bradthepilot

Have you done the math on what the current first order derivative is doing? How's the second order derivative going? That needs to turn around and what I was referring to, because if it doesn't the TSA screening numbers will start shrinking again within the next 14 days. I'm not predicting that - I'm not predicting anything - but applying simple math as you've done is not helpful for identifying trends.

If you do want to predict what traffic looks like at some point, look into how various forecasts are done and what a Gompertz curve is. That will be still be guesswork, but it'll be far closer to whatever ends up happening than will a simple exponential rise.

Excargodog

What I was responding to was YOUR statement:

This is what I said:

I was not never PREDICTING what the outcome would be, I was responding to your hope that the growth trend would reverse its decline and go back up, and demonstrating how improbable - essentially impossible - that a sustained growth rate week over week of over 19.4% is. Certainly there can be variance - we might see another growth rate of over 19.4% a few times, but whether we can define the exact limit or not, we are gradually approaching a limit where even the 19.4% can certainly not be maintained, far less reversing the trend to a higher growth rate. If you thought I was making some other point you misinterpreted my intentions.

RiddleEagle18

230,367
9.3% of last year
30% gain over last week

Same week over week gain last year was 5%


Sent from my iPhone using Tapatalk

md11pilot11

Predicting around 270,000 for the 21st. Should be a new record with just short of 300,000 on the 24th. Going to guess around 290k for the 24th


Sent from my iPhone using Tapatalk

ACEssXfer

He predicted Sweeden would have 40,000 deaths by May 1 and 100,000 by June 1. Sweeden didn't "do anything". Their current deaths are somewhere around 2800 based on the article. To call this political over reach is ignoring how wrong he actually was.

md11pilot11

I have circled the last two weeks Tuesday to Wednesday numbers. Obviously I can’t predict what is going to happen I have a very base level knowledge about these things but I wanted to get someone whose smarter than me to take a look. Each week follows a pretty clear pattern but if we look at today it seems to me that this Thursday through Sunday is going to blast right through 10% and 300,000.



Sent from my iPhone using Tapatalk

bradthepilot

I was referring to the second order derivate, which is a calculus concept. The first order derivative is the rate at which it the data is changing, which is the growth rate. The second order derivative is the rate at which the growth rate is changing. Prior to today's numbers, it was in an accelerating free fall. It's leveled out today, which is the change I was hoping for.

I think something like that needs to happen if we're going to get to 80% of 2019 numbers by Q1-Q2 next year. Again, check out what a typical growth curve looks like - it accelerates, stays steady for a while, then slows. We need some steady growth in the neighborhood of 20% for several weeks before it starts to slow again to hit those kinds of numbers. The curve will will be a sigmoid and will look something like the one below. In the graph below, the second order derivative increases then goes to zero for a while and the longer it is above zero the greater the rate of growth is during the second part of the curve. During the second part, the growth rate is essentially steady, which means the second order derivative is approximately zero. It stays that way until the third portion, where the growth rate slows - thus the second order derivative going negative - and the upper limit (whatever that may be) is approached.

Probably too much math. I went to school for engineering (lots of math) and later went back and added computer science (lots more math), and used both in modeling and developing control systems along with various other things before picking up this flying gig. Call it a force of habit.

Knobcrk1

Sweden also as of last week had a really big spike in deaths per capita. They were scrambling to put a stop on increase of the virus, again per capita.


As far as the Tsa numbers, there’s no chance we will have any kind of normalcy this year, there’s a pandemic going on wether people choose to believe it real or not. There has not been a successful drop yet just a flattening. Until we’re down to near zero cases most won’t fly. Vegas and Disney land will be there next year.

Excargodog

Seems like we are saying the same thing, just differently. At the start there is a geometric growth, but as you approach a limiting factor, the growth rate must taper off, yielding the curve you show. My point was that the initial growth rate was unsustainable, your point appears to be that you do not think we are yet out of the geometric portion of the growth rate. Seems like we were largely talking past each other rather than disagreeing.

The bottom line - in any event - is that we are coming back. The real question is :

At what point to we hit a limiting factor that (groan) “flattens the curve.”

4V14T0R

If the 2nd order is flirting with 0 (bouncing around either side of it) wouldn't that mean that the current growth rate is holding relatively steady or am I thinking about that wrong?

Red is growth rate for same day over week prior. Blue is the difference in growth rate same day from the week prior.





Sent from my iPhone using Tapatalk