Sopwith

Technically correct, but I didn't want to get overly complicated. Use a log-normal distribution if you crave for precision. But for simplicity, take a population of arrival times (ignore the actual schedule), calculate the mean and standard deviation, add two standard deviations to the mean and you'll get the latest probable arrival time with 97.5% confidence. Or close enough for all practical purposes, as we say in the engineering world.

KenHamer

Only 2 SD?

I dunno... isn't Air Canada a six sigma outfit?

Adam Smith

Nice of you to agree with me, but would have been good if you didn't have to throw in an insult about cherry-picking, which is totally untrue, to diminish it.

But you're looking at this the wrong way, because it's not just about how many flights are delayed by less than the 10-15 minutes by which AC may under-block, because delays accumulate. If a narrowbody does ~5 segments a day when it's flying, here's how that might work out:

• Segment 1: 10 minutes over block, arrival 10 minutes late, officially on time
• Segment 2: start 10 minutes late, 10 minutes over block, arrival 20 minutes late, officially late by 5 minutes
• Segment 3: start 20 minutes late, 10 minutes over block, arrival 30 minutes late, officially late by 15 minutes
• Segment 4: start 30 minutes late, 10 minutes over block, arrival 40 minutes late, officially late by 25 minutes
• Segment 5: start 40 minutes late, 10 minutes over block, arrival 50 minutes late, officially late by 35 minutes

In this example, under-blocking by only 10 minutes has led to OTP of 20%. Only one of the flights was "late" by less than the 10 minutes by which the flights were under-blocked, yet 100% of the delays were caused by under-blocking.

This is obviously simplistic and assumes the plane doesn't have slack time at a hub built in. Some turns are not nearly so demanding. But there are also times when a plane is scheduled to fly even more turns all day on YYZ-YUL, YYC-YVR, etc and where it's even worse.

Obviously not every flight is 10 minutes over block time. And sometimes a delay on one flight can be made up on the next flight. But this is an illustration of how delays caused by under-blocking can compound throughout the day, having an outsized impact on OTP. And if flights are consistently under-blocked, that compounding impact will be meaningful when measured across the whole schedule.
​​
I have not had this experience. If anything, I've heard pilots apologize a number of times for small delays that don't even officially count as late.

Delays can be negative, it just means a flight was early. That being said, there's generally a practical limit to how early a flight can be given the amount of time needed in the air can only be compressed so much, so I agree that the distribution is not even. In fact, I noted that in a post upthread. (Although I was thinking F distribution rather than Poisson, but it has been a very long time since I took a stats class). I also linked to an interesting paper on delays that you gentlemen might find interesting.

SparseFlyer

Just assume the delay data is log normal distributed.

Then you can just ln() your data and work with a normal distribution. Make everyone's life easier, we already have it hard enough posting on the same topic every month lol

mellon

Air Canada has a shareholders meeting this week I wonder if they discussed being the worst in the world for OTP? Or of they had any plans to improve?

rankourabu

They ll get on that right after discussing fixing the broken J seats*

*assuming there is time leftover from counting all the profit money

WaytoomuchEurope

"Okay, Shareholders, we have a choice to make! I know we continue to expand our EBITDA each quarter and our earnings per share continue to rise, but there's a couple of troubling things we may want to address.

Firstly, there's the deflating seats - a real pain in the butt. Secondly, there's a monthly report that comes out showing that our OTP is one of the worst in the world.

The problem for us is that DOZENS of people on this online forum called FlyerTalk are upset about both. The seats are a relatively easy fix and in the grand scheme wouldn't cost us much, but I say we continue to ignore it. Fixing the OTP, however, will definitely affect our bottom line. Who thinks we should become less profitable to fix this problem so we don't get these complaints from dozens of anonymous people who probably wouldn't fly AC even if it were fixed?"

Sopwith

In an ideal world fixing the OTP would improve the bottom line. But then we have all those infrequent flyers who make their selections based on the absolute lowest stripped-down fare quotations without considering (a) all of the hidden add-ons; and (b) what they are actually buying.

How long will it be before airlines start charging resort fees?

RangerNS

And also FF who are smart enough to factor OTP into schedule and still are happy (enough) with AC.

Symmetre

Factoring OTP into the schedule was just reason for this FF's decision to abandon blind loyalty to AC

mellon

another reason to avoid AC, is IRROPS.

WaytoomuchEurope

Interestingly, this is one of the foremost reasons I stick with AC. As an SE they work pretty quickly to get my situation fixed. UA wouldn't do much for me in IRROPS but I'm confident they would move mountains for you, a 1K.

rankourabu

Easy to say when you dont have to fly for work domestically and/or dont live in a city with other options.... Apparently it seems that quite a few here have resigned themselves to the OTP, and are actually wasting their own time trying to account/plan for ACs unreliability.... And have managed to convince themselves that its ok ....

That they do. Every time. Except they dont have to do it as often as AC does for you.

WaytoomuchEurope

I'm a sample size of 1, and have only been a "FF" for 1 year and an occasional traveler for several years before that, so I would never apply my stats to anyone beyond me and I couldn't care less what airline a person prefers, however...

Roughly 150 AC segments
Roughly 35 UA segments

AC - 3 IRROPS, 2 weather, 1 mechanical - all fixed faster than I could hope
UA - 4 IRROPS, 3 weather, 1 mechanical - 2 of them dealt with to my satisfaction, 2 flubbed so badly I actively avoid UA

So for this guy at least, I'm flying the right airline - the one that has to fix their mistakes less frequently. I'm certain you feel the same for yourself. I'm glad UA takes good care of you. Happy flying!

mellon

since AC has a 60% OTP, and you had 150 segments you need to go to Vegas, since you only had 3 IRROPS