The Theory of Games: Diplomacy

Like many IT people, in college I was a gamer. And like many gamers, I spent time playing Diplomacy. Most board games are strategy games, but in Diplomacy you make deals with other players, and there are no rules that require you to keep them. Everyone writes down their orders and then they're all opened simultaneously, so you have to depend

This is an intoxicating freedom for people who've played bad chess for years: all of a sudden, they can win by making a deal and then betraying them. It's as though you could take back moving a piece in chess!

They soon learn that while the chess board has no memory, their fellow Diplomacy player does. Betray someone on Friday night and next week they'll return the favor. I once saw a guy wiped out on turn one of a game (this is nearly impossible) because he was so untrustworthy that, when opening round orders were opened, we found out that every one of his fellow players decided to launch an all out attack on him as their opening move.

What does this have to do with IT at a University? Just this. On a college campus, our customer base doesn't change. The people we work with this year are the people we'll be working with in ten years. Thinking short term doesn't work, because we live with the long term consequences of our actions. We need to find "win win" solutions

When several of us took a short course at the Business School, some of our fellow students were surprised to find out that we actually recommend that customers not buy some of our solutions, because there were other choices that were both cheaper and better, and our bosses knew this and were okay with that. My boss at the time knew that (for those solutions) we couldn't be the cheapest or the best, because we didn't have the volume that some commercial operations did. So we offered the service for when it "absolutely, positively, has to be there overnight" and steered everyone else to more reasonable solutions.

It's one of the things that went wrong in the mortgage market. People made subprime loans passed them on to others, and so there was no incentive to be ethical. The saying on Wall Street was "You'll be gone, I'll be gone, so let's do the deal". University IT can't live that way. They'll still be here and we'll still be here, so we need to do the right deal.

Starting a strategic effort: What to do first?

What's the first thing you do when starting a knowledge management project?

There are lots of choices, but I'd pick "Establish your feedback loop with your users" as an early favorite. We're doing something new for our users. They don't know exactly how they'll use our new service, so why should we?

Look at Flickr, the photo sharing service. It was intended as a feature for an online game called "Game NeverEnding", but it was completed before the rest of the game, the startup company realized they were on to something and threw away their original idea in favor of their photo sharing subsidiary, and a few years later sold Flickr to Yahoo for $30 million dollars.

You can no more know the future of how your users will use your work than Dilbert's boss can be alerted in advance of unscheduled outages. All you can do is plan to adapt, and for that you need to listen to your users.

Friends: The One Where We Talk About Technical Support

Years ago, people would ask me "What kind of computer should I get?". Like any academic, I answered their question with another question: "What do your friends use?", because the technical support available from their friends is far more important and useful than anything the company could provide. They will give you more time, you know how to communicate with them, and they probably understand the context of your problems better than some person in a call center halfway across the country.

These days I have friends ask me "Nikon or Canon?", and I say "Nikon", because that's what I have.  When I take a hiking trip with my friends Sean and Mary, for example, Sean and I can share a $1400 lens (which we'd use maybe one week out of the year) by renting it ($74/week at http://lensrentals.com). If he'd gone with Canon, he couldn't.

Technical specs matter to a point. But once you reach "good enough", the support of the people around you will matter far more.