Stories
Slash Boxes
Comments

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 12 submissions in the queue.
posted by martyb on Sunday July 02 2017, @05:45PM   Printer-friendly
from the dicey-analysis dept.

Passengers have more chance of winning the National Lottery jackpot than being allocated middle seats at random on a Ryanair flight, according to new Oxford University analysis.

In recent weeks Ryanair have faced mounting customer criticism, with some accusing the airline of splitting up groups and families, who do not pay an additional charge for reserved seating. These claims have been rejected by the airline which says that customers who do not wish to pay for their preferred seat are randomly allocated one, free of charge.

Last night, the BBC Consumer affairs programme, Watchdog, ran its own investigation to test how random the airline's seating algorithm is.

As part of their tests, groups of four people were sent on four separate Ryanair flights. In each instance every single person was allocated a middle seat. Dr Jennifer Rogers, Director of the new Oxford University Statistical Consultancy was then invited to analyse the data, to work-out the chances of every person getting a middle seat allocated randomly.

By looking at the amount of window, aisle and middle seating available on each flight, at the time of check-in, Dr Rogers, calculated the chances of all four people being randomly given middle seats on each of the flights, to be around 1:540,000,000. The chances of winning the National Lottery jackpot are 1:45,000,000. (This means that you are 10 times more likely to win the lottery than be in a group who are all randomly allocated middle seats.)

Source: Oxford University


Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 1, Informative) by Anonymous Coward on Sunday July 02 2017, @05:59PM (3 children)

    by Anonymous Coward on Sunday July 02 2017, @05:59PM (#534195)

    For the most part, flying on European or Asian airlines is still beats flying on 'any' U.S. based airline by a wide margin. I've used Ryanair before while in Europe and I was treated like a King compared to even flying business class in America. If you are flying on an economy ticket in the U.S., the airlines look at you as worth less than dirt. Good service in the U.S. means making it to you destination without being forcefully dragged off the plane.

    • (Score: 0) by Anonymous Coward on Sunday July 02 2017, @06:50PM (2 children)

      by Anonymous Coward on Sunday July 02 2017, @06:50PM (#534208)

      Guess it's better to stay in Europe then..
      Come to think of it. There's a whole world outside US ;-)

      But those Russky airlines with drunk pilots are something special. British Airways also have their incidents, but they are very few compared to Russky.

      • (Score: 2) by Gaaark on Sunday July 02 2017, @09:01PM (1 child)

        by Gaaark (41) on Sunday July 02 2017, @09:01PM (#534250) Journal

        That's why im glad i live in Canada, though we're so close to the U.S. that all their bad things seem to waft over to us. We just got rid of our own Bush/Trump (Harper) and now we got the new, geeky, lying Trudeau, son of 'Put Canada deep-deep into debt' Trudeau.

        Sigh. I guess Canada has shit of it's own... too bad America's shit over-flows onto us sometimes as well.

        --
        --- Please remind me if I haven't been civil to you: I'm channeling MDC. I have always been here. ---Gaaark 2.0 --
        • (Score: 0) by Anonymous Coward on Sunday July 02 2017, @10:51PM

          by Anonymous Coward on Sunday July 02 2017, @10:51PM (#534277)

          reason I chose not to relocate up there.

          If any of the northern colonies were to defect and reinstate privacy, limited copyright enforcement, no internet spying, etc, I'd move up and become a colonist up in the far north.

          Too bad the populations there aren't big enough to either defend themselves, or get supplies in without southern canada's infrastructure and financial support :(

  • (Score: 1, Informative) by Anonymous Coward on Sunday July 02 2017, @06:13PM (2 children)

    by Anonymous Coward on Sunday July 02 2017, @06:13PM (#534199)

    Oh come on. "Allocated randomly" does not mean that the results are necessarily uniformly distributed. As director of a statistical department, Dr. Rogers should know better.

    • (Score: 5, Insightful) by Mykl on Monday July 03 2017, @03:25AM (1 child)

      by Mykl (1112) on Monday July 03 2017, @03:25AM (#534323)

      This article should have the by-line of 'Duh'.

      • Those people who pay to select seats would vastly reduce the number of non-middle seats out of the 'random' pool, so assuming a random distribution across the whole plane is foolish
      • The sample size of 16 is statistically invalid anyway - not only for all flights on the day of the test (which would have been tens of thousands of passengers), but in fact for each individual flight as well (4 out of a couple of hundred)
      • The article is poorly worded, suggesting that getting a middle seat is a 'good' outcome.
      • (Score: 1) by chair on Monday July 03 2017, @03:54PM

        by chair (6194) on Monday July 03 2017, @03:54PM (#534466)

        Those people who pay to select seats would vastly reduce the number of non-middle seats out of the 'random' pool, so assuming a random distribution across the whole plane is foolish

        from the summary:

        By looking at the amount of window, aisle and middle seating available on each flight, at the time of check-in

        ---

        The sample size of 16 is statistically invalid anyway - not only for all flights on the day of the test (which would have been tens of thousands of passengers), but in fact for each individual flight as well (4 out of a couple of hundred)

        From the (very short) FA:

        To support her analysis, Dr Rogers was also given access to data from a further 26 individuals, from nine groups,

  • (Score: 4, Touché) by maxwell demon on Sunday July 02 2017, @06:34PM (21 children)

    by maxwell demon (1608) on Sunday July 02 2017, @06:34PM (#534204) Journal

    Passengers have more chance of winning the National Lottery jackpot than being allocated middle seats at random on a Ryanair flight

    Sounds good. Who wants to sit on a middle seat anyway?

    As part of their tests, groups of four people were sent on four separate Ryanair flights. In each instance every single person was allocated a middle seat.

    Wow, I didn't know that the chance of winning the National Lottery jackpot are that high. I guess I should play the National Lottery. ;-)

    --
    The Tao of math: The numbers you can count are not the real numbers.
    • (Score: 5, Informative) by deadstick on Sunday July 02 2017, @06:50PM (20 children)

      by deadstick (5110) on Sunday July 02 2017, @06:50PM (#534207)

      Sounds good. Who wants to sit on a middle seat anyway?

      Whoosh...

      Point is, if you don't pay the airline for an assigned seat, they stick you in the worst kind...and the analysis shows that it's deliberate.

      • (Score: 2) by Adamsjas on Sunday July 02 2017, @07:19PM (1 child)

        by Adamsjas (4507) on Sunday July 02 2017, @07:19PM (#534214)

        Still, the first quoted sentence is wrong.
        Your changes are very very good of being allocated the middle seat in their version of "random".
        The sentence implies that your chances are nil, does it not?

        • (Score: 1) by Roger Murdock on Sunday July 02 2017, @10:23PM

          by Roger Murdock (4897) on Sunday July 02 2017, @10:23PM (#534273)

          The sentence implies your chances are virtually nil if the seat allocation is random and if the pool of seats from which the selection is made contains all of the available seats on the flight, which apparently isn't the case. There's nothing wrong with the sentence. Incidentally, it looks as if they did take into account the popularity of window and isle seats (to a certain degree) by doing the analysis at check in time.

      • (Score: 2, Insightful) by Anonymous Coward on Sunday July 02 2017, @07:23PM (9 children)

        by Anonymous Coward on Sunday July 02 2017, @07:23PM (#534215)

        Is it though? If some people are buying reserved seats in the window or aisle, then the odds that the remainder get middle seats go way up. The article did not factor this into their math.

        • (Score: 3, Interesting) by kaszz on Sunday July 02 2017, @07:45PM (3 children)

          by kaszz (4211) on Sunday July 02 2017, @07:45PM (#534222) Journal

          And if the statistical truth about people is that they pay for window seats etc. Then you want non-paying seat customers assigned to seats where they will not occupy anything that you can sell. But once the non-desirable seats are full. Then even window seats etc will be filled with non-paying seat customers.

          Algorithm 101?

          Perhaps RyanAir should state "customers that don't pay for a seat allocation will be allocated to seats others historically didn't wanted to pay for. And will then be allocated from least wanted location to the most wanted, in that order."

          Kind of reminds me of the keyboard order of keys.

          • (Score: 4, Interesting) by tfried on Sunday July 02 2017, @08:22PM (2 children)

            by tfried (5534) on Sunday July 02 2017, @08:22PM (#534231)

            Perhaps RyanAir should state "customers that don't pay for a seat allocation will be allocated to seats others historically didn't wanted to pay for. And will then be allocated from least wanted location to the most wanted, in that order."

            Or put shorter: "If you don't pay up, we'll assign you the worst seat we can." Which does not actually stand out as unprecedented evil, among the many schemes to look good on price comparison sites, while making sure customers will actually pay much more than they expected. Getting caught lying about it is still (and will hopefully remain) newsworthy.

            • (Score: 0) by Anonymous Coward on Monday July 03 2017, @10:38AM (1 child)

              by Anonymous Coward on Monday July 03 2017, @10:38AM (#534407)
              There's a difference between Ryan Air purposely giving you the worst seats and you just getting the leftovers.

              If you don't pay extra to pick a particular seat, from a coding and business perspective it might actually be easier and better to not assign you any seats till much later.

              So the ones who pay extra get their choice of seats first, and are likely to take the aisle or window seats and the ones who don't pay extra get what's left which tend to be crappier seats or maybe no seats even - bumped off.

              If the study didn't take stuff like that into consideration then the study is worse than Ryan Air ;).
              • (Score: 2) by AthanasiusKircher on Monday July 03 2017, @02:51PM

                by AthanasiusKircher (5291) on Monday July 03 2017, @02:51PM (#534446) Journal

                If you don't pay extra to pick a particular seat, from a coding and business perspective it might actually be easier and better to not assign you any seats till much later. [...] If the study didn't take stuff like that into consideration then the study is worse than Ryan Air ;).

                From the summary:

                By looking at the amount of window, aisle and middle seating available on each flight, at the time of check-in, Dr Rogers, calculated...

                Almost all airlines assign you a seat at check-in if you haven't already chosen/been assigned one before (unless you're flying standby or something). So, yes, it seems they took into account the fact that seats were assigned at the last moment before a final ticket was generated.

                The point seems to be that the algorithm isn't randomly choosing among remaining available seats -- it's filling up the "less desirable" seats first. And from a business perspective, that's probably a good policy, but it's not the "random" selection process that the company is advertising.

        • (Score: 3, Interesting) by Joe Desertrat on Sunday July 02 2017, @09:33PM (1 child)

          by Joe Desertrat (2454) on Sunday July 02 2017, @09:33PM (#534257)

          It is the same sort of way hotel rooms get allocated. Buy the cheapest one or get one through a discount service (Expedia, Priceline, etc.), you get a room facing the street, wall, whatever is less desirable in that location. Pay more, you get the oceanfront, higher floors, whatever constitutes good in that location. Except in hotels they usually try to sell you an upgrade at check in.

          • (Score: 2) by driverless on Monday July 03 2017, @04:34AM

            by driverless (4770) on Monday July 03 2017, @04:34AM (#534337)

            It is the same sort of way hotel rooms get allocated. Buy the cheapest one or get one through a discount service (Expedia, Priceline, etc.), you get a room facing the street, wall, whatever

            Or one with a dumpster fire below the window. Literally. Made for quite an exciting first night in SFO.

        • (Score: 2) by driverless on Monday July 03 2017, @04:51AM

          by driverless (4770) on Monday July 03 2017, @04:51AM (#534342)

          If some people are buying reserved seats in the window or aisle

          Those do seem to be the preferred seats. When I fly I get asked whether I want a window or aisle seat. I don't recall ever being asked whether I want a middle seat. So it does seem that those are the ones that'd get left over once all the preferred/paid-for-allocation seats are taken.

        • (Score: 0) by Anonymous Coward on Monday July 03 2017, @05:00AM (1 child)

          by Anonymous Coward on Monday July 03 2017, @05:00AM (#534346)

          What idiots are reserving an isle and a window seat and leaving the middle for some random joker? You are probably forgetting groups of people will much prefer to reserve adjacent seats, leaving less middle seats. You would have to know who are reserving the seats, single people, couples, family groups.

          • (Score: 0) by Anonymous Coward on Monday July 03 2017, @07:49AM

            by Anonymous Coward on Monday July 03 2017, @07:49AM (#534373)

            People flying alone will choose a window seat (best view) or aisle seat (easier to get to the toilet or (emergency) exit).

            A family of four people will take one window seat, one middle seat and two aisle seats.

            Groups of three people will take one of each, but even that won't change the fact that there will be most middle seats left as long as there are even a single booking of any of the above groups.

            Only groups of two will leave an aisle or window seat. Unless they choose the two aisle seats next to each other.

      • (Score: 2) by maxwell demon on Sunday July 02 2017, @08:47PM (6 children)

        by maxwell demon (1608) on Sunday July 02 2017, @08:47PM (#534243) Journal

        But the sentence I quoted says the exact opposite: That they virtually never stick you in the worst kind.

        If it is intended to be a joke, it's a very strange joke indeed.

        --
        The Tao of math: The numbers you can count are not the real numbers.
        • (Score: 0) by Anonymous Coward on Sunday July 02 2017, @10:46PM

          by Anonymous Coward on Sunday July 02 2017, @10:46PM (#534276)

          But the sentence I quoted ...

          That dead horse is dead.

        • (Score: 2) by Magic Oddball on Sunday July 02 2017, @10:55PM (4 children)

          by Magic Oddball (3847) on Sunday July 02 2017, @10:55PM (#534279) Journal

          No... It's an if:then statement. IF the seating is random, THEN several groups of 4 people all getting middle seats would be extremely rare. The fact that their volunteers were all given middle seats therefore indicates that seating is not random.

          • (Score: 0) by Anonymous Coward on Monday July 03 2017, @07:42AM

            by Anonymous Coward on Monday July 03 2017, @07:42AM (#534368)

            That only holds as long as you assume either an infinitely large plane or that nobody is willing to pay extra for an aisle or window seat.

          • (Score: 2) by Wootery on Monday July 03 2017, @09:17AM (2 children)

            by Wootery (2341) on Monday July 03 2017, @09:17AM (#534385)

            You're ignoring that some people pay to reserve a seat, and out of these people, probably none of them reserve a middle seat. So if you're randomly assigned one of the 'leftover' seats, odds are pretty good it'll be a middle seat.

            This has been pointed out repeatedly in this thread.

            • (Score: 5, Informative) by AthanasiusKircher on Monday July 03 2017, @02:56PM (1 child)

              by AthanasiusKircher (5291) on Monday July 03 2017, @02:56PM (#534447) Journal

              From the summary:

              By looking at the amount of window, aisle and middle seating available on each flight, at the time of check-in, Dr Rogers, calculated

              In other words, they took into account the fact that many people had already reserved aisle and window seats, but at the time of check-in given the mixture of seats still left "available," they were still disproportionately assigned middle seats in a fashion that was decidedly non-random.

      • (Score: 1) by khallow on Monday July 03 2017, @02:35AM

        by khallow (3766) Subscriber Badge on Monday July 03 2017, @02:35AM (#534314) Journal

        Point is, if you don't pay the airline for an assigned seat, they stick you in the worst kind...

        I think it's obvious that they're not going to stick you in a seat that someone else paid to get assigned to. So that rules out the "best" seats, whatever those might be. And getting randomly assigned to the dregs of that process, means you sit in the dregs.

  • (Score: 2, Informative) by Anonymous Coward on Sunday July 02 2017, @07:27PM (8 children)

    by Anonymous Coward on Sunday July 02 2017, @07:27PM (#534218)

    I would think that the people who did reserve a seat most likely chose a window or isle seat, leaving only middle seats to "randomly" be assigned to the non-reserve people.

    • (Score: 0) by Anonymous Coward on Sunday July 02 2017, @08:12PM (2 children)

      by Anonymous Coward on Sunday July 02 2017, @08:12PM (#534227)

      By looking at the amount of window, aisle and middle seating available on each flight, at the time of check-in,...

      • (Score: 0) by Anonymous Coward on Monday July 03 2017, @07:46AM (1 child)

        by Anonymous Coward on Monday July 03 2017, @07:46AM (#534370)

        Are you saying that they deliberately stuck people in all the middle seats, and left the aisle and window seats empty?

        If not, then those seats were taken either by people who paid for them (which is what people have been saying) or randomly assigned to other people showing that the experiment really did just demonstrate bad luck.

        • (Score: 2) by AthanasiusKircher on Monday July 03 2017, @03:04PM

          by AthanasiusKircher (5291) on Monday July 03 2017, @03:04PM (#534449) Journal

          Are you saying that they deliberately stuck people in all the middle seats, and left the aisle and window seats empty?

          Yes, at least at the time they checked in. Assuming the summary is accurate, it seems to indicate the algorithm likely prioritizes filling in "less desirable" seats first when people check in. The hope being, I assume, that the airline can still sell SOME people checking in later on "upgrades" to take the remaining aisle and window seats, etc. If the system is consistent in prioritizing "less desirable" seats, I'd assume after it exhausted middle seats, it would start filling window and aisle seats from the back of the plane or something, leaving the generally more desirable ones toward the front of the plane available.

          Nobody wants to pay extra for a middle seat, unless I suppose you have a tight connection and need to be really close to the front of the plane. So if the algorithm truly randomly filled up seats, there might not be any aisle or window options left by the time the last groups of people are checking in, meaning there's less chance for the company to make extra money off of seat upgrades.

    • (Score: 0) by Anonymous Coward on Sunday July 02 2017, @08:13PM (4 children)

      by Anonymous Coward on Sunday July 02 2017, @08:13PM (#534228)

      By looking at the amount of window, aisle and middle seating available on each flight, at the time of check-in, Dr Rogers, calculated the chances [...]

      Emphasis mine. Though the fine summary is as badly written as it is, I take it this sentence means that all free seats, including some unoccupied windows seats, were taken into account for the probability calculations.

      • (Score: 2) by maxwell demon on Sunday July 02 2017, @08:49PM (3 children)

        by maxwell demon (1608) on Sunday July 02 2017, @08:49PM (#534245) Journal

        including some unoccupied windows seats

        Windows seats? Do they offer Linux seats, too?
        SCNR

        --
        The Tao of math: The numbers you can count are not the real numbers.
        • (Score: 2) by Joe Desertrat on Sunday July 02 2017, @09:37PM (2 children)

          by Joe Desertrat (2454) on Sunday July 02 2017, @09:37PM (#534258)

          Windows seats? Do they offer Linux seats, too?

          Linux? That's a hacker's O/S. Mention Linux and you get an aggressive pat down and a spot on the no fly list!

          • (Score: 2) by Magic Oddball on Sunday July 02 2017, @11:21PM (1 child)

            by Magic Oddball (3847) on Sunday July 02 2017, @11:21PM (#534282) Journal

            I imagine at least some of the TSA agents booting people's laptops for examination react pretty much precisely that way to Linux:

            "Hey, that don't look like Winders! You some kinda hacker or somethin?"
            "No, no, it's Linux, just another operating system, like Mac OS X."
            "I ain't heard of no Linux 'fore." *shouts to burly pig-eyed guard* "Hey, Jim-Bob, we got another one of them hacker types tryin to pull a fast one, says he's got the Laynuts." *he turns back to you with a menacing scowl* "Don't you try to talk bullshit to Jim-Bob, he's been usin the Innernet since 2000 an' even knows that H-M-T-L stuff."
            "Uh, I wasn't going—"
            "You better not. He's gonna check you out, you better not have one of them remote det'naters hidin in yer pants since he's gonna find it if you are."
            "What!? Look, no, I can prove Linux is just—"
            "Don't you argue with me!" *he leads you over to the guard* "Hey, Jim-Bob, we gots us a figher here, he freaked when I said you were gonna check for remotes, so you better make sure he ain't got one up his butt or somethin..."

            • (Score: 2) by Magic Oddball on Sunday July 02 2017, @11:29PM

              by Magic Oddball (3847) on Sunday July 02 2017, @11:29PM (#534285) Journal

              (Apologies to Soylentils from the deep South or wherever the hell that accent comes from. I was tempted to use a surfer stereotype instead, but making someone sound 'stupidly menacing' rather than 'stoned out of his gourd' when he's talking like Keanu Reeves in Bill & Ted's Excellent Adventure is a bit too difficult. Trust me, I've tried.)

  • (Score: 2) by opinionated_science on Monday July 03 2017, @12:47AM

    by opinionated_science (4031) on Monday July 03 2017, @12:47AM (#534295)

    Doesn't look like much data. I would expect at least 100 individual trials...

  • (Score: 1) by shrewdsheep on Monday July 03 2017, @09:22AM

    by shrewdsheep (5215) Subscriber Badge on Monday July 03 2017, @09:22AM (#534387)

    Probably, the algorithm used is not random at all. There is probably a list of free seats sorted by "quality" and a new customer gets assigned the best seat he or she is qualified for (which is the worst on the list for the non-reserving customers: quality-eligibility of 0). The randomness is produced by the random seat allocation order as induced by human behavior. So the claim of Ryanair of producing random seat allocation is correct. Collecting many seat assignments will reveal the precise distribution in effect.

    TFA is a staggering display of trying to be sensationalist at the cost of displaying absence of common sense.

  • (Score: 2) by wisnoskij on Monday July 03 2017, @02:19PM (5 children)

    by wisnoskij (5149) <reversethis-{moc ... ksonsiwnohtanoj}> on Monday July 03 2017, @02:19PM (#534429)

    Dr Rogers, calculated the chances of all four people being randomly given middle seats on each of the flights, to be around 1:540,000,000

    OK, lets look at this. OK, so some "Dr." runs an experiment with 4 trials, claims 99.999999999999999999999999999999999999999999999999999999999999999% confidence.

    Lets look at the math ourselves.

    Depending on the seating arrangement we most likely get 1/3rd of the seats being middle seats. It is possible for none of the seats to be middle seats, it is possible half the seats are middle seats. But 1/3rd is very common and it appears to be what Ryanair uses.

    So grade 9 statistics give us.

    Four independent events, assuming that we start with airline with 0 reserved seats (or reserved seats that were uniformly reserved across aisle , window, and middle)
    1/3 * 1/3 * 1/3 * 1/3
    = (1/3)^4
    = 1/81
    Or a 1:81 chance. A 7 million percent difference from what Rogers found.

    Now lets look at how confident we are at this answer.

    Standard Deviation = sqrt(sum(difference from mean)^2/num_trials)
    = sqrt((((2/3) squared) * 3)/4)
    = 0.58
    This is also pretty close to what we get for our confidence interval with a 95% confidence level

    OK, that is pretty horrible, so we found that the chance of randomly being assigned a window seat is 1 +or- .6

    • (Score: 2) by wisnoskij on Monday July 03 2017, @02:25PM

      by wisnoskij (5149) <reversethis-{moc ... ksonsiwnohtanoj}> on Monday July 03 2017, @02:25PM (#534433)

      * Middle Seat

    • (Score: 1) by chair on Monday July 03 2017, @03:56PM (1 child)

      by chair (6194) on Monday July 03 2017, @03:56PM (#534467)
      from the (very short) FA:

      To support her analysis, Dr Rogers was also given access to data from a further 26 individuals, from nine groups, who had been separated from their party when travelling with Ryanair.

      • (Score: 2) by maxwell demon on Monday July 03 2017, @04:18PM

        by maxwell demon (1608) on Monday July 03 2017, @04:18PM (#534472) Journal

        To support her analysis, Dr Rogers was also given access to data from a further 26 individuals, from nine groups, who had been separated from their party when travelling with Ryanair.

        Sounds very much like biased data sampling.

        --
        The Tao of math: The numbers you can count are not the real numbers.
    • (Score: 2) by FatPhil on Monday July 03 2017, @04:19PM (1 child)

      by FatPhil (863) <{pc-soylent} {at} {asdf.fi}> on Monday July 03 2017, @04:19PM (#534474) Homepage
      There aren't 4 events, there are 16 events.

      Which implies, all things being equal, a probability of .28 of getting a middle seat for the Doctor's calculation. Which seems to imply many pairs making adjacent seat reservations, and thus snarfing middle seats making them rarer than the no-assumption 1/3 /a priori/ probability.
      --
      Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
      • (Score: 2) by wisnoskij on Monday July 03 2017, @08:08PM

        by wisnoskij (5149) <reversethis-{moc ... ksonsiwnohtanoj}> on Monday July 03 2017, @08:08PM (#534557)

        Ah, I missed that. That still only gives you a 50 million to 1 probability. Another sentence I missed "By looking at the amount of window, aisle and middle seating available on each flight, at the time of check-in" might imply that many of the middle seats were already taken, presumably by already checked in not paying extra customers.

        Basically, if you want an aisle or window seat, check in late.

(1)