In many cases, positive predictions are impossible. But the
impossible might be predicted, like in thermodynamics. Models can
help to map the impossible and thus limit the possible
[log in to unmask]">I agree with Gary in so far as "predictions" are as possible as dangerous: Our models cannot guarantee that their "predictions" will come true at any rate as this is only possible for very simple systems, but they can shed light on what could happen if certain empirically measured initial conditions are input to the model and the model describes the mechanism which leads to an outcome. Currently we do not seem to know enough about the mechanism of the current epidemic (most unknown parts of this mechanisms have already been named here), but various conjectures about the mechanism are certainly possible. And the empirically measured initial conditions are far from precise enough (as you never know whether yesterday's new infections were reported yesterday or only today or tomorrow, or how the patients were tested in this or that country, let alone who infected whom and who sat next to whom, and too near). "Predictions" claim to say what will happen, perhaps with a certain probability distribution — projections only say what might happen if the initialisation and the mechanism are reliable enough and can also say what would happen if alternative initialisation and/or mechanisms were assumed.
It might be good to reduce simulationists' claim to projections ...
Best
Klaus
By the way, as for the data: Day after day a approximate a logistic function to the covid-19 data, and day after day the approximation is surprisingly close, but the next day the coefficients of the logistic function need to be changed considerably to achieve the same precision as yesterday ... and much less do we know about the micro level generating these macro data as they are published with some differences by WHO, worldometer or Johns Hopkins!
Prof. (i.R.) Dr. Klaus G. Troitzsch
Schillerstraße 26
79713 Bad Säckingen
oder
Aeussere Dorfstrasse 39
3718 Kandersteg, Schweiz
Telefon +49 7761 9982074/5 bzw. +41 33 6750166
[log in to unmask]
https://www.uni-koblenz-landau.de/en/campus-koblenz/fb4/iwvi/former-rg/rgtroitzsch/klaus-troitzsch
https://klausgtroitzsch.academia.edu/
########################################################################
To unsubscribe from the SIMSOC list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=SIMSOC&A=1