Lets say we have some indie game, but good looking, quite succefull etc., lets say vanishing of ethan carter (just an example). Lets say that according to steam spy, the game sold 200 000 copies. And lets say that it came out in the beggining of of 2014 (just and example). So it was out for 2 years, endured around 6 "holiday" sale actions (summer sale, autumn sale,winter sale X 2 (for each year)) and 2 discount action/flash sales during the year. The game (besides the action sales) still sells for the full release price.

The steam spy says 200 000 copies sold. Is there a way how to at least a little bit APPROXIMATE the number of sales that happened for the full price and how much copies were sold for the discounted price? Before you say NO, lets at least try to think about one.

1) Majority of the games sell majority of their coppies in short space right after release. Whats the average time frame and copies sold? Lets say that on average game sells in its first month around 50% of its total sold copies (measured two years after the release). True?
2) Humble bundle- if the game sells on humble bundle you can easily subtract the ammount from steamspy numbers.
Whats other tips you have on this issue?

Is it safe to say that in my example, 100 000 copies (50%) were sold in the first month for full price , additional 20 000 (10%) sales were made slowly through the 2 years for the full price, and the rest (80 000 copies= 40%) were sold during various sale/discount actions (-25%, -50%, -75%)?

Lets say we have some indie game, but good looking, quite succefull etc., lets say vanishing of ethan carter (just an example). Lets say that according to steam spy, the game sold 200 000 copies. And lets say that it came out in the beggining of of 2014 (just and example). So it was out for 2 years, endured around 6 "holiday" sale actions (summer sale, autumn sale,winter sale X 2 (for each year)) and 2 discount action/flash sales during the year. The game (besides the action sales) still sells for the full release price.

The steam spy says 200 000 copies sold. Is there a way how to at least a little bit APPROXIMATE the number of sales that happened for the full price and how much copies were sold for the discounted price? Before you say NO, lets at least try to think about one.

1) Majority of the games sell majority of their coppies in short space right after release. Whats the average time frame and copies sold? Lets say that on average game sells in its first month around 50% of its total sold copies (measured two years after the release). True?
2) Humble bundle- if the game sells on humble bundle you can easily subtract the ammount from steamspy numbers.
Whats other tips you have on this issue?

Is it safe to say that in my example, 100 000 copies (50%) were sold in the first month for full price , additional 20 000 (10%) sales were made slowly through the 2 years for the full price, and the rest (80 000 copies= 40%) were sold during various sale/discount actions (-25%, -50%, -75%)?

AFAIK that information is confidential, and people are not allowed to tell the public... but:

Just look at the sales over time, and compare it to the times games where on sale. During a sale a game participates in, it is always sold at a reduced price.

Don't know if the information what game participated in what sales is out there somewhere (without you having to watch the game during all sales just to notice when this particular game gets on sale, and how much).

On the other hand, if there is a sales spike during a sale, you can bet with a high certainity that the game was on sale. You don't know how much, just pick a sensible average and it should even out over time (50%? 60%? 40%? Depends on how old the game is, if it still sells well and many other factors).

Now, it could still be that the game was on sale, but there wasn't a spike. Does that matter though? If the sales where bad, the amount of error introduced by counting these sales at full price is quite meaningless compared to the sales over the full lifetime.

So, TL;DR: even if you don't find the exact numbers, you can make a pretty educated guess that might get into +/- 10% of the real number... at least the ballpark is right.

And really, do you need more exact numbers? I guess not. Because all games are different, yours will likely sell at a very different pattern.

1) Majority of the games sell majority of their coppies in short space right after release.

That "fact" is a well-established fact for AAA games, and prior to digital distribution.

Unless you have modern up-to-date information for indie games with digital distribution, you're walking on false assumptions.

Whats the average time frame and copies sold? Lets say that on average game sells in its first month around 50% of its total sold copies (measured two years after the release). True?

That varies wildly per game - if you merely look at the mean average, you are looking at the information with one eye closed. For example, AAA games that are harder to pirate, continue to sell for longer, and sell more copies. AAA games that are easier to pirate sell less copies and with shorter lifespans.

Which completely invalidates the claim pirates like to make that "I wouldn't have bought it anyway". Obviously they wouldn't buy every game they pirate, but equally obviously, piracy does cost real sales, as proven by multiple games that had overwhelmingly incredible sales because it took pirates over a year to break the DRM, instead of the usual one week. Note: I'm not in favor of DRM, and prefer buying DRM-free. AAA studios know that for each additional week they can delay piracy, they make millions more dollars in sales.

Is it safe to say that in my example, 100 000 copies (50%) were sold in the first month for full price , additional 20 000 (10%) sales were made slowly through the 2 years for the full price, and the rest (80 000 copies= 40%) were sold during various sale/discount actions (-25%, -50%, -75%)?

No, that's not a safe guess. What with Steam sales, and how prominently the specific game is promoted in those sales (i.e. top-of-page banner, or front-page 'recommended titles', or etc...). it is entirely possible that for indie games that become popular by word-of-mouth, their later sales might eclipse their first-month sales.

One example of this, even without Steam sales - and with zero sales ever, is Minecraft. While granting that it is an extreme outlier, in later years, it sold more copies than it did even during it's entire first year of sales.

This may be a common trend actually, with pre-KickStarter alpha-funded games.

1) Majority of the games sell majority of their coppies in short space right after release.

Eh, not always.

I agree with Servant. Alphafunded (and episodic) games get sales over their lifetime. The developers develop then spread the word in weekly updates, rinse and repeat until the game/season is done. More exposure means more sales, all other things being equal.

Ok, so for a majority of games (not for all for GOD SAKE) , they sell a BIG BIG chunk right after release. What is the time window in which this happens, and after which the sales slow down considerably... I believe its 1 month after release... after that the sales slow down usually. Is this corrtect time window (1 month) or is it even shorter (1st week)?

Ok, so for a majority of games (not for all for GOD SAKE) , they sell a BIG BIG chunk right after release. What is the time window in which this happens, and after which the sales slow down considerably... I believe its 1 month after release... after that the sales slow down usually. Is this corrtect time window (1 month) or is it even shorter (1st week)?

I know your response, is trying to flippantly dismiss my post as pointless nitpicking, but my post really wasn't nitpicking - I'm trying to say you're using "facts" that are at least a decade outdated and no longer hold true, and you might want to re-think your assumptions.

The majority of triple-A games, prior to digital distribution, if their DRM was cracked quickly, sold the majority of their copies within the first three weeks or so, depending on the time of the year they were released in, and even then, other factors could also influence things - popular mods for a game could re-invigorate sales months or even years after release (Halflife and Battlefield 2 for example).

The mod example is semi-nitpicking, because those are outliers (as is Minecraft, for other reasons). The other four points (highlighted in different colors) are not nitpicking - they are very important qualifications that you shouldn't brush under the table, because they all apply directly to your situation and your games (or to whatever modern games you are wishing to examine).

Part of the reason for the big first few weeks of sales was the emphasis on graphics (every few months, a game with better graphics came out), and the huge marketing budgets. With modern indie games and digital distribution, big releases are still very important, but Steam sales and bundles and word-of-mouth are also very important, as there's a (somewhat) slower curve of indie games becoming "outdated" compared to AAA games.

The old statistics have changed. How exactly they've changed, I don't know; because most sales data isn't public hidden, there hasn't been huge studies on it (that I've seen) that analyze a large enough sampling of games, so we can only guess here and there. But we know the statistics have changed, because what sales information is released has proven the old "facts" either invalid, or much less absolute than they used to be - especially for non-AAA games.

As a human, I like to reduce down complex things to simple rules - which sometimes is useful, and othertimes can be harmful. But if you take my statement and boil it down to "the majority of copies are sold within the first three weeks", you might be doing yourself a disservice by boiling down the information too much.

Thus is my opinion from what I've read. If you do in-depth research and find solid conclusions, please share them! :)

You can find some sales data for some games here and there (sixty or seventy indie or mega-indie games), but so far I've never found anything large enough to do real studies on.