An article on VGPG showed that the value for Michael Jackson's "Moonwalker" game on the Genesis rose from $15.00 to $200.00 in a matter of hours. Moonwalker is rated "3" on the Digital Press rarity scale, which in human terms means "somewhat uncommon but not hard to find." The value has since started to trickle back down.
My theory about video game pricing goes like this; all video games follow a curve. The price starts out high on release day, then slowly falls down towards zero as the game and associated system go out of vogue. There's a time when the game reaches its "sweet spot", the cheapest it will ever become, before it slowly starts climbing back up in value as it becomes collectible. The "sweet spot" is when the game is in clearance bins and the like. Earthbound for the SNES, fetching prices of $100 and above at local flea markets, at one point could be found for $9.99 or less in bargain bins across the country.
Along a game's journey upwards to collectability, there are a few anomalies affected by outside factors. It's like landing on some of the good squares in Monopoly. Final Fantasy VII became very expensive thanks to Advent Children. Moonwalker became expensive due to Michael Jackson's death. Super Mario RPG has been falling in value since it was released on the Wii's Virtual Console.
Brian Hughes has a different theory. He figures that because the trend in video games is moving towards an all-online distribution model, all games can do nothing but drop in value. He envisions a world with stacks of NES games worth less than five dollars a piece, since they can all be had online for a few bucks. This is an interesting world view for someone who runs a video games store.
I've always been fascinated with the concept of rarity. A game that is 1 ("R1") on the Digital Press rarity scale had millions of copies printed, and is thus far more common than one that is 7 ("R7"), which probably only had a print run of a hundred thousand or so. But yet, we see common games at high prices because of the demand. Zelda II: The Adventures of Link, is dirt common at R1. Because there are millions of copies of this game out there, the game should be practically worthless. But, the popularity has ensured that prices like $25 are common at the local flea markets.
On the other hand, VGPG shows that the value of The Adventures of Link is steadily dropping. Brian might be right; its release for the Wii's Virtual Console has probably helped to depress the price.
I have a different theory. I believe that rarity will ultimately determine the price of an item. It only makes sense. Something that is rare is more valuable than something that is not. This is why an ounce of gold is worth far more than an ounce of copper. Paying $25 for The Adventures of Link makes no sense, because the odds of finding a copy for $5 or less (say, at a garage sale) are far higher than finding a copy of Secret Scout (R7) for the same price. (Secret Scout, by the way, has done nothing but rise in value.)
Rare games are also not likely to be published on the likes of Virtual Console or Sony's Playstation Store. Rare games like Secret Scout, Tengen Tetris, and Cheetahmen II are unlicensed. Stadium Events was rebranded as World Class Track Meet. Nintendo World Championship is unlikely to be re-released online due to the specific nature of the game (it's more of a competition cartridge than a "game".)
It's going to take a while to get there, of course. Current demand still results in ridiculous prices, like $40.00 at a local flea market for a loose copy of Burgertime for the NES, despite the average price on eBay being $2.50. On the other hand, since rare games are usually not desirable, you can pick up some rarities for a bargain. A list of SNES games that I own shows that the rarity distribution is weirdly lopsided. I'm closer to completing full collections of rare games than commons. This is because a lot of vendors still price their common SNES games very high ($10 or more) despite being about as common as dirt.
So, I figure the market will correct itself slowly. Common games will become cheap, as they should be. They're also most likely to be re-released for online distribution. Rare games will become more expensive as time marches along. If this theory is correct, then the time to buy up rare games is now, before they rise in value, and then fill out with commons later, after they become cheap.
I've been working on VGDB for some time, and have struggled with a useful pricing model. There exist a few "states" that a game can be in: loose (game only), complete (game, box, manual), new (factory sealed, never opened), or somewhere between (game, manual, but no box, for example.) I've always tried to maintain two sets of prices: a "loose" price, and a "complete" price. This is further complicated by the fact that some systems tend towards loose games (like the NES, SNES, or Game Boy) while others towards complete games (Genesis, 3DO, Saturn.) Most of the time, if cardboard boxes are involved, most games on the market will be loose. If the boxes are made of some kind of tough plastic, then most games on the used market will still be available with their boxes intact.
This makes maintaining a database of relevant pricing data rather difficult. How do I present data that represents a sale of a game with manual, but no box? How do I authenticate that a certain sale is, in fact, of a sealed game?
I think the answer lies in moving to a flat pricing model. To simply poll all prices, and present the viewer with a minimum, a maximum, an average, and the boxing trend for that particular system (i.e., are most games on this system loose, or complete?) It makes it easier to process.
For example, let's say you see a copy of Super Turrican 2 for the SNES at a garage sale. This game is very rare, and priced at $15. The minimum pricing data on VGDB is $18, and the maximum is $39. That makes it a no-brainer: this is a steal of a deal. But let's say it's priced at $30 at the garage sale. The price is now on the higher end of the pricing spectrum. Maybe if the game is complete with box and manual, it's worth it. If the game was priced at $60, regardless of condition, it's likely too much.
That's the power of a min/max/avg "scale" system, I think. It allows you to easily make a decision based on the game's completeness, condition, rarity, and any other number of things you could think of. It's hard for a computer to determine these things for you, but easy for a human. It's easy for us to say, "values at the top end are probably for a complete game in good condition, while values at the bottom are for a loose game, maybe in rough shape."
I haven't started converting VGDB over to that kind of a system yet, but I think it's the best way to go. It makes it easier to dump pricing data into the database in bulk, since I don't have to categorize it first. There aren't as many numbers for your eyes to look at, making it easier to process in a situation where you're reviewing many titles at once. It gives you a broader outlook of what the pricing situation for that game is.