FX President John Landgraf, who started a loud rumble in the television industry last summer by declaring that there are too many TV shows, returned this summer to further explain why that's true.
If you have 500 scripted shows, he told TV writers Tuesday, hundreds of them are going to lose money - and by the basic laws of capitalism, producers will not do that indefinitely.
To reinforce his point, he cited research by the FX staff into the viewership and cost of scripted TV shows.
The top 20% of scripted shows across all platforms, he said, have an average audience of 10.8 million viewers.
The second quintile averages 4.4 million, and it descends to the bottom 20%, which averages 380,000.
At the same time, he said while that while the cost of producing and marketing shows can vary widely, the average hour of scripted television costs between $4 million and $5 million to produce and market.
"You can make money on the top three tiers," he said. "You can't make money on the bottom."
So hundreds of shows "are losing a lot of money," and at some point, he suggested, those unprofitable shows will no longer be made.
Landgraf admitted he was wrong in his prediction last summer that scripted show production had peaked at a little more than 400 shows in 2015.
While production at broadcast, basic cable and premium cable networks has leveled off, he said he had underestimated the number of new shows that would be produced by streaming services like Netflix, Hulu and Amazon.
The explosion there will likely push the total number of new scripted shows to 500 or more by 2018, he said.
After that, he said, he still expects the number of shows will level off and then slowly decrease.
Landgraf also said there is an upside to this explosion of shows.
"There are more opportunities for creative people," he said. "It also gives us a chance to broaden our storytelling.
"There is more great television now than at any time in history."
At the same time, he cautioned that the sheer volume of TV shows means we risk losing "the thread of a coherent conversation about what's good and what's great."
What won't be lost in the end, though, he predicts, is what pays and what doesn't.