Sensible and silly criticisms of modelling. (1/7)

Sensible: The model misses out some important feature of reality, or is needlessly complex. This is hard to get right, so usually has some validity. Epidemic models don't need to represent every individual, but they almost always need something beyond the simplest approaches.(2/7

Silly: The model is "too theoretical" or "just a model". People trust their lives to infrastructure models and use devices that rely on esoteric quantum phenomena. There's no science without models, we just need to get them right (3/7)

Sensible: The model is mis-calibrated to the data, or poorly fitted, or identifiability has not been checked, or uncertainty not properly quantified ... Also very hard, so usually a good point. (4/7)

Silly: The model predictions didn't come exactly true, particularly when the policy the model was simulating wasn't enacted. (5/7)

Sensible: The model has not been compared / averaged with independent approaches. Epidemic models are just catching up here. But the wisdom of crowds is powerful. (6/7)

Silly: There is "too much" modelling. Compared to the scientific effort expended on all sorts of other things, epidemic modelling is tiny. We probably need more of it, but arguably less "front and centre" and more auxiliary. (7/7)

PS Your own additions welcome!