The weekend after Christmas is usually blissfully quiet for us at Cranky Network Weekly as airlines usually just do a little clean-up and let their network teams catch their collective breath. That was not the case for American this year which completely remade its schedule in several different ways. The overall theme of the schedule change seemed to be to improve the operation and invest in strategic flying above all, an interesting move for an airline that continues to suffer from a revenue disadvantage. At best, this likely won’t help that situation. At worst, it could hurt it. But the airline has decided that the benefits outweight the potential costs.
Some of the changes were minor. They got the April and mid-May schedule closer to accurate, put the A321XLR in the Boston and New York/JFK – San Francisco markets, and filed the premium-heavy B787-9P to fly from London to both New York/JFK and Los Angeles starting this summer. On a somewhat larger scale, American added one extra bank to Philadelphia in the evening, so that it could increase the number of international flights it runs on the airport’s limited international-capable gates. But those were minor compared to the two big changes that I will cover in greater depth this week: the quick and relentless ramp-up in Chicago and the de-peaking of the Dallas/Fort Worth hub.
Today, let’s start with DFW.
DFW is not lacking for runways, but as the hub has grown and weather has worsened, it has found itself in an operational nightmare more often than should make anyone comfortable. This isn’t exactly new. Under legacy American before the US Airways merger, the airline had gone to a “rolling hub” concept where it smoothed out the operations during the day to be more even. This was meant to be great for operational integrity, and it was helpful for maximizing utilization of employees and gates throughout the day. Here is what that looked like:
American DFW Departures by Quarter-Hour July 11, 2011
Data via Cirium
Sounds great, but you know what it was bad for? Selling tickets. Rolling hubs created longer connecting times, and flight displays often ranked what they showed by total duration of the trip. This penalized American, and so, post-merger, it was re-hubbed again to optimize for short connections.
This new re-peaked structure has stayed in place for over a decade. You can see what that looks like just by comparing the above chart to summer 2025.
American DFW Departures by Quarter-Hour 2025 vs 2011
Data via Cirium
For the last decade-plus, American has run a nine-bank hub at DFW. Some of those banks were directional, meaning planes would come in from the west and depart to the east. Others were omnidirectional meaning planes came from everywhere and turned around to go to everywhere. As DFW has now crept closer and closer to 1,000 daily departures for American — last June hit 921 average daily departures with more planned this summer — the airline has been faced with the prospect of operational gridlock. This summer showed that.
In July 2025, according to Anuvu, American canceled nearly 3 percent of its departures from DFW and saw only 61.4 percent of those departures that did operate arrive on-time (within 14 minutes of schedule). That is terrible, especially with so many short connections being made (er, missed).
Now, American is going back to the future, or, well, sort of. It is trying to relieve stress on the DFW operation by… depeaking once again. This time, however, it isn’t going to a true rolling hub. Instead American will convert from having a nine-bank hub to 13. And it will ditch the omnidirectional banking. That being said, there are still clearly defined hub banks. They are just more spread out.
American DFW Departures by Quarter-Hour 2026 vs 2025
Data via Cirium
I was going to compare the before and after for 2026, but that wasn’t (and still isn’t) a final schedule yet anyway. I thought it would be more useful to compare to another Monday last year to show what is really happening.
In the above chart, you can see that flights are scheduled out with no 15-minute period having more than 33 departures. That is a far cry from the old plan which had nearly 80 in that big morning bank last year.
Operationally, this will be very welcome, but it’s apparently not enough to ensure operational success. In its public announcement, American said it will also be increasing block time. Block is the scheduled time between departure and arrival, it’s what you see printed in the schedules. It has nothing to do with the time in the air, so if an airline adds more block, it can pad its schedule to buy better on-time performance. After all, if you leave late but have a lot of block padding, you can still arrive on time, and that’s what the official metric counts.
This isn’t generally a great plan for efficiency or for improving financial performance, but with American running a bad operation last summer, this isn’t a terrible idea to try and get things back on track temporarily. It’s something many airlines have done at different times.
All of this means trips will take longer. Is that going to be a problem? Or will it be a benefit as American has been touting? That’s a discussion for the next post where we’ll look into this in an even more granular way.
