I honestly don't think the problem is how you portion out the tickets. The system you describe does not seem to be any better at getting the tickets in the right hands. I'm not so concerned with the unfairness of the current system, but with the fact that many of these tickets will be completely wasted.

The real problem is that they've combined a 1 hour media event with a developer conference. So many people every year use their ticket JUST for the keynote. It's like combining the Super Bowl and JSConf, then reaching the conclusion that obviously JSConf needs a bigger venue because millions of people tried to get a ticket.

I'd be really curious to see how many developers would seriously be willing to throw down $1600 for a keynote-less conference that notoriously offers no free stuff (ala Google IO) and has many of the materials available afterward. Then instead of trying to solve the intractable problem of a lottery that leaves more people satisfied, you'd actually see who really cares about the content.

> It's like combining the Super Bowl and JSConf, then reaching the conclusion that obviously JSConf needs a bigger venue because millions of people tried to get a ticket

JSConf also sells out in like 10 minutes, which is why they moved to a lottery system. There are no media events or giveaways there (mostly), just a really really good event that refuses to get larger because getting larger may (probably will) spoil some aspects of the event that make it really good.

Think about the lottery this way: if WWDC sold out in two minutes, it's likely that more than the number of people that can attend were on there trying to buy tickets at the same time, so it was really down to a lottery of timeouts, latencies, and overwhelmed systems. A real lottery at least makes it slightly more civilized, doesn't give geographic or equipment-based advantages, and spreads out the entry times over a day or two to let all timezones take equal advantage.

So, there's not really a gain in the quality of the person who gets to go, but it does call a spade a spade and takes some of the frustration out of the physical process of getting a ticket (if not with the result of the attempt).

edit: I do like non-transferable tickets, however (with some negotiated process of refundable tickets). This can be mostly enforced by essentially requiring will-call for all ticket purchases. While some people might be willing to let you borrow their credit card/id, not that many strangers will. You will have to be prepared to be a jerk and turn people down, though.

> doesn't give geographic or equipment-based advantages, and spreads out the entry times over a day or two to let all timezones take equal advantage

(I agree totally, am just adding some things I think are important.) It also levels the playing field in other ways: imagine someone who has a slight tremor in their hand, and can't use a mouse as quickly as others. Many people have jobs (including relevant ones) or families that might require their attention during the critical minute-long window.

So, unless your goal is to have a conference filled to the brim with the world's most dextrous, carefree, and well-off people, "first come first serve" doesn't make any sense... it is "fair" for some definition of "fair", but it is using a criteria with existing fundamental biases. As another example: "auction off to the highest bidders" is also a fair allocation scheme, yet many would balk at that.

At least this year they pre-announced the tickets; last year the time zone advantages were more drastic, and the "I have other things in my life that require me to be on certain schedules" issue was more costly. You also run into human advantages, such as "do you have friends (or minions) who are also on crazy schedules and can alert you". People like myself even had programs (and/or services like Pingdom) watching for changes to the website (although at least that is in some way related to the concept for the conference ;P).

Was this for some previous JSConf? I ask because from what I can recall of JSConf "back in my day", the price was pretty cheap and the seats were incredibly limited -- so I can definitely see ~200 seats @ $200 a pop selling out in 10 minutes and a lottery being necessary. I'm not so sure the same holds for WWDC however. I mention this because if I go to the JSConf US 2013 site right now, there seems to be 32 days till the conference and I am still able to buy a $2000(!) ticket (no sellout apparent at this price point, which leads me to continue believing my analysis of WWDC). This might be some sort of "special" jsconf though so I don't know.

This year was 100 early bird tickets at $650 with the usual ticket rush, then 200 tickets at $750 selected by lottery. I don't know what those $2000 tickets are...maybe extras held in reserve for sponsors or something that they ended up not needing.

I'm still not sure about your idea...it just seems to boil down to reducing the incentives while keeping the price high, so of course there is some point where you can adjust that mixture to where you will only sell the number of tickets available.

I agree with the argument that the keynote attracts some people that otherwise don't care (though the WWDC is always super crowded post-keynote), but considering the other conferences that have this problem, the volume of people buying tickets, and the volume of people upset with the process that denied them tickets, I'm not convinced that going keynote-less would be sufficient to erase the issues with ticket buying. And that's really the only knob you have except for raising prices, which certainly does your attendee demographics no favors.

I also really have to agree with saurik that, even if you can reduce demand to have tickets sell out in, say, an hour, simple distractions that morning (kids, forgotten meetings, car problems) still mean you miss it, which will again heavily skew your demographics to the set of people glued to twitter who would never have missed the 30-minute, 15-minute, and 5-minute warnings of impending ticket sales. Some of those people are going to indeed be the true fans, but many will just be seeing those by virtue of attributes that bring no advantage to the conference, and may limit its reach. It takes all types :)

The building is packed all week. I don't think that keynote-only attendees are a significant factor here. If the two were separated, then maybe it would have sold out in 80 seconds this year instead of 70 seconds. Big deal.

The imbalance between the number of people willing to pay $1600 for the conference and the number of tickets available is so enormous by this point that there's simply no way to make it fair. It doesn't matter if it's first-come-first-served, or a lottery, or if you give preference to people who went before, or people who haven't gone before, or anything. No matter what you do, you're going to leave out a ton of people who deserve to go.

The event needs a bigger venue or some other radical change if it's to avoid leaving a lot of people out in the cold. Maybe Apple doesn't care, and nothing says they have to make any changes at all. But if they do care, then something really has to give. They need a vastly bigger space, or multiple conferences throughout the year, or a higher price, or something to manage the insane demand, if they actually want to improve this stuff.

Who really cares about the content? A ton of people. Buying a ticket only for the keynote is simply not a problem, and it's obvious just from being at the conference and talking to people who go or who want to go but can't because they couldn't buy a ticket.

So one interesting thing is that the "two disparate events, one conference" problem goes a little deeper. There seem to be four "things" you get out of WWDC:

1. Meeting with people (can be achieved without a ticket and just going to SF that week)

2. The sessions (many of which are recorded and thus largely don't require a ticket, but some are not I think?)

3. The keynote (not actually important for developers)

4. 1-on-1 with engineers. As far as I can tell the only truly unique and universally considered valuable aspect of the conference that absolutely requires a ticket.

So it seems that (4) is the real value here, probably more to do with the fact that its so hard to get developer help from Apple any other time of the year. Since you mention radical changes, I wonder if it would be worth it for Apple to just prerecord ALL sessions, make them all available after a Keynote that has nothing to do with the conference, and then offer a WWDC week that just focused on direct contact with engineers. This would actually be better for the sessions too: many times I've had a question about a session, and either there wasn't time to answer it then, or I didn't think of it until much later. By having a conference full of people that had plenty of time to review content beforehand, you'd probably get more questions answered.

I think that #1 is still enormously enhanced by the conference itself. You don't get the same experience if you just hang out in the area that week. (Or I assume you don't, I've not tried it.) The serendipitous socialization aspects are a big deal, even only looking at doing it with people who don't work at Apple.

I agree that #2 isn't all that important (although there are probably some people who learn better in that sort of environment) and #3 is basically just a free show that you get with your ticket and could easily be made completely separate.

I like the idea of a week that's just focused on direct contact with engineers, but I fear that it would end up in very inefficient uses of their time due to being flooded with people asking basics because they couldn't be bothered to watch the intro sessions. In general, I fear what would happen without enough structure. But perhaps it could be made to work, and it certainly seems worth considering.

The obvious first step seems like a bigger venue, though. Apple keeps saying they can't... but other companies pull it off, so I have to wonder why Apple can't.

> They need a vastly bigger space, or multiple conferences throughout the year, or a higher price, or something to manage the insane demand, if they actually want to improve this stuff.

Regarding multiple conferences, Apple decided not to do the Tech Talks for 2012. I hope they'll reconsider for 2013. That really was a great resource to have, and the 1:1 with the evangelists and DTS was excellent.

For higher prices, I keep on hearing this rumor and it never happens. There was a bump in the prices around 2010, and it never went further up from there. Maybe they're afraid of pricing out the indie developers? I have no idea.

The problem with a cheap Lotto is it levels the playing field TOO much, everybody and their mother who has any interest at all even considering going to the event will put their name in the hat, why not? Now, people who make their living for themselves or are representing a multi-million dollar company have to compete with hobbyists or really anybody who just wants to hang out

I'd prefer a solution that scales the cost so as to make tickets available for every level of 'desire':

-first 1000 tickets cost $500 to get people in who may have never been able to otherwise -the next 4000 cost $1500 -the next 2500 cost $3000 -the next 2000 cost $4000 -and the final 500 cost $6000 or more so there are ALWAYS some tickets at the upper end available if you REALLY have to go.

Or perhaps more conferences?

Geez back in the Microsoft heyday there were many different conferences. Sure, the PDC had 20K attendees if you wanted to go to that but there were other smaller conferences as well.

Hell, I'd settle for an east coast conference so I don't need to hang out with the skinny-jeans hipsters :)

More or bigger. Allowing for 20,000 attendees would be huge for WWDC. Currently there are only 5-6,000 tickets available. I imagine it would probably still sell out at 20,000, but it would be a lot less crazy, at least.

And a lot less useful. For the "talk to Apple's engineers" part, the current head count already is stretching it, if not overstretching it.

I must hang out in the unpopular labs. My experience has been that the engineers currently have a lot of down time and it's really easy to get their attention if you want it.