I write to you today with feelings of warmth and appreciation for the elitist jerks of WoW. I’m not necessarily talking about the site, but about the people for whom the site was named: the people who endlessly debate the smallest minutia of optimization and then rail against anyone who disagrees with them, no matter to how small of a degree. These are the bullies who just don’t know how to communicate anything other than their smug superiority.
Don’t get me wrong, these people can be huge asshats. But the delightful thing about them is that they debate using data: raid logs or simcraft or Female Dwarf or even just raw math. They aren’t always right and there’s always room for more debate and name calling about the veracity of various models and corrections for real-WoW environments but darnit, they’re still using data to argue.
I’m feeling extra fond of these people because of my day job. Here is a long example of what I mean:
Once Upon a Time, in the Day Job
Frost: Hey boss, I have an idea called Plan A that will make us at least 30 million dollars! Here it is.
Boss: What? Really? How did you get your numbers?
Frost: Here was my method. As you can see it’s incredibly conservative and realistically I think it will be more like 70 million, but we can’t support any solid estimate for that extra — the 30 mil I can basically prove we can get though.
Boss: And you’ll also get some extra money from this other thing you ignored.
Frost: Yeah, but I suspect that will be a rounding error compared to the rest.
Boss: You are uncommonly good-looking! Let’s make this happen.
Boss: So Frost, bad news. Ops says that they can’t do this?
Frost: Why not?
Boss: They say it just can’t be done.
Frost: That’s ridiculous. Here is a list of sites that are capable of doing this, including every one of our competitors and including my WoW blog!
Boss: Good arguments, I shall assault them with it.
Boss: Bad news, they say they can’t do it because it will slow the load time of pages by 0.01 seconds.
Frost: Um… that sounds worth 30+ million dollars to me. How much do they say slowing the site will cost us?
Boss: They don’t. They’re measured on site uptime and speed. So making money doesn’t matter to their reviews.
Frost: Fair enough. I talked to some developer friends and here is their solution to do this and literally not slow load time of the site at all!
Boss: Awesome! I shall deliver the techno speak to them.
Boss: They say no, because they’re too busy.
Frost: I can’t help but notice that their reason for saying no keeps changing.
Boss: Yes. I hate them with every part of my body, including my wee wee.
… months pass and after schedules have cleared a bit we start making progress on forcing them do implement Plan A, which everyone we talk to continues to resist forcefully….
Boss: So, some other guys have another idea, Plan B, that they want to do.
Frost: I see how that will work well with Plan A in synergy.
Boss: No, they want to do Plan B only, and not use Plan A at all.
Frost: Damn them. How much money will Plan B make?
Boss: They have no idea.
Frost: Let’s test them!
At this point we meet with the Analytics department (and how cool is it to have an Analytics department?) about setting up a test of the two plans. Testing is fairly complicated, but the analytics people come up with a clever way to test it in phases — because if the first phase comes up one way then we don’t have to do the complicated stuff.
There is some obstacles as the Plan B people don’t want us to test Plan A the way we want because, as they say, “We can’t deploy what you describe — we’d have to build it first” so we agree to test it in a way that can be just deployed.
We ask for the test to happen asap, but then the Plan B people tell us they’ll need a couple months to build their solution.
Boss: Bad news Frost, the big wigs decided to use Plan B and not Plan A at all.
Frost: Wha…. what? Are they at least going to test them?
Boss: No, they don’t want to take the time to test them.
Frost: Have they come up with an estimate for how much their clearly inferior plan will make?
Boss: No. But they said we can revisit considering Plan A in a year.
Day Jobs Suck
… So that was a condensed and cutified version of what happened over the course of six months. I was in a lot of the meetings (which usually drove me crazy) and others only my boss was in.
The thing is, Plan B wasn’t a bad plan — it was way better than no plan at all — but it was also clearly worse than Plan A. All of the money Plan B would make was the rounding error that Plan A would also have made, in addition to the 30-70 mil.
And it’s far too common at just about any job I’ve worked at (not just the current day job) this is how decisions are made. By what appears to be randomness. It’s not just a matter of making decisions in the absence of data (bad) but in this case deliberately deciding not to collect any data at all and ignoring what data we did have. It’s baffling — I see it over and over and over again but I’ll never understand it.
Why not just get the data? Why not just make decisions based on the data you have? Why the hell just pull decisions out of the air???
Not in WoW
And this leads us back to the elitist jerks of the WoW world. For all their asshattery and poor communication, at least all their bickering is about data. Whether data is good or better or worse or how to interpret it — these are arguments I can get behind.
And you know what — most of the elitist jerks conversations out there are about the minutia because they agree on all the major points. Whichever rotation/gearing/glyphing you go with will probably very close to optimal, because their analysis is based on real data.
And that’s what separates WoW from the real world. And they wonder why we spend so much time in WoW…