Tonight marks the 77th Emmy Awards ceremony, honoring the best series of the past year as chosen by voting members. Tomorrow, we’ll be flooded with articles tallying each streamer’s wins, but today I wanted to revisit an analysis published recently by Nielsen, whose methodological shortcomings astonish me once again for such a well-regarded institute.

In this study, Nielsen compiled all the titles nominated this year in the main Emmy categories and aggregated their viewing times between June 1, 2024, and May 31, 2025, in order to determine the most-watched nominated programs, measured in hours viewed.

Predictably, this is Nielsen, after all, the methodological biases are significant, since multi-season series had the viewing times of all their seasons added together, giving them a huge advantage over shows with far fewer episodes.

The observation period is also problematic, as I’ll illustrate with the most striking example: the TV Movie category.

In this category, we find films released as early as September 2024 (like Rebel Ridge), which had eight months to rack up viewing hours, but also films released literally just two full days before the end of the observation period, like Mountainhead on HBO Max, which premiered on May 30, 2025. And all of these titles are treated on equal footing. I’m not even mentioning the differences in each title’s runtime, but once again, these huge biases astonish me coming from Nielsen, whose studies are then cited across all U.S. specialized film and TV media.

As usual, when something bothers me, I wanted to tackle the question myself: ‘Which Emmy-nominated series this year were the most-watched?’ The twist here is that I’m not looking at hours viewed, but at U.S. audience numbers.

The answer seems impossible to determine, yet it’s surprisingly simple — because five months ago, Nielsen published a far more interesting ranking: the most-watched new series of the September 2024–early April 2025 period, across streaming and TV, measured over their first 35 days, in millions of viewers, which I’m sharing here:

This ranking has almost every advantage: it uses a common time frame for all titles (the first 35 days of release), it removes runtime biases (by using average viewers), and most importantly, it compares streaming titles to broadcast titles on the same scale, something other available rankings don’t do. In short, it’s a near-perfect measure, proving that Nielsen can indeed get it right when it wants to.

This ranking still has a few gaps, particularly for Emmy-nominated series released before September 2024 or after April 2025. But thanks to other Nielsen data (Top 10s and more), I can estimate the missing data with varying degrees of certainty and that’s exactly what we’re going to look at now.

Let’s start with the drama series category. Here, Nielsen crowns The White Lotus as the most-watched series in terms of minutes viewed, and this is also the conclusion I reach when switching to the 35-day average viewers metric for its season 4. The only unknown in determining the leader was the performance of The Last of Us season 2 released in April 2025 but with the data I have, I estimate it averaged just over 10 million viewers across its first 35 days.

Severance season 2 and The Pitt rank much lower in the methodology in average viewers compared to hours viewed, likely due to two factors: Severance accumulated a lot of minutes viewed from its season 1, which was rediscovered during season 2, while The Pitt has 15 one-hour episodes, which boosts its ranking in Nielsen’s original metric compared to say, Paradise’s 8 episodes. Andor season 2 was released after the observation period, but the figures available through Nielsen and its weekly Top 10s allow us to estimate it at around 3.5 million average viewers for the season. I have no way to gauge the audience for the latest season of Slow Horses, but it must be very, very low.

On the comedy side, it’s a bit of a mess, if you’ll pardon the expression, because Nielsen ranks Abbott Elementary and Only Murders in the Building as the top two in minutes viewed, while in their ranking by number of viewers, these two shows had almost half the average unique audience of Nobody Wants This on Netflix — which is actually the most-watched series in this category according to Nielsen, when measured using this alternative viewers-based methodology.

According to my estimates, The Bear season 3 was watched more than the latest seasons of Abbott Elementary and Only Murders in the Building. Nielsen’s study also finally allows us to estimate a viewership number for The Studio, and it’s very, very low, as expected. When a series doesn’t appear in any serious Top 10, regardless of methodology, it means it’s being watched at very minimal levels, which is likely also the case for Shrinking and What We Do in the Shadows.

Let’s wrap up with the miniseries ranking, which also highlights all the biases of the minutes-viewed metric. Adolescence ranks 4th out of 5 in minutes viewed, but when switching to the viewers-based metric, it comes in first, by far, with the largest audience of any series released on U.S. screens (both streaming and traditional networks) during the 2024/2025 season, second only to Squid Game according to Nielsen’s viewers-based ranking.

The reason is simple: the four short episodes of Adolescence heavily penalize it in Nielsen’s hours-viewed metric. But this clearly shows the absurdity of that approach, since Adolescence was watched by more viewers than 99.99% of all other series in the U.S. audiovisual landscape this season.

Behind it, The Penguin edges out Black Mirror, likely because it was released much earlier and therefore had more time to accumulate viewing hours. But according to my estimates, Black Mirror actually ranks ahead.

Nielsen’s study also provides a first viewership figure for Dying for Sex with Michelle Williams, which totaled around 1.2 million EVCs over its first 60 days. A very low total, but viewership isn’t what matters when it comes to being nominated for an Emmy.

It’s always a battle of methodology, but also of what we actually mean by ‘most-watched series.’

For Nielsen, a 20-episode series followed by 2 million viewers will always be considered more ‘watched’ than a 4-episode series followed by 9 million viewers — and this is where I fundamentally disagree with them, especially in the context of this study on Emmy-nominated series.

I really wish Nielsen would finally adopt a more rigorous approach to the studies they publish, with truly comparable scales from one title to another. Let’s say the publication of the Top 112 last May was a step in the right direction. In the meantime, we’re stuck with their flawed studies, which lead to flawed conclusions — a real shame when you aim to be the ultimate arbiter of an entire industry.

Enjoy the rest of your weekend! We’ll be back on Wednesday with a new study of Netflix viewership around the world.

If you enjoyed this issue of Netflix & Chiffres, why not share the newsletter with someone who might also be interested?

You can find me here: