On Sunday, October 25th, the NFL (with the help of Yahoo! and others) streamed the first, globally free NFL game. No cable subscription needed. No online credentials required. Just a connection to the Internet and a desire to see the Jacksonville Jaguars match up against the Buffalo Bills. 15.6 million unique visitors “tuned in” to see what turned out to be a barn-burner.
This was truly a watershed moment in today’s media landscape. Even as more people are cutting or shaving the cord, the vast majority still cling to their cable subscriptions because of live events (mostly sports) that are only available through traditional broadcast. Only this event demonstrated that it’s possible to deliver “broadcast-quality” live events over the Internet and generate the kind of viewership that we generally see with terrestrial feeds.
So what did it take to pull this off? If you take a step back and let your imagination go, you can begin to see all the moving parts that were required to make Internet magic. From encoding the broadcast streams to something we could view online to distributing that content on any device to the advertising—whew. It’s a lot to imagine. The illustration below (provided by Yahoo! Engineering on their Tumblr page) gives you a 50,000-foot level picture of the mechanics in delivering a live, online event at broadcast quality with the scale to reach a global audience on any device.
And here are the specific details on what was actually delivered:
- 33.6 million streams and 15.2 million unique viewers
- an average rebuffering ratio of nearly 1%
- over 8.5 petabytes to end users
- the stream reached HD levels, with max bitrates over 6.74 Megabits (Mbps) per second and 60 frames per second (fps).
From a business standpoint, though, what this event demonstrates is that “it takes a village.” Lots of organizations looking to deliver streaming video content, especially live, believe they only need a single CDN like Akamai. Only as Yahoo! and the NFL demonstrated, that’s not the case. In fact, according to streaming-industry experts like Dan Rayburn, there were a host of CDNs involved including Akamai, Limelight, Level3, and Edgecast.
Putting aside the mechanics of making the event happen, what was it like to watch the game? For the most part, it was equal in quality to a television broadcast. In fact, according to Chris Chase at USA Today, “other than the medium, it was virtually indistinguishable from a CBS production.” It would seem that Yahoo! was, in fact, successful in delivering an online event at broadcast quality. Even in the hands of Dan Rayburn, the stream was pretty much flawless:
“I tested the stream being played back via TiVo, Roku, Xbox, Fire TV, Apple TV, MacBook, iPhone and iPad and experienced very few problems. The stream was split between multiple content delivery networks including Akamai, Limelight, Level 3 and Verizon amongst others. Streams started up fast, within 1-2 seconds at most and I never experienced any buffering or stuttering, aside from the stream to the TiVo. That stream, being delivered by Akamai, had multiple issues…”
But given the nature of the Internet (and its variability of quality; there are no QoS services running on the Internet to ensure flawless delivery), as you might expect, the distribution wasn’t without it’s problems. In fact, Chase further cites that, “[I] talked to about a half-dozen people across the country and half seemed to have at least double or triple the buffering problems that I had in the D.C. suburbs…” And according to Julia Boorstin at CNBC, “There were many complaints on social media and other outlets that parts of the game buffered or were pixilated. Others noted that watching the feed on a TV — via the NFL app on Xbox Live — was not as smooth as on mobile devices. (And of course, people are likely to want to watch football on a giant screen, rather than a smartphone, given the choice.)” There’s even a great list of Tweets from people about the broadcast over on Geekwire.
It’s not surprising that people would encounter some issues, though, because when it comes down to it, there are simply factors outside of Yahoo! (or their partners’) control—bad Internet connections, ISP congestion, over-taxed computers. Still, all-in-all, the NFL was extremely pleased with the outcome:
“We’re thrilled with the results of our initial step distributing an NFL game to a worldwide audience and with the work of our partner, Yahoo,” said Hans Schroeder, Senior Vice President, Media Strategy, Business Development & Sales for the NFL. “We are incredibly excited by the fact that we took a game that would have been viewed by a relatively limited television audience in the United States and by distributing it digitally were able to attract a global audience of over 15 million viewers.”
As an avid football fan (and a Limelight employee) I naturally logged on to see what all the fuss was about and was quite amazed by the level of streaming quality. Sure, there were times the picture was pixilated (I was watching through an AppleTV) but I probably have a better understanding of streaming video than your average consumer. I wasn’t going to blame Yahoo! or the NFL when there was a lot of moving parts between my AppleTV and the content for something to go wrong.
Now a week later, in looking back and thinking about this game, I firmly believe it was a watershed moment. Not because this is what ESPN, MLB, and others have been doing for a while now, but because of what it represented—the NFL took its normal television-broadcast workflow and transported most of that to the Internet (the original signal was still backhauled via satellite). Not only could you watch the broadcast from a variety of screens (which you really can’t with normal television), but there was no authentication involved and it included a full compliment of advertising. To parrot again what Chris Chase from USA Today said, “…it was virtually indistinguishable from a CBS production.” That’s the watershed moment here, the real tipping point. I have no doubt that we will look back on this moment as a pivotal juncture when we transitioned from the traditional television experience to something new.
“Looking to the future, we expect live sporting events to be routinely streamed over the Internet to massive global audiences. People will expect these broadcasts to be flawless, with better than HD quality. October 25th 2015 was a significant step towards this vision.” [Yahoo!]
What’s next for the sporting world online? I’m excited about what this portends for the future. Will the NFL consider online-only packages for individual teams? If the recent class-action lawsuit against the NHL portends anything, the answer is a resounding yes (go Steelers). Soon, I’m sure, I’ll be able to just order up the games or teams I want. And combined with data and interactivity that’s only possible with an online broadcast, it will be a much richer experience than I currently get from a traditional feed. Sure, I’ll still use my 60” television to watch the signal, but it will finally be on my terms.
 What’s a good rebuffering ratio? According to firms like NICE PEOPLE AT WORK, the publishers of the video analytics tool YOUBORA, 0% - .5% is considered mild, .5% - 2% is considered moderate, and anything over 2% is critical. What’s more, when users experience a rebuffer ratio of 1% or lower, they watch 2.3x more content.