Subjects A and B. Subject A remains the directional winner irrespective of what else is on the field. → Meta-analysis §3.4, moderator analyses
"You're biased. You wrote a paper specifically to prove this."
Correct. This is also true of every meta-analysis: someone formed the question, then answered it. The protections against bias are not the absence of an investigator's opinion (impossible) but the pre-registration of methods, the documentation of analytical choices, the publication of the full extraction dataset, and the independent reviewer assessment of risk of bias. All four are present. → Meta-analysis §2 (Methods); PROSPERO record

Audio Supplement

The Official Anthem of the Pooled Estimate.

// recommended listening while reviewing §3.3 of the full paper

Hosted on Suno. Public release; no login required. If embed fails to load, the song is also accessible directly at suno.com/s/bEhtEJa49OClb9Y2.

Source Document

Read the full meta-analysis.

Everything on this page is a summary of a longer document. The full peer-reviewed paper contains the PRISMA flow diagram, the complete forest plot (all 27 studies), the subgroup analysis by terrain, the funnel plot and Egger's test, the GRADE summary of findings, and a 28-entry reference list.

Download full meta-analysis (PDF · 6pp · 586 KB)
Cite as
Harrington, E. W., Tanaka, H. K., Wu, J., Vivienne-Graham, M., & Serizawa, D. R. (2026). A systematic review and meta-analysis of combat outcomes between titan-class organisms: establishing the dominance hierarchy in the Godzilla–Kong paradigm. Journal of Hypothetical Zoology, 47(3), 214–247. https://doi.org/10.48174/jhz.2026.047.0214