bradthepilot

If the rate of growth keeps changing at the average pace it has over the last few days, it will be in negative territory within two weeks which means actual screening counts start going back down. I think we’re past the point where using the rate of change to identify a bottom is helpful. Rather, it identifies inflection points and that shows a growth slowdown is absolutely underway. With changing conditions another inflection point or ten is almost certain it it’s guesswork to predict where that might be.

kingairfun

This^^^^^^

Realistically there's no where for people to go at the moment. States are generally open, but who's gonna go on vacation to be regulated to 25 or 50% capacity, have to worry about "rules" of social distancing, and with bar's not being open.... If some form of normalcy occurs in travel areas, especially Disney and other parks, IMO a pretty large spike will occur.

I think there is pent up demand and not nearly as many people are still worrying about dropping dead on the spot, but for now there is still way too much hassle to make traveling worth it.

kingairfun

I think looking at M-Su daily pax load trends, and finding the trends in those days over several weeks, is useful..
We really don't have much data history to draw from yet..

Seeing a 25% gain one week vs. the last is probably not a very good indicator as the numbers grow larger.

It will be a slow climb at first, barring any huge inroads to a vaccine announcement. Without a vaccine, we'll have to rely on people's propensity to "move on" from disasters over time.. Question is how much time and will it be soon enough for management?

Excargodog

190,477 yesterday versus 163,205 the Tuesday previous. That’s a solid 16.7% increase in a week. If that were to continue until 1 October we could expect to see something on the order of (1.167)^18 x 190,477 which would equal 3.07 million.

Of course, it won’t keep up at a 16.7% increase - not if more destinations aren’t opened (seriously, who wants to go to Hawaii and spend two weeks looking out the window of your hotel room?) - and International flying will be a particularly difficult nut to crack. But for airlines that don’t do a lot of Hawaii or international flying, it really is looking fairly promising to be avoiding any sort of massive furloughs.

LoneStar32

Until International opens back up fully, there is a ceiling much lower than last year's TSA numbers that we will hit. And yes, that will negatively impact NB and RJ numbers too.

velosnow

The thrust of the criticism was using the worst case numbers in the event of unmitigated action and saying the model was wrong. That still doesn't add up.

Debate among scientists on how or not to code a model is most certainly fair and expected. There will always be academic back & forth in nearly every study conducted. A lot of the articles concerning this seem to be conflating professional disagreement to outright malfeasance. That isn't the case either. We all know how bad reporting on aviation can be, so trying to decipher reporting on something none of are truly experts in is challenging. From this article...

"Epidemic modelers are the first to admit their projections can be off. “All models are wrong, but some are useful,” statistician George Box supposedly once said—a phrase that has become a cliché in the field."

https://www.sciencemag.org/news/2020...owns-and-other

The model for the UK ranged from 5,600 deaths to 550,000. But no one likes to talk about the low number, it isn't sexy in supporting the argument on how 'wrong' the IC study was. There are now over 35,000 deaths in the UK with the strategies that have been put in place. Sweden as we know took more of the herd immunity approach and have a higher death rate over their neighbors. Which approach is better over time we'll see but you can't claim other countries didn't fare worse. It also isn't an apples & oranges comparison with so many different health care systems, demographics, etc.

The IC study like so many others were upfront about the challenges they faced in modeling and the potential for error. I'm just not on board with trying to publicly hang academics for doing their work. Stating the work has been judged and deemed unworthy by posting blog posts of other scientists isn't conclusive. If there is room for future improvements then by all means learn from it and make the next one better.

Ferguson's resignation over his misconduct is both fair and human, I won't argue that. No doubt being in the position he was he should've stayed home and played five on one instead of a booty call.

galaxy flyer

The model for the UK ranged from 5,600 deaths to 550,000.[/QUOTE]

Any “model” where the range of error is orders of magnitude wide is useless. Like the SARS model which was ignored to no great harm.

Excargodog

A ceiling lower I have always granted. “Much” lower I’m unsure of. If you have been able to find a source that indicates with any reasonable degree of accuracy what percentage of the TSA numbers are attributable to international passengers please provide it. I’ve looked and haven’t found such a source.

Excargodog

And any computer model which - when fed the same data on multiple occasions produces different answers each time - is particularly suspect. At a bare-ass€d minimum you need repeatable results from any predictive program except perhaps a random number generator.

LoneStar32

You're right, every international traveler lives in an international hub and doesn't need to fly a RJ or NB to get there. You win the internets again, congratulations! You can now have your thread back.