Extreme events theory

Florian Angeletti
NITheP

Not enough data, information and knowledge
Perfect and reliable prediction is impossible
Probabilistic methods:
weak but reliable prediction

Prediction

Random person: ≈ 0.273 %
Audience of 100: ≈ 24.0 %
Common birthday: ≈ 99.99997769%

Rules:
- Roll n dices
- Keep the highest score

Rules:
- Roll n dices
- Reroll 6, add 6 to their result
- Go on until their is no new 6
- Keep the highest score

Rules:
- Roll n dices
- Reroll 6, multiply by 6 their result
- Go on until their is no new 6
- Keep the highest score

Weibull

Upper limit

Gumbel

Logarithmic growth

Fréchet

Power growth

Limitations

Exact and non-interacting repetitions
- Trend detection?
- Interactions?

Random daydreaming physicist
Physicists don't teleport
Physicists don't wander too far

My work

What happens when a physicist wanders far away?
Mathematics extremoscope to describe this situation

Thank you

Questions?