I realize Thanksgiving is a distant memory at this point, but I keep finding myself thinking back toward that weekend and Delta’s huge number of cancellations that plagued holiday travelers. With capacity down so much, operational disruption hasn’t been much of a thing this year, and that’s one reason why this was so notable. On top of that, it’s rare that Delta ever blows it operationally, but it happened, and I wanted to see what more information I might be able to glean by looking through the data. Thanks to masFlight, I was able to dig in.
Let’s start with a look at the big picture. Below you’ll see a chart showing number of canceled flights by day around the Thanksgiving holiday. (Thanksgiving was November 26.) Now, I realize this might be subtle, so you may have to look really closely to see where the problem was.
Delta and Regional Partner Canceled Flights by Day

Or maybe you won’t have to squint at all. That is one big spike in cancellations with the worst coming on Thanksgiving Day itself.
Keep in mind that this chart includes regional partners. Don’t see them? No kidding. That’s because there wasn’t a day with more than 1 canceled flight on any of these airlines, so they don’t show up on the graph at this scale. Endeavor canceled none during the entire period. Republic had one on the 1st of December, and that’s it. SkyWest was the regional problem child with five scattered on five different days. To be clear, that is sarcasm. This was not a regional problem. This was a Delta problem.
My next thought was to look at the fleets. Delta keeps saying that it was a variety of factors that led to the problems, but ultimately it looks like an issue of being short of crews. Was this something that impacted all crews? Or just some?
I grouped the fleets together so we could evaluate where the problem children were. Sure enough…
Delta Completion Factor by Fleet Type

It becomes abundantly clear that the real problem here was with the narrowbodies. More specifically, it was the A319/320/321 and 737-800/900 crews. The A220s and 717s held up well, as did the 757s which are grouped with the 767s for pilot purposes. This makes sense. There have been so few widebody flights that I imagine there are plenty of pilots to operate them. But for narrowbodies, those widebody pilots are being retrained to fly narrowbodies(Update: I’m told this isn’t the case and the widebody pilots are mostly still on their aircraft, waiting for airplanes to fly) and there are a whole lot of training events that need to happen. It looks like Delta just didn’t have the crews it needed to fly those narrowbodies.
But was Delta able to keep this isolated to a part of its network? No. No it wasn’t. Though do keep in mind that unlike the ones above, the scale on this map starts at 65% to make it easier to see. In other words, it’s not as dire as it might look at first glance.
Delta Departing Completion Factor by Hub

Sure, some hubs held up better than others. Minneapolis, for example, was able to keep its completion factor above 90 percent on Thanksgiving. But that was still 93.1 percent; not a very good number. LA was at the other end of the spectrum with only 83.2 percent of departing flights completed.
I can poke around the data all I want, but it won’t tell me WHY all of these cancellations occurred. I suppose it doesn’t matter. As long as Delta knows what happened and can prevent it from happening again, that’s what we should care about. Sure, Delta was able to smooth things over with a Skymiles apology this time, but that doesn’t salvage the holiday for those who may have missed it, and it won’t help if it happens again. The burden is on you, Delta, to make sure this coming holiday season goes smoothly.