Hacker News Reader

Comment by dmurray | Hacker News Reader

The article assumes that Ryanair always gives middle seats to the first passengers who check in and don't pay for middle seats. The eventual conclusion is that you should check in later to avoid getting the middle seat. I don't see why the R analysis and pretty visualizations are necessary for that.

Sadly the article lacks evidence that this is the algorithm Ryanair uses. They could equally well hold back the middle seats and offer them to the passengers most likely to pay for an upgrade. In fact, since on a full flight the hypothesized algorithm (give middle seats to the first passengers to check in) should not improve Ryanair's outcome over the old algorithm (give middle seats to the last passengers to check in) I'd guess they are indeed doing something like that.

View on HN

The article specifically says that there is statistical evidence showing that Ryanair does, in fact, preferentially allocate base-fare paying users to middle seats. Though the article does not explicitly say that the evidence supports a particular variant of the algorithm, your particular alternative does not seem to alter the underlying reasoning here, which is that if you check in late enough, all the middle seats have been allocated. Once that happens, you are not going to be allocated a middle seat, regardless of how the airline allocates the remainder. (This could, however, change your odds of getting a middle seat earlier in the process, and whether a later check-in leaves you more prone to being bumped (which I assume Ryanair does unless it is forbidden by law) is another issue to consider.)

Update: I see that Ryanair does not overbook, which puts some perspective on the complaints about it - I don't care about meals but I do about the risk of being bumped.

Update 2: I overlooked one part of the alternative algorithm, but I think that it actually makes it less likely: if they are holding back middle seats on the chance of upselling to later arrivals, where are they going to put those people? It seems that they are likely to end up with a bunch of middle seats with no chance of upselling even to passengers who would have done so, or possibly having (even more) disgruntled passengers who paid to choose but got a middle seat anyway, depending on exactly how this scheme works.

Interestingly, Ryanair claim that they do not overbook their flights, and claim to be the only airline in Europe which don't. See section 9 here: https://corporate.ryanair.com/about-us/passenger-charter/

Yes - this seems to support that claim: https://conversation.which.co.uk/travel-leisure/flight-airli... This also mentions that EasyJet claims to limit its overbooking to about 1%. You are probably aware that overbooking has become a problem in the US - I don't know whether European airlines are allowed to do it to the same extent.

The compensation for being bumped is defined in consumer law in the EU, and it's very punitive - ~300 euros minimum.

Yeah, thanks to 261/2004 (https://en.wikipedia.org/wiki/Flight_Compensation_Regulation...) airlines generally have incentive to make sure people get to their destination ASAP. Although it's kind of a prisoner's dilemma at the gate when they ask for volunteers to take a later flight, if someone agrees to compensation which is less than the regulated amount the airline will of course take that.

At least if someone volunteers for less, no-one can claim to have been unfairly treated. I would argue that this is so even if the volunteer was unaware that the regulated amount was higher.

Overbooking doesn't seem to be a problem in Europe. In hundreds of flights I've never missed one because of overbooking and I don't know of anyone else missing one either. In the same time I've had 3-4 flights cancelled or rescheduled due to weather or mechanical issues.

The queues or electronic waiting lists for standby seats (people at the gate who aren't yet confirmed on this flight) are something I notice when travelling in the US that I don't really see in Europe.

You're right. I'm making an assumption of what Ryanair's system is based on the outrage people have at being deliberately assigned middle seats. Ryanair have admitted however that they prioritise window and aisle seats for paying passengers. My analysis took this to an extreme, which may not be the case.

And I agree, an R analysis is unnecessary but I've been telling people this conclusion for ages and they don't seem to get it!

Ryan air might use a system where the probability of getting a middle seat is biased upwards, but not quite guaranteed. In that case, there has to be a critical point for the bias where it switches from "Better check in late" to "better check in early". Because with no bias (the first simulation) we have "better check in early" and with maximal bias (second simulation) we have "better check in late". I'd guess that this critical point depends on the proportion of pre-booked seats.

I wonder if Ryan air balanced the bias so it lies roughly at this critical point. This comes with the 'upside' of being more fair, as this system can't be gamed by changing your check-in time.

Reply on news.ycombinator.com