Superconductor

My Progress Bar

As I've mentioned in the past, I'm a big fan of the on-line game Progress Quest, a parody of MMORPGs in which you don't have to actually have to do anything to get your character to increase in level. If you really want to, you can watch the progress bar for your character slowly move towards 100 percent completion, but since that can actually take a few months at the higher levels, that's probably something that only government employees will be able to find the time for.

It seems that someone has taken the idea of a progress bar and applied it to another situation that many people might find useful: counting down to the end of their working day. The application that does this is My Progress Bar. Here's a screen shot of this fine application in operation.

It even has a "Disguise" setting, which labels the application's window as "Critical Patch Installation" instead of the default "Work Day Progress." Here's what that looks like.

If your company's security policy allows it, this is definitely the sort of application that you should consider installing and running.

Usability lessons from Progress Quest

Voltage is known for its innovative encryption technologies, but we're also known for how easy our products are to use. Not too many years ago, it was extremely hard for the average person to encrypt their email. The classic paper "Why Johnny Can't Encrypt" describes exactly how hard this can be for a typical user and anyone interested in the usability of encryption should read it.

With Voltage's SecureMail, on the other hand, a user doesn't have to do anything more than click on the "Send Secure" button instead of the "Send" button. If you're implementing SecureMail at a gateway appliance, they don't even have to do that – it can just happen automatically. Decrypting is just as easy.

Because we worry so much about the usability of our products, I'm very interested in seeing any enterprise security products that might actually be easier to use than SecureMail. If we ever find one of these, we'll probably be able to learn a thing or two from it. That's why I got so excited when I recently learned of an application that may actually be easier to use than SecureMail. In this case, however, it's not enterprise software. It's the game Progress Quest.

Progress Quest is a massively multiplayer online role-playing game (MMORPG). Before I heard of Progress Quest, I had never actually played a MMORPG, but that didn't stop me from being a government expert on the topic. I say that because I was actually the invited speaker at a government workshop on MMORPGs a couple of years ago. Unfortunately, the fact that I had to sign an NDA for this event means that I can't say much more about it.

Here's how the manual for Progress Quest describes the game:

Progress Quest is a next generation computer role-playing game. Gamers who have played modern online role-playing games, or almost any computer role-playing game, or who have at any time installed or upgraded their operating system, will find themselves incredibly comfortable with Progress Quest's very familiar gameplay. Progress Quest follows reverently in the footsteps of recent smash hit online worlds, but is careful to streamline the more tedious aspects of those offerings. Players will still have the satisfaction of building their character from a ninety-pound level 1 teenager, to an incredibly puissant, magically imbued warrior, well able to snuff out the lives of a barnload of bugbears without need of so much as a lunch break. Yet, gone are the tedious micromanagement and other frustrations common to that older generation of RPG's.

You start Progress Quest by picking the class and race of the character that you'll be playing. After that, the game does everything else for you. I even created a Progress Quest character: Elrond Hubbard, a Demicanadian Ur-Paladin with a name that's almost funny. If you're more adventurous you can pick races like Double Wookiee or Enchanted Motorcycle and classes like Fighter/Organist or Battle-Felon. I wasn't.

If you let Progress Quest run, your character will gradually increase in power and gain useful magical treasures. As I write this, Elrond Hubbard is currently Level 60 and has +23 Fine Gilded Plasma Vambraces. I'm not really sure if that's good or bad, but I certainly didn't have to pay any $9.95 monthly fees to get my character to where he is now.

Surprisingly enough, or at least surprisingly enough to surprise to a one-time government expert like me, Progress Quest seems to be fairly popular. The good reviews of it dramatically outnumber the bad reviews. And that's for a game where the player does absolutely nothing.

I'm never surprised to learn that most people really don't want to worry about encryption at all - they're too busy doing their jobs to worry about fighting with software that's hard to use. But I never would have thought that people would actually enjoy a game in which they do absolutely nothing.

In any event, I suppose that the bottom line is that we haven't quite figured out what we can learn from Progress Quest that will help us make SecureMail better, but that doesn't mean that we won't keep trying.

(If anyone wants to quote me about Progress Quest, here's my position on it: "Of all the games available for the PC, this is one of them.")

Pokemon combinatorics

Collectable card games are incredibly popular. The cards for these games are typically sold in packages of several cards. The manufacturers print several sheets of cards, and include cards from the different sheets in these packages with different frequencies. They might include four cards from sheet A, two cards from sheet B and a single card from sheet C, calling the cards from sheet A “common,” the ones from sheet B “uncommon” and the cards from sheet C “rare.” Your ability to complete a collection of the entire set of cards is clearly limited by your ability to collect all of the rare cards.

How many packages of cards can you expect to have to buy before you complete a set?

A solution to this problem was described by Paul Erdős and Alfréd Rényi in 1961. Here’s roughly how their solution goes.

Suppose that we already have k different rare cards out of a total set of n possible cards. The next time we buy a pack of cards the probability to not get a new rare card is k/n so the probability that we need to buy exactly p packs of cards to get another new rare card is

(k/n)^p-1 (1 - k/n)

so the expected number of packs of cards that we’ll need to buy to a new rare card is

Σ_p≥1 (k/n)^p-1 (1 - k/n) p = 1 / (1 - k/n)

= n / ( n - k)

So the expected number of packs of cards that we’ll need to buy to get all n rare cards is

Σ_k=0:n-1 n / ( n - k) = n/n + n/(n - 1) + … + n/2 + n/1

= n (1/n + 1/(n - 1) + … + 1/2 + 1/1)

= n H_n

So you should expect to buy packs equal to about H_n times the number of rare cards to get a complete set. If there are 50 rare cards in a set, for example, you should expect to buy about H₅₀ (approximately 4.5) times 50, or about 225 packs to get a complete set.

Here's how H_n grows as a function of n:

So for the number of rare cards in a typical set of collectable cards, a good rule of thumb is that you can expect to have to buy a number of packs of cards equal to between four and five times the number of rare cards to complete your set.