Biting the Silver Bullet
Yesterday I had a discussion with a friend about the Silver Bullet Syndrome. Well, it's actually an ongoing discussion we have had for months. We've both been in companies that victim to this particular malady (it wasn't our fault, honest!), and we see it happen over and over again. We love grousing about it.

The Silver Bullet Syndrome is a name applied primarily to the IT industry for what basically amounts to a company falling for a Get Rich Quick Scheme. Some miracle technology or paradigm promises to solve all or most of an IT department's ills. It could be some tool that promises to improve productivity to ten times its current level. It could be the lure of offshore outsourcing, which promises to get the same job done by internal departments for a quarter (or less) of the costs. Or it could be some new programming language that's going to solve all the problems of the previous languages because it's so much easier to use.

Who wouldn't get weak in the knees at the prospect of becoming ten times more productive, or cutting their costs by 75%? Imagine me - a small business owner (very small, I'm afraid...) and game developer - if I could crank out a completed, polished game in two months instead of over a year... WOW! And as most of my expenses are living expenses... what if I could cut that to 1/4th of what it is now... I'll bet I could quit my day job and devote myself to making indie games! (Of course, I'm a weird case, in that my day job right now IS making games... so there's not a big difference.)

It is an incredible prospect. But like most things in life, if things sound too good to be true, they usually are. Why businesses keep falling for 'em is beyond me. I've been in a meeting where the IT Director (who had enough experience to know better) confidently promised that after only nine weeks of development on this new miracle platform, the consultants from the platform's company and select internal staff (both of whom were new to the company) had the new system that would replace our entire infrastructure 90% complete. Yep, we just needed five or six more weeks to integrate it with a few external systems and we'd be done.

About a year (and, depending on how you count, 2-5 million dollars) later, the entire project was scrapped, and said IT Director was replaced.

I've heard first-person horror stories of "offshore outsourcing" that ended with the company pulling everything back in-house after one to two years, and being forced to scrap nearly everything done in that time because it was simply unusable. Suddenly a 75% discount doesn't seem like such a bargain when you've wasted all that money AND two years of opportunity.

I'm still waiting for the CASE tools to revolutionize software development and make my job obsolete, as I was told in the late 80's. The promise of Object Oriented Development, which was preached to me at the same time in college, has largely been unrealized. But I continue to hear people loudly proclaim, "But this time it's different," for whatever the new system or technology has them excited.

Now here's the thing. I don't believe in silver bullets. I don't believe in any other kind of "get rich quick" scheme. Oh, I know it's technically possible, but it's usually non-repeatable or useful only in a very specific opportunity. However, I have found that if you haven't paid too much for that silver bullet, if you strip away the silver plating the marketers throw on it, you may find a high caliber hollowpoint round that is perfectly lethal for most of the more mundane monsters we face in this business. Just try to avoid any werewolves, and you should be fine.

In these cases, if you adopt the technology or methodology, you'll typically go through three phases:

Stage 1: The Honeymoon Phase - "This technology is the greatest thing since sliced bread. Looky-here, it used to take me a week to do this one thing, but now it only takes me an hour! I'm in heaven!"

Stage 2: The Welcome-To-Reality Phase - "Uhoh, this isn't doing what I thought it would do. And all of these other things I normally do are taking even longer to do. I've got to re-learn all this other stuff, too. Oh, crap, what did I pay for?"

Then, assuming you haven't been sold pure snake-oil:
Stage 3: The Acceptance Phase - "Okay. So maybe this isn't the be-all, end-all. Maybe it's got some real ugly limitations I have to work around. Still, I think the pros will outweigh the cons in the long run, and it'll be a net gain. I just have to learn how to do it."

Some Examples:
Extreme Programming (XP). I'm a big proponent of agile methodologies after my experience in a shop that practiced XP. Were the results dramatic over other methodologies? Well, it depends on what you mean by dramatic, and which methodologies you used. If you are talking pure waterfall development, sure. Just about anything is dramatically better than waterfall in MOST real-world applications. XP seemed to dramatically increase the costs of some phases of development, but dramatically decrease others, and coupled with the secondary benefits of the methodology it seemed like a net win. I think other agile practices might yield better results, though.

Game Engines - Hey, this IS a gaming site after all, I should at least mention games or game development, right? Game engines are a mixed bag. You can get fairly dramatic results if you are willing to accept some extreme limitations with some game development "kits." Or you can accept the steeper learning curve and more work involved with a highly flexible engine that exposes its source code. With luck, it'll take less time and effort than creating your own engine from scratch, but it's still not a "slam dunk." Making a commercially viable game isn't easy no matter what tools you adopt.

Admittedly, I'm finding myself in a bit of a honeymoon period with the Torque Game Builder. I'm pretty impressed with its functionality and ease-of-use (though I'm irritated by a couple of glaring bugs, too). Fortunately, GarageGames hasn't been blatently billing this thing as a miracle machine or a silver bullet, and it's priced well below the point where it would need to be. People are having a lot of luck with it in developing casual games (specifically), but those typically have short development times to begin with. It is possible to follow the tutorials and create a pretty respectable tech demo in a matter of a couple of hours. Just don't make the mistake of assuming your new game is 90% complete at that point.

Just remember to keep your expectations realistic, especially when deciding what you are going to pay - in terms of actual cost and in terms of the time and effort it will take to adopt it. Some caution and good sense can keep you from turning a frustrating situation into a disaster.
- posted by The Rampant Coyote @ 11:01 AM (Permalink)