Let's coin toss 100 times.
Let's put 'em in orbit and assume they survive reentry and land where we can evaluate heads or tails, so we have some flight time in this analogy.
Just for shits and giggles.
It takes a coin anywhere from two weeks to a month to land.
We have tossed all 100 coins, 25 per week over the course of a month.
Our stats could look like this.
Tossed coins = 100
Landed coins = 50
Tail landings = 4
To make it a bit more interesting, we could also say we are tossing 25 new coins per week right now, and that could change up, or down.
New coin tosses this week = 25
Right now, we know the outcome for 50 coins, and 4 of them landed tails.
(4/50) = 8 percent tails rate
We know at least 50 coins are in flight, outcome unknown.
It could be that every other coin toss lands heads. We could also say, there are no more tosses to consider. So let us add the 25 new tosses this week too.
(4/125) = 3.2 percent tails rate. This would be the best case.
It could also be every other coin lands tails too.
(54/100) = 54 percent! That is the worst possible case.
In reality, for this example, we know there are more tosses coming and we also could model poor testing and say we are not aware of all the tosses.
Some error factor could be computed to refine best and worst case too.
The better we are at collecting data, the more those outcomes will tell us.
The more time that passes, the more outcomes will tell us.
We may find tosses at night double risk of landing tails, for example.
Or maybe coins tossed mid week land tails consistently.
Ignoring that which has not come to conclusion biases the risk assessment toward lower risk with a larger error.
Low numbers with high error rates are not all that useful in this context.
Current global stats are:
Coronavirus Cases: 19,725,865
Deaths: 727,548
Recovered: 12,660,022
That's 5.7 percent deaths overall known outcomes with 7,065,865 cases in progress right now.
If people test positive, don't end up in the hospital and we have no death certs filed, they are recovered, but may be harmed in some way. Poor data on that.
Otherwise, we know they died.
Some argue a percentage of those deaths are with COVID causing a death due to other factors.
You can bias that overall 5.7 percent number up or down based on your own personal risk assessment, but dead is still dead, and catching the virus only adds to ones risk. These all count for purposes of policy and understanding danger.
I know at least one person who did that low risk assessment, went to a State with high case load and who has been on a ventilator for a month now.
They may live, and that looks increasingly likely, but will definitely not be the same after it is all said and done.
"I am unlikely to die" carries greater weight in that light, yes?
All I am really saying here is people would be wise to treat this with more respect than we may otherwise feel is warranted.
The basic reasons are simple ignorance (we are doing the science and it takes time) and lack of effective medicine at this time.
Both of those warrant the more firm, lower error numbers we get from actual outcomes.