Amazon Prime’s well-reviewed Philip K. Dick adaptation “The Man in the High Castle” is now the service’s “most-streamed” original series, surpassing the slightly more low-profile cop drama “Bosch.” That’s lovely for Amazon, which sent out a spiffy press release yesterday to mark the occasion, and it’s also nice for the talent both in front of and behind the camera on “High Castle.”

But how many people actually watched the show? Well, Amazon’s not telling.

Like most other subscription streaming services, Amazon doesn’t release viewership numbers, and it shouldn't have to. Since shows on Prime are not advertiser-supported, Amazon is under no obligation to report numbers to advertisers, which was one of the ways that ratings wormed their way into the public discourse, much to the chagrin of certain networks.

But the streaming service will occasionally issue press releases about the relative success of its series, just as Netflix executives will talk up how well its original shows are doing, without providing concrete numbers. This isn’t outright charlatanism, but it is smoke and mirrors. Reading reports and articles about streaming viewership is like watching someone attempt to solve an equation composed entirely of undefined variables: a long, frustrating and ultimately fruitless endeavor.

Subscriber Subterfuge

Unlike Netflix, which at least releases subscriber numbers every quarter, we don’t even know for sure how many Amazon Prime customers there are. (Arguably, Amazon isn't under any obligation to disclose that number either, although shareholders may be curious.) Analysts estimate there are anywhere from 44 million to 50 million Prime-ates in the United States, more than Netflix’s last count of 42 million U.S. subscribers. However, all Netflix subscribers are Netflix viewers; because Amazon Prime is not simply a streaming service — it began as free two-day shipping on Amazon purchases — not every Prime subscriber translates to a Prime viewer.

The lack of viewership information is just as frustrating to production companies, particularly those tied to traditional networks, which are nearly rabid for data at this point. Nielsen, the third-party media research firm that collects the ratings data the TV ecosystem runs on, had been thwarted in its attempts to measure streaming viewing in the past, particularly with Netflix, since Netflix strips out the audio codes that signal to Nielsen devices what episode of what show is being watched. But the ratings company says it’s found a workaround and has begun providing viewership data on acquired programming to the studios those shows were acquired from.

Reality vs. Perception

Streaming originals have a long theoretical shelf life, so it’s odd to trumpet a show’s performance within any kind of window, particularly without any context clues. “Bosch” may have been Amazon’s previous record-holder for streams within the first month of release, but without so much as a percentage indicator, that doesn’t mean anything to anyone outside of Amazon HQ. Telling visitors to the Amazon Prime section that a series is more popular than others is good marketing; media outlets doing the same is just odd — like reporting sound bites with no context. The week before Christmas is notoriously slow, news-wise, but that doesn't mean critical thinking should go out the window.

Ratings have never been an indicator of quality — the current ratings system was devised to make it easier to sell advertising, that’s all — but they are an indicator of popularity, and their absence shouldn’t mean any show can now be anointed a hit.

Or as one traditional TV exec noted to us, if anywhere near the same number of people were watching “House of Cards” as “NCIS” — with well over 17 million viewers a week — we’d definitely be hearing about it.