Whether at the casino or the race track, the house always wins. That’s the way it has always been, too, with Internet advertising. Nearly all Internet ad dollars are spent in two ways: 1) buying ads from advertising networks whether that network is Google or Yahoo or even IDGTechNet, which sells space on this rag, or; 2) buying search terms — the right to have your ad shown every time someone searches on the word hermaphrodite, for example. Network profit in those transactions comes from arbitrage — buying low and selling high. But what if there was a more efficient way to buy and sell Internet ads? As of this morning it looks like there is a better way called Clickochet (for Click Richochet), the first ad trading network.

Clickochet is the baby of my old friend Paul Tyma and I have no financial or other interest in the product or the company. It’s just something cool to write about. I want you to know about it because it is different, probably a lot more efficient than competing solutions, and it might just change the world.

Paul is a great programmer who spent a number of years working as a senior developer at Google where his major achievement was a major rewrite of Gmail. He is also the author of Mailinator, one of the first and best anonymous e-mail services. But the idea for Clickochet sprang not from anything at Google, but from a Christmas gift Paul gave to his girlfriend Tanna, who runs a celebrity gossip blog called StarSnarks.com.

“She started it last fall and is pretty passionate about it,” Paul explained. “At some point she put up AdSense on the site and, no kidding, was making about five cents a day. She didn’t have a ton of traffic so it wasn’t terribly surprising. As a Christmas gift I bought her $100 worth of AdWords to help her get traffic. With no exaggeration this campaign lasted three days and yielded a total of 100 clicks. If you do the math, I successfully turned $100 into about 15 cents by converting to and from ad impressions.”

Ad networks work by selling ads, but if you close the loop between publishing and advertising there is an exchange rate between money and ad impressions. Existing ad networks create a market inefficiency in that exchange to make money. Using Tanna’s example, above, $100 bought $0.15 worth of ad impressions. That’s $99.85 in gross profit there, some of which goes to web sites participating in the program but most goes straight into the massive fuel tanks of Eric Schmidt’s G5 jet.

That is the way Google makes so much frigging money.

All web sites want traffic, yet I won’t buy ads for cringely.com. Why not? Because it is a non-economical transaction. The only revenue I get from this site is from ads, but as the example above shows it would be insane for me to buy ads to sell ads. That’s a perpetual motion machine and perpetual motion machines are always defeated by friction — in this case the friction of the ad network’s revenue cut. This is the market inefficiency that ad networks introduce. Ad networks provide value in other ways, but this closed-loop is not it.

It might make sense to buy ads for this site if I was offering more than just ideas. If I was selling diamonds or penis enlargement pills, okay, but buying ads just to sell ads is a fool’s game. As always, though, there are plenty of fools.

That’s when Paul came up with the idea of trading ad impressions. Tanna could show an ad for other small sites like hers and those other small sites could show her ad. It’s like a link exchange but way smarter because it isn’t just for Search Engine Optimization (SEO). It’s an ad trading network.

Clickochet is free, social, and tied to Facebook and Twitter. Clickochet is a social ad community for web site owners like me for whom buying ads doesn’t make economic sense.

“I realized right away that ad-for-ad didn’t exactly work,” Paul said. “First, when you show a normal text banner ad, you’re not just showing one ad — you are showing three. There are three text ads inside a 728×90 banner ad creative. I can pass that multiplier back to the Clickochet member. So to start, if Tanna shows 1000 banners in one day, I can show 3000 of her smaller ads across the network.”

Enter at this point the inevitable honking-big algorithm to make this all work, because there are many factors that come into play when doing transactions like this — factors like ad placement, ad quality, and even my reputation as a journalist — to determine the exchange rate for each ad. So instead of taking the traditional ad network route of converting clicks directly into dollars (CPM), Clickochet converts ad impressions into virtual currency.

Here’s where it gets very geeky and leaves traditional ad networks in the dust, because where a DoubleClick would take the money and run, Clickochet introduces a clever multiplicative effect.

On Tanna’s blog right now she’s showing a skyscraper containing five ad-equivalents. So showing one ad on Starsnarks could get up to five ads for her site showing on the network. She can get more out than she puts in.

There are 40 billion webpages in Google’s index, almost none of which make a profit by selling ads. Clickochet is, in some respects, an ad network for the long tail — the 39 billion web pages where participating in the present monetary ad system makes little sense. For those pages gaining traffic is important and forgoing nickel-a-day ad revenue is painless.

Clickochet removes the market inefficiency that ad networks create to make money. There are two facets to that inefficiency. One is money: the networks take their cut. But the other is time. If you took your AdSense money and paid into AdWords with it you’d be delayed 30 days because Google waits that long to pay and so do all the other networks. You finance Google. But Clickochet is an instant transaction. The moment you get an ad impression, you can configure the system to either bank the credits, or spend them immediately. Show an ad on your site, and 3 of your smaller ads show nearly instantly elsewhere on the network.

And unlike the $100 AdWords campaign Paul bought for Tanna, a Clickochet campaign runs indefinitely. You get out of the system only what you put in, but on a perpetual basis. If Tanna is only putting 1000 ad impressions a day into the system, she won’t get 1000 new users overnight, but will increase her traffic over time.

Because Clickochet converts ad impressions into virtual currency, you can save them instead of spending them right away. Say you are launching a new product or even a new company in six months. You can run ads on your knitting blog, bank the credits, then spend all your money in a big ad blast when the product or company is ready to go. Your friends and family investors can even pay you in ad impressions if they like.

Here’s another even more subtle effect. I have a link on this site for portraitquilts.com, which is run by my little sister. She wants the link mainly for Search Engine Optimization but doesn’t mind if you buy a quilt or a pillow with Grandma’s picture on it (or Elvis’s). I get a lot of traffic here, which is good for Sis to some extent, but my readership is also very stable. The same people keep coming back week after week after week. Eventually you’ll stop clicking on that quilt link and my good brother brownie points will cease to accumulate. The link will go stale as advertising, if it hasn’t already.

With Clickochet, however, I could distribute ads for my sister’s site elsewhere on the ad network — perhaps even on sites more relevant to photo quilts than this blog. I can directly convert my ad impressions into my sister’s. To be fair, Google would let me do this too — make money off cringely.com and buy ads for portraitquilts.com — but I’d never do that because it would be a losing proposition.

Clickochet is a great idea and since it is coming from a great developer I’m pretty sure it will work technically. But can Clickochet gain critical mass? That’s the problem faced by all new networks: they need lots of traffic to succeed but, absent some viral effect they don’t have any traffic to start. PayPal famously handled that simply by sending people money, but most networks can’t afford that gimmick so they die. What makes Clickochet any different?

The clever answer here is that Clickochet may well have all the mass it needs to start, courtesy of Paul’s other site — Mailinator.com.

“I’m doing a virtual economic stimulus,” Paul explained “I’m injecting Mailinator’s several hundred thousand ad impressions into the system every day, without taking any out. Active users could get a lot of free ad impressions.”

Mailinator’s ad inventory allows Clickochet on its first day (today) to start where it might otherwise be a week or a month into viral growth. Paul hopes that’s past the most likely failure point.

Yeah, but how does Clickochet expect to make money?

Initially it will be as a cheaper competitor to AdWords. Many sites will find this to be a new (and cheap) venue for advertising. They will simply buy ads on the network just as they would through Google. But where the term “mesothelioma” costs about $50 on AdWords, it could be much cheaper on Clickochet.

“We’ll level the cost of ads, ” says Paul. “This won’t help with cheap search terms but creates a new outlet for currently expensive ones. With the right ad sales team in place what percent of that market could we capture?”

We should all want Clickochet to succeed just the same way we want Southwest Airlines to start serving our local airport — because it will drive down the profit margins for all ad networks, increasing transparency and efficiency and saving money for us all.

The house will still win, but their profits won’t be so obscene at our expense.