So, right now we're in the midst of a month-long blitz by... AmazonAppleGOGGamersGateSteamUbiSoftFuckingEveryone as they desperately cut prices on digital games in a "summer sale". If you haven't been exposed or if you're reading this after the fact, this is a sale where literally 20-50% of the games on the site are discounted by 30% or more - often things like 75%.
This simultaneous fire sale is bad. Or, rather, it's a sign that everything about the game retail market is shit. This sale indicates that all of the major digital retailers are fine taking a 50% hit to their prices and, in most cases, the game devs are similarly okay.
One of the reasons for this is that we're in the summer doldrums. Classically, studios don't like publishing in the summer, preferring to publish in the October-September sweet spot. So it is a time when there aren't very many new releases.
Another factor is the "long tail" idea. The overhead in distributing a digital game is very low, so nearly everything is revenue. After the first few months a game has been out, you've already gotten over your initial sale bump and now you're into vestigial sales. So your marketing preference should change from selling at what the market will bear to selling at what the market will be interested by. Hence many titles going for 75% off - if they can get people to buy their game this far past the initial hump, it's all just a final profit. It's a radical version of GameStop steadily marking down their new copies of games that never sold, except that in this case it works ten times faster and, due to a lack of inventory, it applies equally well to good and bad games.
This is especially true of the many games embracing pay DLC, or even as simple marketing for a game that you're about to put out. In those cases, even giving the game away for free would probably be a profit - but, of course, giving it away for free would be a bit weird-feeling. It's much easier to pretend it's a "sale".
And that's the issue.
We're not seeing something that is driven by the classic brick and mortar algorithms. There is no stock. There is no supply-side cost. The price of a digital game has nothing to do with anything store-related. It is 100% created out of whole cloth - every sale is revenue. So you try to find a line, a sweet spot where lowering the price doesn't increase the number of purchasers enough to be worth it, but raising the price would result in enough fewer that you'd lose out.
We like to pretend this price exists as some plausible ideal, but it doesn't. Not only is it largely defined by the whims and norms of the market, it also diminishes over time. Although in practical terms a game from 5 years ago is probably still just as fun as it was when released (aside from the very tippy-top of the graphics ladder), in reality the game is worth less, because fewer people are buying it. Decreasing the price over time will maximize your return, even if your game still looks fantastic and plays sharply.
These sales aren't sales. They are the prices these games should be, given those facts. Combined with pay DLC, the prices may still be too high.
So, this "fire sale", this clumsy mass-markdown which requires the community to do all the sorting and advertising, this is the way the market is. This is how prices will decay.
We just haven't normalized to it, yet, so we get there in these clumsy jumps.
Addendum: the worst part about these fire sales is that the games are often already pretty cheap. Lowering the prices doesn't actually make the games much more interesting, past a certain (already reached) point.
The thing older games really need is marketing, to remind people of how awesome they are. By simply lumping everything together under one discount banner, these fire sales do the exact opposite. Fortunately, the obsessive gamer geeks that watch the market are already around to repair the damage by marketing their favorite games via word of mouth. But that's a pretty clumsy way to do it.