You know what you should be asking?

Why haven’t the media been covering modeling? Why haven’t they asked actually people doing the work (not x professor in slightly related field) what goes into the PROCESS of creating a model

1/thread

I’m not surprised that places like CNN haven’t asked these, but you should be concerned that no conservative media has bothered to ask any of the following.
Question that have not been asked:

1) How is the data accumulated for a model? How many sources?

2/x

- How do you determine if the data is valid?
- What do you do if there is bad, inaccurate, incomplete, or missing data?
- How do you clean data?
- How can the public tell if data in a model is reliable?

3/x

- How many sources of data does *this specific* model use?
- How do you account for differences in documenting data from various sources.
- What data is useful?
- How much of this data is in a standardized format from the sources?

4/x

Notice how I haven’t even gotten to the MODELING stage of it yet?

- What do you do to the data before you put it in the model?
- What is a model?
- What is a model good for? What does it do?
- Is any one model the best?
- What makes up a model? What are the parameters?

5/x

- How do you decide what parameters to include in a model?
- why goes into deciding what kind of model it will be?
- How is a model programmed? Related, how is a model updated or debugged?
- are there different types of models and what are they?

6/x

Notice, again, how we haven’t even gotten to this being a VIRAL model forecasting human behavior, and outcomes.

And I’ve only given SOME questions.

- Is there any virus or past situation that allows us a baseline for how to create this model?

7/x

- How does one create the model parameters for an unprecedented virus? Or an epidemic? Or a pandemic?

- What did you decide for *this* model to do differently than other models in the face of no prior history to work from?

8/x

- What have been the accuracy rates of past models created in unprecedented situations (literally the first ones to model a particular situation)?
- What is the Make-up of the team creating the model? (Epidemiologist, Data Scientists, Drs, Behavioral Scientists, Virologists)

9/x

- What are the strengths and weakness of *this* model?

- What is the difference between a worst case scenario projection, an average projection, and a light projection?

- What is a Meta-model? How many models goes into the Meta-models the Task Force uses?

10/x

- How many days behind the real-time data updates is this model (read: how long does it take to clean and put new data into the model)?

- Has Social distancing ever been modeled at this scale?

- What do you do when it hasnt?

11/x

- How would you compare this model to other models out there?

- How can this model be better?

- What habe we learned about modeling new scenarios and unprecedented situations so far?

- What is the list of models that go into the meta-models the task force uses?

12/x

There.

I could go on. But I think I’ve made my point.

I’ve not seen one goddamn person in the media ask these question. Not one liberal, independent, or conservative.

It’s a disgrace.

Yet they want to complain and get those clicks.

I’m not going to blame the non-media people who trust the media, even if I think it’s naive.

But I will blame the media, all of those in it, who are currently specifically railing on about models without ever having asked even one of those questions

You should be ashamed
/fin