Dutch Auctions are the answer for how tickets should work

[bradtem]

The thing I "won't let go" is not accepting the idea that the tickets "sell out in minutes." I am sorry to be insulting if this is insulting, but that just can't happen, It is simply impossible in an auction designed with reasonable care. If you think it's possible, you may be misunderstanding because this system is designed to make that impossible. If what I wrote made you think it is possible, it is my error. So of course I will not "let go" disagreeing with claims that it's possible, let alone likely.

I used a hypothetical dutch starting point of $1,000. Yes, it is possible at that price to sell out on the first day, but that's not because the system is wrong, it's because my example price was wrong. I am happy to consider debate about what starting price is high enough to assure it won't sell out the first day, but that's the side issue. The key thing is that there is some starting price at which it is impossible. (Example, 1 billion dollars, because there are not 50,000 billionaires on the planet. The real sure-fire price is much lower of course, though you can actually start at a billion if you do an exponential dropping curve in real time, but that's silly.)

The Dutch style has a starting price which could be chosen badly. The sealed bid variant has no cap price, so it is quite literally impossible for it to sell out in minutes. Again, this is not an opinion, it is how it is designed. What can happen within minutes is that, unbeknownst to the auctioneer or any of the bidders, enough bids are in that bids below a certain value are no longer capable of winning, but it is always possible at any time in the auction, to submit a winning bid.

So if your focus was on the hypothetical start of $1,000, it's the wrong focus. It's just an example. in order to write an explanation with real numbers rather than the way I would normally write it as a mathematician where I would have said "$X." I think $1,000 is sufficient, some other's don't, which is fine. I would not actually set a real starting price without doing research and knowing everything BMOrg knows. If I used Dutch style at all.

So pick at the other flaws of a 2nd price multi-unit auction, but, "selling out in minutes" just isn't one of them. If you don't trust me, ask other friends you have with knowledge of auctions. And if this is not something you can do, consider only the sealed bid variant, where there is no starting price and the impossibility of selling out is clearer because you can't even make a mistake in setting it up.

The real issues are not whether such auctions work -- they are very widely used, including on high demand scarce items. The real issue is how, since they are so well tested and effective, to use them within the ethos of Burning Man. Those who have said to me, "this makes this block of tickets too high" have a reasonable case to make and debate. I expect the auction sale tickets to be beyond the means of at least half the burners participating -- this is an explicit part of how it works. The hope is to find a way that this still does not price people out because it funds a large block of below market tickets which do not price people out, but select who goes based on some other criteria than willingness or ability to pay.

[maladroit]

This is reminding me a lot of the type of fervor you see within the free-energy crowd. Too much invested in the idea to ever consider the possibility that their underlying premise might be flawed.

[bradtem]

Um, I know you are trying to be insulting, but we're talking about something a bit different. These are not my ideas. These are multi-century old, tried and tested ideas, and unlike "free energy" or other pseudosciences where a small minority defends a crazy idea against the mainstream, in this case (though I realize you don't know this) I am the one in the (vast) mainstream, even though within this particular thread, you might see it the other way.

Understand I am not saying there is an opinion among the mainstream in game and auction theory about burning man. The question of how to apply these principles to burning man is what I came here to discuss, and whether there is away to do it within our ethos. But it's silly to discuss things that are simply false like whether tickets will "sell out in minutes." I want to discuss the real issues -- what would a high price for the auctioned block of tickets mean for the attendance profile of the event, does it skew it too far in one direction or another? Can the subsidy it offers to other blocks of tickets, can it compensate for that, or even improve on things? I don't promise I know the answers to those questions.

Now you are demonstrating one element of how these auctions don't always work in the perfect way they are designed, which is that there are failures when people don't understand them. And that is a challenge, but I didn't do this thread to show how i would explain them in a real auction. For that, I would use software tools that help people pick their bid well, not a thread. So that's a real issue.

[Elderberry]

Would everybody just quit posting in this thread and let it die a quiet, natural death? Some people believe in Dutch Auctions as much as others in Mohammad (or Jesus or what erver). You can never change the mind of a fanatic.

(Yeah, like that would ever happen.)

[Captain Goddammit]

Robin Hood... you know, take from the rich, give to the poor. It worked in the fairy tale...

[maladroit]

My last post:

[vargaso]

I used a hypothetical dutch starting point of $1,000. Yes, it is possible at that price to sell out on the first day, but that's not because the system is wrong, it's because my example price was wrong. I am happy to consider debate about what starting price is high enough to assure it won't sell out the first day, but that's the side issue. The key thing is that there is some starting price at which it is impossible. (Example, 1 billion dollars, because there are not 50,000 billionaires on the planet. The real sure-fire price is much lower of course, though you can actually start at a billion if you do an exponential dropping curve in real time, but that's silly.)

Now we're in the realm of the theoretical. Sure, you could set the starting price ridiculously high just to avoid a quick sellout, but the end price would still be much higher than the current ticket price of $390. THAT'S THE POINT. Your proposal would result in favoring those with more money, and even with some kind of subsidized ticket program, which is already in existence anyway, the BRC demographic would suddenly be highly stratified, socio-economically, thanks to over-engineering. That's bad. Who care if it takes 10 minutes or 10 months to sell out if tickets are suddenly financially out of reach.

[forty_eight]

what is so important about not selling out in minutes?

auctions set efficient prices, but that's about it.

it's an answer to a question no one asked.

[Eric]

what is so important about not selling out in minutes?
[snip]
it's an answer to a question no one asked.

[Simon of the Playa]

Wear your manly man pants.

[mdmf007]

Um, "the tickets will be gone in an hour" is not how a Dutch auction works. Or a more basic 2nd price sealed bid multi-item auction. There is no time constraint really, unless you are bidding near the final price and it is announced that they might decide who among those in a last place tie gets the tickets if there are too many. (I have generally used random for tiebreaking because if you bid the exact price, you are supposed to be fairly cool with either winning or not winning. If you are not cool with not winning at that price, you should have bid more.)

I am in fact starting to believe that a single sealed bid is better than the Dutch. The Dutch does disclose more information about what is going on, and can induce fear as the supply starts running out. If 50,000 tickets run out on the first day for $1,000 I will eat my playa hat, though it would be interesting to see what happens to have an event with $50M and tons of people who got in for near free based on luck or merit or some other mix.

Yeah, we all know how the auction works, but start your Dutch auction at just about any price and the event will still sell out waaaaaaayyyy before you get close to a 390.00 figure. Start at 2,000.00 a ticket and I guarantee it will still sell out in an afternoon at the 2,000.00 level. If it did not I would eat my hat and yours.
###

Make the event harsher. Actually pretty easy. Go back a few years. No onsite water deliver. No RV delivery. No commercial airlines. No grey water removal. Definitely no gas delivery for art cars, genny delivery, the list goes on & on. All of these things make camps less radically self reliant, and make it easier for those that buy their burn.

Exodus was easier because a lot more people stayed until Tuesday.

Might want to read the permit - most of these items are required for safety and sanitation. Otherwise people are dumping RV shitters on the playa, and clogging Gerlach with deliveries that have to be picked up there.

MDMF

[Lonesomebri]

Once you people tire of having "the free market", whatever that is, and the Dutch take care of ticket distribution, the BRCCP will still be here.

That's my problem with all this, the whole faith that "the market" is something that can figure out how best to handle human affairs, not humans.

[pink]

Um, "the tickets will be gone in an hour" is not how a Dutch auction works. Or a more basic 2nd price sealed bid multi-item auction. There is no time constraint really, unless you are bidding near the final price and it is announced that they might decide who among those in a last place tie gets the tickets if there are too many. (I have generally used random for tiebreaking because if you bid the exact price, you are supposed to be fairly cool with either winning or not winning. If you are not cool with not winning at that price, you should have bid more.)

I am in fact starting to believe that a single sealed bid is better than the Dutch. The Dutch does disclose more information about what is going on, and can induce fear as the supply starts running out. If 50,000 tickets run out on the first day for $1,000 I will eat my playa hat, though it would be interesting to see what happens to have an event with $50M and tons of people who got in for near free based on luck or merit or some other mix.

Yeah, we all know how the auction works, but start your Dutch auction at just about any price and the event will still sell out waaaaaaayyyy before you get close to a 390.00 figure. Start at 2,000.00 a ticket and I guarantee it will still sell out in an afternoon at the 2,000.00 level. If it did not I would eat my hat and yours.
###

Make the event harsher. Actually pretty easy. Go back a few years. No onsite water deliver. No RV delivery. No commercial airlines. No grey water removal. Definitely no gas delivery for art cars, genny delivery, the list goes on & on. All of these things make camps less radically self reliant, and make it easier for those that buy their burn.

Exodus was easier because a lot more people stayed until Tuesday.

Might want to read the permit - most of these items are required for safety and sanitation. Otherwise people are dumping RV shitters on the playa, and clogging Gerlach with deliveries that have to be picked up there.

MDMF

When I started going, yes, you could get your RV pumped out. But you couldn't get water delivered. Or arrange for camp very water removal.

[mdmf007]

When I started going, you could shoot rifles at shit - everything evolves...

[Tiahaar]

This thread is enjoyable on an academic level, admit it

Turns out there are a WHOLE BUNCH more auction variations around than I ever guessed...wiki links

Brad's Dutch Auction (which I find ebay never really used...though they called it that...sorry...more below)

Vickrey Auction a type of sealed bid auction, check out the theory @@!

Multi-Unit/Uniform Price Auction,nearly the ebay 'dutch' auction except for the sealed part, to quote:

"A uniform price auction otherwise known as a "clearing price auction" is a multi-unit auction in which a fixed number of identical units of a homogenous commodity are sold for the same price. Each bidder in the auction may submit (possibly multiple) bids, designating both the quantity of units desired and the price he is willing to pay per unit. Typically these bids are sealed - not revealed to the other buyers until the auction closes. The auctioneer then serves the highest bidder first, giving them the number of units requested, then the second highest bidder and so forth until the supply of the commodity is exhausted. All bidders then pay a per unit price equal to the lowest winning bid (the lowest bid out of the buyers who actually received one or more units of the commodity) - regardless of their actual bid. Some variations of this auction have the winners paying the highest losing bid rather than the lowest winning bid."

And there's enough analysis on game theory and market weights to make a statistician giddy, and links to many variants.

In the end, we can all just ignore this and remember to log on at 12 noon and hope for the best

[southern crone]

I think after the 2015 tickets sold out the people left in line should get the first tickets for 2016, 2016 for 2017, etc. Maybe then I might get one!!! Of course I didn't even get to try this year, had to work, too many calling in sick.

[Captain Goddammit]

When I started going, tickets were cheap and available at the gate. If anything, it was easier back then. The reason there were "only" 20,000 people is that almost no one knew about it.
If you'd all kept your mouths shut... I mean, hell I didn't even bring a camera the first five or six times, and I learned real quick not to talk about BM to "default world" people.
Now look what youve done.
In that respect, Brad this situation is largely your fault!

[Elderberry]

^^^hahaha, that's true^^^

[bradtem]

Yeah, we all know how the auction works, but start your Dutch auction at just about any price and the event will still sell out waaaaaaayyyy before you get close to a 390.00 figure. Start at 2,000.00 a ticket and I guarantee it will still sell out in an afternoon at the 2,000.00 level. If it did not I would eat my hat and yours.

MDMF

Yes, and I think this is the 2nd or 3rd time I've had to make this correction, of course the tickets sell for more than $390, probably quite a bit more. This is what I said in the start, it is the purpose of this approach, to capture the extra money burners are willing to put into going and channel it into the community and subsidies for those who don't have the money. (I am not referring to low income tickets where you have to document your low income, but rather the more classic method of making it hard to get cheap tickets, so that they go to people with more time (or devotion) than money, while people with more money than time just pay the money.)

(Note: For those uninterested in game theory of auctions, probably best to skip this and the rest of the thread as well.)

I threw out some example numbers, but in fact with study of the real demand, you can tune it as you like. The pre-sale is, in effect, an example of this philosophy. It gives people the ability to trade money for greater certainty of a ticket, and that extra money reduces the price of the other tickets (at least in a non-profit, it should.) You could express this by doing 10,000 tickets at auction and 60,000 tickets by lottery. You could very closely match the current structure with 4,000 tickets in a uniform price auction, and the rest by other means. Indeed, the only difference between this and the current structure is that the market would set the price of the block of 4,000 vs. the BMOrg setting the price. If you think 50,000 tickets could sell for $2,000 in a day, you will presumably predict an auction of 4,000 would reach some astronomical price -- I am less sure.

There are reasons to make the auction pool larger than 4,000. A larger pool can still raise more revenue, but at a price that seems fairer. The bigger the pool, the lower the price (though total revenue which is pool size times price can actually be higher.) The lower the price, the fairer it feels. In particular, with this approach you have one group trading money for certainty and time, and another group accepting the uncertainty of the lottery (ie. trying to hit refresh at exactly noon) in order to save money. It's important to preserve fairness, so you want people to feel that this is a reasonable trade-off. If the auction tickets are $2,000 and the lottery based tickets are $390, that's too much of a gap for many people to consider that a fair trade. That will provide some downward pressure on the auction, but mostly it creates dissatisfaction. As said before, the uniform price auction variants are designed to produce the maximum satisfaction for both buyers and sellers. It is not necessary that the buyers understand how this works, what is important is to maximize their satisfaction with the result.

If the people buying the auction pool feel they are donating to support the community, they can get more satisfaction from their higher price. In fact to make that stronger, I believe it might even be legal (though this is not certain) that any amount paid above a certain base price, such as the individual sale price, could be treated as a tax deductible donation, which reduces the cost -- though it's slightly regressive because higher income people get more from the deduction.

As long as the market price is well above the base ticket price, lots of people will spend lots of money to go to BRC, but all that money will go to speculators and not to the community. I do not think that just having low prices for all is an available solution.

[maladroit]

Yes, and I think this is the 2nd or 3rd time I've had to make this correction, of course the tickets sell for more than $390, probably quite a bit more. This is what I said in the start, it is the purpose of this approach, to capture the extra money burners are willing to put into going and channel it into the community and subsidies for those who don't have the money.

I had to break my own rule and post one more time in this thread.

Again: FALSE ASSUMPTIONS.

We are not willing to put in this "extra money" you keep mentioning. For many of us, there is no extra money we'd be happy to pay. Doesn't exist. Now, if you create a situation to extort more money from us, then a lot of people will very begrudgingly pay it in order to avoid the alternative outcome.

Your proposal (to the best of my ability to determine) is supposed to solve the problem of scarcity while making everyone happy with the outcome. I can tell you right now that's fucking impossible, ESPECIALLY with your idea. Even the people who are forced to pay $800 for each ticket will feel cheated. The people who were priced out of the market will feel cheated.

And your justification "we can channel it all back to the community" is bullshit. There are thousands of ways to donate to the ORG and to specific projects and many of us already do.

[Eric]

(I am not referring to low income tickets where you have to document your low income, but rather the more classic method of making it hard to get cheap tickets, so that they go to people with more time (or devotion) than money, while people with more money than time just pay the money.)

Your tenacity to utterly ignore the fact that there is a market for ~160,000 tickets (80,000 people registered in for the Main Sale, each could buy up to 2 tickets), and that people who lost out in the first "expensive" batch would flood the second "cheap" batch as well in their desire to get tickets. You're sitting in your ivory tower, playing with ideas that only work if you ignore the basic reality that four times as many people want to go as there are tickets.

No matter how many times you think we don't understand you:
you have yet to explain how your system is any fairer in making sure the maximum amount of people (at all income levels) at least have a chance to get a ticket at a "reasonable" price, like they do now.


This has become mental masturbation with no money-shot and a hand that doesn't understand your needs.

[maladroit]

One more (it's just so darn salty, I can't stop):

Um, I know you are trying to be insulting, but we're talking about something a bit different. These are not my ideas. These are multi-century old, tried and tested ideas, and unlike "free energy" or other pseudosciences where a small minority defends a crazy idea against the mainstream, in this case (though I realize you don't know this) I am the one in the (vast) mainstream, even though within this particular thread, you might see it the other way.

Hey, WE ARE the community you're trying to sell this idea. Fuck off with the attitude that we should do things the way the rest of the world does them.

[bradtem]

One can conclude maximum demand for tickets at roughly $420 is no more than 150,000+20,000+1,000. (Individual sale did not quite get 2 tickets per winner, and a share of vehicle pass price should be included.)

However, the number of 171,000 is probably a fair bit less than the maximum. That is because all 80,000 registrants did not participate in the sale (BMOrg has not released that number) but most importantly because if history is a guide, a large number of people who participated in the sale were participating twice, or more than twice. Not just the scalpers, of course, but the pairs of people who both registered and both tried to participate. In addition those who asked friends and associates to attempt to purchase even though they are not going, and those who asked campmates who only needed one ticket to ask for 2 to help the camp get tickets.

There is strong evidence from the lottery in 2012 that the duplication is massive. That lottery had, as I recall, as much as a 4x oversubscription, similar to the ratio of potential ticket registrants to tickets in 2015. But as we all know, by the time of the event, tickets were going begging, being sold for half price or less. Part of this can be attributed to the fact that the lottery stopped certain groups of people from coming that year, and those people were not ready to come at the drop of hat when the tickets became easy to find. (The Org will be able to calculate how many theme camps and regulars avoided the event that year.) However, the evidence is strong that actual demand in 2012 was probably only modestly above capacity. In 2012 there were still tiers of pricing, so the fixed $390+pass system today would have reduced demand a bit.

But it's clearly very, very unlikely that there are/were 171,000 people willing to buy a ticket to Burning Man 2015. Thus my "utter" tenacity to not treat that as a fact. A more likely guess, but I am just guessing, would be something a bit under 100,000. Still more than city capacity of course, and thus a shortage.

Now, demand for 100,000 tickets at $525 is on a demand curve that we don't yet understand. Demand at $800 is going to be quite a bit less than 100,000. Some burners will be price insensitive up to a point, but it's hard to say how many there are. Those who rent RVs for 3 or fewer occupants or contribute heavily to theme camp construction might be a clue (has anybody got a count on the number of RVs?) Guessing again, I would venture probably at most 10,000 of those. However, those are just my estimates, and the point is not to debate against other guesses, but to understand that the total demand number is nowhere near 171,000.

Anyway, now to the question of why this system might be superior of getting people to have tickets at prices they can afford, we go back to a few key points:

• While the lottery/click-reload-at-noon-and-pray system is fair in that it allocates mostly regardless of income, it has many flaws which have been discussed.
• One flaw that I forgot to mention is the flaw so obvious in 2012 -- people "cheating" and trying for more tickets than they need to make sure they get what they need. Because a large auction blocks scalping, there is much less need to have a limit on how many tickets a registrant can buy. A larger limit, like 8-10 does not help scalpers and vastly reduces over-buying.
• Buy moving that extra money into subsidy for below-market tickets, we do exactly the goal you state, getting more people the chance to get tickets at lower prices. For the lower price, they sacrifice the certainty of getting a ticket, and they probably should sacrifice the ability to resell the ticket

But let's avoid the ad hominem, shall we? ePlaya hasn't got a very good reputation in that department, but I hope people can rise above that.

By the way, the suggestion that those who lose out one year get priority next year has been bandied about. It has merit, but is challenging to enforce (without non-transferable tickets, which BMOrg does not wish to do at this time for general sales.) Burning Man actually scares a remarkable number of people away every year, it amazes me what fraction of the population are virgins each time. Only a minority of us get converted to the type who want to come every year -- which is good, or the population would have exploded a great deal more. The vast majority of people who have come to Burning Man never came back.

[Captain Goddammit]

A really good writer knows the art of getting a point across efficiently, without boring the audience to sleep.
Burning Man doesn't use the Dutch auction and it isn't going to.
You lose, good day sir.

[Eric]

There is strong evidence from the lottery in 2012 that the duplication is massive.

You mention the 2012 sale in this post a couple of times, even going so far as to use it as your "evidence" that "a large number of people who participated in the sale were participating twice", even though you don't work in Ticketing and have absolutely no proof of that, other than words tossed about on the internet. Yes, we've all heard of people who claim to have done it, but there is no actually proof they did it, let alone that "a large number" did. Also, this completely ignores that the way the sale has been run for the past three years is totally different than 2012, thus making any points using that baseline moot.

You're going further down into the same rabbit-hole as conspiracy theorists, who turn unrelated information into "facts" to prove their theories.

Also, joining maladroit, I am also part of this community you're trying to get to pay more, just to fix something you see wrong with the system (and since you still haven't CLEARLY stated what that "something" is, I can't even decide if it actually needs fixing). Personally, this is looking more and more like a little plan to get 40,000 people to pay a huge chunk more than the current price, all to give yourself the Squeeee's, and that's not going to happen.

[Captain Goddammit]

Fortunately, this isn't a board meeting, and whoever wins this debate doesn't get to make a new policy.
Yay for that!

[maladroit]

Fortunately, this isn't a board meeting, and whoever wins this debate doesn't get to make a new policy.
Yay for that!

While true, I think many of us felt compelled to point out exactly what was wrong with this idea, just in case someone who DOES make policy ever saw it. We don't need ideas like this floating around for BMORG to pick up because "hey this guy seems pretty confident in his qualifications."

[bradtem]

I don't think anybody knows just how many people were trying for the same pair more than once. Clearly it happened a great deal, massively, in 2012. As such, I fear the burden of proof is on those who want to assert that it has gone away (or reduced to some fraction of its value then.) If you look in the comments on the BM Blog thread, you will see at least some evidence of people doing it.

If anything, one might wonder if it didn't increase. After all, back then people didn't even know how it would go down, and they figured out to triple and quadruple register in the lottery. Now they have years of experience, and we even knew before the sale (though after the registration) that there were 80,000 registrants for 40,000 tickets.

The point is that the idea that the 80,000 registrants represents 160,000 (or even 150,000 at the announced ratio) people wanting to buy tickets at $390 is, to put it bluntly, ludicrous. Nobody here has data from this year. Even BMOrg has only limited data on this year -- they might have some data on how many registrants came from the same address, but I suspect most people doing it multiple times would use different addresses.

What we do know for a fact is that the over-participation in 2012 was massive. Truly massive. Three times the number of entrants than tickets in 2012, rising to 3.75x in 2015. (It should be noted in 2012 the next sale was turned into a directed sale to get theme camps tickets, and in 2015 the directed sale took place first.) So we see a climb, but not an overwhelming one. And the bizarre result that the 2012 event ended up not truly being sold out, with significant tickets left unused or sold below face value -- attendance was actually down from 2011. The decline can be explained by those who failed in the lottery and just gave up, then finding themselves unable to purchase in the ticket glut.

But still, the assertion that suddenly it's all switched and there were 150,000 people looking for tickets would need a lot of evidence to back it up. It's a big change fro the facts we have.

But far be it from anybody on ePlaya to admit they could have been wrong about something. Well, that's not fair, there are good people here, but it's one of the least civil message boards I participate on, if not the least, which is a shame because when people get to BRC they turn into much nicer folks, much more accepting of new ideas.

So enough of this.

[Lonesomebri]

Whenever the plans for human activities devised by technocrats don't work, they always blame the humans, never their supposed perfect plans.
Yes, if only the markets and people with the most money would run everything, that would solve our problems.

[Eric]

But far be it from anybody on ePlaya to admit they could have been wrong about something.

So much truth in this statement, coupled with so much irony.