Funny, you ignore the French case that does have evidence. Regardless, thanks for putting your ignorance of exponential growth on display.
- Let's split the middle and say doubling of death is 2.5 weeks (your numbers, not mine)
- Let's also qualify several months as 12 weeks (3mo x 4wk/mo, fair right?)
- I think you also said 300-ish deaths per day from flu?
So, let's give the benefit of the doubt and say....increase in daily deaths by 25% for health department to realize something is wrong, so....approximately 75 additional deaths per day is your detection threshold (could be a little less, but it doesn't have significant impact, maybe by half a month)
Starting with patient 0's death as the start, and 2.5 weeks as 17.5 days:
- Wk2.5: 2 deaths (0.05 deaths/day)
- Wk5.0: 4 deaths (0.11 deaths/day)
- Wk7.5: 8 deaths (0.46 deaths/day)
- ...
- Wk27.5: 1024 (58.5 deaths/day)
- Wk30: 2048 deaths (117 deaths/day)
I'm going to treat this like a step function just to make the math easier - but you wouldn't really reach the alarm threshold until you reach between week 27.5 (6.9mo) and week 30 (7.5mo)
So, this, my friend, is why COVID
could go undetected for several months (or even greater than half a year).
Regardless, this isn't the point of my original question. All models have generally been based around day of first identified case more or less. And these models are all in the Feb/March timeframe. This is what I stated:
Models incorporate assumptions and new information - that's why they're refined over time. They don't need to be "hard" evidence based - how many assumptions you make dictate how accurate the models are, but doesn't preclude anyone from actually creating a model. You saying "show me evidence" is either a) merely a deflection on your part in that you don't have a clue on how to model this (in which case, your feedback is pretty worthless) or b) you do know how to model, but the model doesn't reflect what your narrative is (in which case, your feedback is even more worthless...since the longer this pandemic actually has been ongoing, the lower the risk of death becomes).
Now that there is a confirmed case in France of Dec 27th, and verbal evidence from the person having symptoms back until mid-Dec, there is merit to model this based on France data from Mid-Dec (or end of Dec) to now. What you will find is either that the infectivity rate is really really low, or the fatality rate is really low (since, it's been floating around France for 5.5 months now)
Now that you have your confirmed day 1 case in France and data from France, what excuse will you use this time? Can't wait!
Note that I am genuinely interested in what the pandemic's risk profile would be if it's been lingering around 2-3 months longer that what current models start at since this would provide a clearer picture on where in the pandemic we are and in general, how much at risk we are at as individuals.