Question
Under a single locked algorithm and same-method inputs, how does a transparent physics-style model score African Lion vs Bengal Tiger on Land — and what must we refuse to claim?
- Status
- EXPERIMENT
- Version
- Apex Physics algorithm v1.2
- Date locked
- 2026-09-07 (Asia/Kolkata)
- Matchup
- African Lion vs Bengal Tiger · Land
- Attack type (V1)
- Charge (KE = ½mv²)
- Channel
- @animalfightsimulation
Inputs (locked for Video #1)
Only variables used in the locked materials. Speeds tagged as commonly cited bursts, not GPS-collar peaks.
| Input | African Lion | Bengal Tiger | Notes / source |
|---|---|---|---|
| Mass | 190 kg | 190 kg | Adult-male means; tiger = central India framing (not Chitwan ~235 kg) |
| Weapon — bite CBS | 1768 N | 1525 N | Wroe et al. 2005, dry-skull canine bite strength — same study, same method |
| Speed (land burst) | 80 km/h | 65 km/h | Commonly cited burst, not GPS collar peak |
| Charge KE | 46 914 J | 30 970 J | ½ m v² with v = km/h ÷ 3.6 |
| Armor index (1–10) | 5 | 4 | DB defaults for V1 |
| Stamina index (1–10) | 4 | 4 | Both burst/ambush class |
| Domain fit Land (1–10) | 10 | 9 | Domain mult 1.50× / 1.40× |
Assumptions
- Equal mass (190 / 190) for a same-region-style comparison
- one bite method card only — never mix Wroe + alternate studies into one ranking
- Charge body KE for impact category (Strike paw KE deferred — engineering estimates only)
- v1.2 weights: mass 15% · weapon 20% · KE 20% · armor 15% · speed 15% · stamina 15%
Method
Share = value / (A+B); weighted = weight × share; domain multiplier = 0.5 + index/10; win % = renormalised adjusted scores. Alternate bite studies (e.g. Różycka 2024; van der Meijden 2023) are mentioned in research notes but not scored for V1.
Result (model output)
| Lion | Tiger | |
|---|---|---|
| Win probability (Charge, Land) | ~56.1% | ~43.9% |
Headline call under locked inputs: algorithm favours lion ~56% / tiger ~44% on Land. This is a transparent model output, not a claim of a wild fight result. Change an input → change the percentage.
Limitations
Model output, not a real fight. Same-method Newtons (Wroe 2005 dry-skull CBS). Masses = adult-male means (tiger: central India 190 kg, not Chitwan ~235 kg). Speeds = commonly cited bursts, not GPS peaks. Charge KE = ½mv² from body mass × burst. Weights as above. Change an input → change the %.
- Not a prediction of real animal behaviour
- Not scientific proof of combat outcomes
- Not veterinary advice
- No PSI “bite force” figures; no mixing methods; no “proven winner in the wild”
- YouTube views/subscribers are not scientific validation
Next question
Sensitivity cards (e.g. Strike KE, Chitwan mass) remain available as later “what if” experiments — not the scored V1 template. Interactive simulator on this website: NOT YET AVAILABLE — watch on YouTube and read this methodology card.