Olympic Tickets: The Logistical Game

When the Summer Olympic Games come to your city, the chances are it will be a once in a lifetime chance to see the event. The inevitable excess demand creates a logistical quandry — and we’d like to know if you could solve it in a better way than the organizers of next year’s event.

For those who don’t know, the London 2012 tickets are not being distributed to domestic buyers in the way of most major sporting and entertainment events by which they go on sale at a set time on a first-come first-served basis. Officials believed it was likely using such a system would inevitably lead to what could only be dubbed servercide and, given than 1.8 million people made a total of 20 million ticket requests, they were probably correct.

The system they did use has been controversial, however. It involved ticket applications being taken online during a window of several weeks, with applicants required to hand over credit card details. Money is now being debited from accounts in line with the tickets people have been allocated, apparently on a purely random basis. (Spare tickets for events that didn’t have enough applications will be offered again in a second round of sales.)

That’s led to a range of problems. Some are purely administrative (the money is being debited before the notification of which tickets people have been allocated), but others are more fundamental. Applicants were left with no idea what proportion of tickets they applied for would be allocated to them, meaning a choice between picking a few key events and being left with nothing, or applying widely and praying they didn’t get everything they applied for: successful applicants must pay immediately and must wait next year before they have the option to sell any unwanted tickets (at face value only.)

The results have also demonstrated the way random distribution works. With 20 million applications for six million available tickets, the average applicant should get one-third of the total they applied for, which does appear to be the case. But it’s been reported that 250,000 people will get no tickets, while some are already reporting they’ve got everything they applied for (and are facing a lean period in their finances.)

So while the outcome appears to have been exactly as would be expected from such a system, there are plenty of complaints about the perceived fairness or efficiency of the set-up. Here’s where you come in: we’d like to hear your ideas for designing a better system, a choice that of course involves a subjective criteria for what counts as fair and/or efficient.

To get you going, here are a few rival systems we’ve thought up for debate:

  • The same application system, but with tickets allocated more proportionally: everyone gets roughly the same “tickets received:tickets applied for” ratio (rounding up for those who’d get less than one ticket!) and the luck is in which events that proportion covers.
  • An adjusted version of the same application system, but with each applicant being allowed to pick as many events as they like, while setting a personal maximum total spend limit.
  • An adjusted version of the same system, with the application process carried out in several stages (each covering a selection of events) to avoid people having to overcommit.
  • A blind auction: for each event, applicants name the price they’ll pay for a ticket. If there are 10,000 seats at an event, those who put in the 10,000 highest bids get seats, with the highest bid getting the best seat available and so on.
  • An adjusted blind auction: the relevant number of highest bidders get tickets, but they all pay either the average bid of the winners, or the lowest bid of the winners.
  • A true lottery: applicants buy lottery tickets for a small sum, with games ticket winners then picked at random. Tickets could be for specific events or simple pot luck for the entire games. Lottery ticket buyers could buy multiple tickets for multiple seats: for example, if you buy three tickets and one of them wins, you get three seats next to each other at the relevant event. Applicants could be limited to one lottery ticket each, or there could be unlimited sales.

5 Responses to Olympic Tickets: The Logistical Game

  1. Hi

    Enjoyed the article.  A mate and I split the tickets we wanted (19
    events in total, 2 tickets per event) and we have one or two tickets
    (from the ones I went for, it seems none for him).

    I have been pondering a fairer system while waiting for my own results
    of the lottery. I don't like the “tickets received:tickets applied
    for” ratio as surely many people would apply for one or two 'headline'
    events and mop them up, while others who went for a mix of headline
    and other events would effectively have no chance at getting the

    The idea of a 'maximum spend' might seem sensible (I was proposing it
    at one stage) but then everybody would just apply for as many tickets
    as they could, safe in the knowledge that they could not win them all
    and face ruin, so everything would be even more oversubscription.

    Not sure about the 'several stage' model.  What if there were six
    phases, and I had a limited budget.  I really want to get to the
    swimming which is in stage six.  Do I hang off and apply for nothing
    in stages one to five, then apply for swimming and miss it?  Or do I
    apply for a banker or two in the early stages and perhaps blow my
    budget before being able to apply for what I really want?

    I dislike the blind auction because all of the headline events would
    go to the wealthy (and a lot of others would as well), leaving all of
    us who can't afford to spend a lot on the events the rich don't want
    (Given that we applied for 19 events I'm not suggesting I was in as
    bad a boat as many applicants, by any stretch).  It would mean that Mr
    Jones who lives near the Olympic Stadium but has a low paid blue
    collar job effectively have zero chance of getting a ticket to the
    mens 100m finals.  At least with the existing system he does have a
    very small chance.  The 'adjusted blind auction' has little more merit
    for the same reasons.

    I'm not sure about the lottery system.  It strikes me that there would
    have to be price banding to raise sufficient cash for the organisers.
    Therefore, what if a lot of folk 'won' the lottery for, say, diving,
    but they all wanted the cheaper banded seats and there weren't enough
    to go around.  Do they have to 'lose' their 'win'?  Alternatively if
    seats were equally priced then the average seat price applied may be
    too high for many to afford.  Additionally people would complain that
    having paid the same price, some folk got wonderful seats while they
    got less good ones.

    I don't think there is a perfect solution.  I do wonder if one
    sensible option would be that the current system be applied,
    oversubscribed events are ballotted, but then those who win absolutely
    no tickets are offered one (or more, I would imagine most people
    applied for the same number of tickets to each event, so perhaps
    offered the number they requested generally) to one of the sports they
    wanted (ideally) but in another price band or on another date that is
    still available.  As it would be only one or one set of tickets to one
    event, even if it were a higher price band many may take it so that
    they 'could be therre', and knowing that they aren't getting the five,
    ten or twenty tickets they applied for.  If no tickets were available
    for the sports they applied for then they could be offered a random
    event which would give them something to see.  Obviously this would
    have to be geographically sensible for the football…

    Not perfect, but my best solution personally!

  2. Any kind of top-down approach will have problems like the ones listed above. Setting up a market is the best way to ensure that those who want the tickets most will get them. Instead of a blind auction, set up a public auction. This will allow purchasers to update their bids based on new information, and ensure that they get the option to pay more if they are willing to, or bow out if it gets too high. Heck, I'm sure eBay would love to get their stamp on something like this.

  3. A system would be to rate all events, or a good portion of them, with 1 the grade for the most wanted. You indicate the total of events you would like to go to. That way, you would likely have a ticket for one of your ten favorites.
    Such a system is used to assign schools to teachers in France (But each teach has a weigh based on experience and ratings following inspections)

  4. This may be an odd thought, or it may have been covered and I missed it.  Time-released, blocked, banded lottery. I think this will be fair, but I can't be certain as it's a pre-coffee morning and I really don't want to deal with the mathematics involved in testing right now.

    The basics: The ticket seller creates a set number of tickets to be sold to a particular population (think student tickets for university football, like the University of Tennessee sets up). This is then chunked into blocks. Each block contains a certain number of seats in each band, or each block is determined by the band (i.e. each block has cheap, middle, and expensive range seats, or there's a [group of] cheap block[s], middle, and expensive). Anyone who wants a seat in that block or a particular band in that block requests or buys a lottery ticket (if it's a buy system, no more than $1, £1, or €1 or perhaps 2; I don't recommend charging, but that's up to whomever). One week after ticket distribution, numbers are chosen and the block is sold. Ticket winners have 1 week to buy their tickets. If the winners don't all buy their tickets at the end of that week, tickets are released for general purchasing when the lottery's over, a further set of numbers are drawn, or they're added to another block. When the purchasing week is over, a new lottery begins for the next block, "rinse, and repeat". Optionally, and as a nice sort of added value touch, tickets purchased via a "pay to play" lottery could have a reduced price of the lottery ticket price or a proportional reduction based on the number of lottery tickets sold. Not optional: lottery tickets would be limited per person to something like 2 or 3 or another realistic, fairly low number.

    Why I don't recommend a purchase price for the lottery tickets: A lot of places already have that, and call it LOTO or something equally ridiculous. People drop anywhere from 50¢ – $10 for a "Pick 3" to a MegaMillions ticket (Tennessee prices) and can win a lot of money, and they don't have to pay for it. Granted, I know this is different, but people might not view it this way. In addition, while the 1 or 2 unit-currency price is intended to pay for the lottery system, some administrators, politicians, or other might see it as a way to really turn a profit and not offer the optional reduction mentioned above.

    This is not a simple or easy system, I think, and may still have its logistical issues, especially with the idea of counterfeit lottery tickets, but one has to deal with counterfeit tickets of all kinds anyway.

    Some benefits/pros: People's accounts are not automatically charged, so no issues for them there. Rather than apply, people need simply get a ticket from an authorized location, physical or web-based. Limits on the number of lottery tickets allow for some logistical control, counterfeit control, and other controls that aren't coming to mind right now. It also allows for someone who wants to increase their odds of getting 2 or 3 tickets, let's say in a 10 limit system, to return or reject any unwanted tickets he may have won to be added to a different pool or block. For a totally digital version, a confirmation and e-mail system may be set up for access to a secured ticket purchasing portal or somesuch.

    Some cons: This system takes time; time people don't really want to spend these days. The system may also be prone to servercide when lottery numbers are released. As with any system, some form of cheating may occur (counterfeiting, cracking, etc.). Scalpers for lottery tickets and then for purchased tickets. Corporate greed (see the paragraph on why I don't recommend purchase price for lottery tickets).

    I'm sure this system has other flaws and benefits. It isn't perfect, and I don't believe any system is, but it's something a little bit different, I think.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.