Those kind of numbers are eye-popping to me, as he also noted earlier that it was around 40k 18 months ago, and can actually dip to around 15k if he's more realistic with the analysis.
What do you think is causing the dip, if there is one? Gun to my head, I think it's the oversaturation problem Valve is tackling right now with Steam Labs and their other initiatives.
You guys citing Steam Labs:
You do realize that Steam Labs could very well have the effect of driving DOWN the median revenue of all Steam games.
Theoretically, there's an overall spend on Steam that doesn't change for a given set of users (ie - people only have so much money to spend on games); the idea that people aren't spending to their maximal amount (ie - they have money to spend but don't know which games to spend it on) generally means they can't find games that they want to play; perfect singular curation to people will generally drive consumers to spend more concentratedly on high quality games (ie - people won't buy "asset flips or garbage games that shouldn't be on Steam in the first place").
This will drive a larger sum of the overall spend to the highest quality games (on average) which will shift the median income for a game on Steam lower and lower in absolute dollars.
Which is why pointing to the median is kind of useless.
If my belief is that Steam is FLOODED WITH GARBAGE, it should make me happy to know that the median income is absolutely bonkers-low. Because that means garbage isn't being bought, overall spend is concentrated at fewer (theoretically 'the best') titles.
Honestly, if the overall spend on Steam is growing, but the median revenue is dropping, that kind of shows that people are generally buying what they want and happy with it. But we don't know that, because just telling people the median is mostly useless.
An average (mean), with standard deviation points would be far more useful.
Hell, a mode would be more useful.