NBC's 40 cameras should give officials an unprecedented number of angles to analyze in the occasion a replay challenge is issued during Sunday's Super Bowl. While it makes sense that the biggest game of the season would have more replay resources, there are also differences in the video coverage of regular season games—specifically, in the formats used by the various networks. Does the broadcast format of the network covering a particular NFL game affect the outcome of replay challenges?
There are two different HDTV formats used by networks broadcasting in the United States. NBC, CBS, and the NFL Network broadcast in 1080i30, which is an interlaced format of 29.97 frames per second, with each frame rendered at 1920 by 1080 pixels. Fox and ESPN, meanwhile, use 720p60, a progressive format delivering 59.94 frames per second, at a resolution of 1280 by 720 pixels.
Conventional wisdom (and personal experience) holds that the 720p format is "better for sports," because the increased frame rate results in smoother video without the artifacts introduced by deinterlacing 1080i video for viewing on a modern television. That having been said, 1080i video has nearly twice the resolution per frame than 720p. Thus, there's reason to wonder if the different video formats make a difference in whether or not a replay challenge succeeds.
Both kinds of feeds go into the same replay system. Since 2009, the NFL has used a pair of Harris Nexio video servers to store video information for replay at each NFL stadium. Sound And Vision went into great detail about this when the new system was implemented, but the gist is that the broadcast trucks send video feeds from each camera via fiber to the servers (one primary, one backup) which can then be independently managed by the on-field referee or the replay official in the booth. They view the replay footage on identical 26" Panasonic BT-LH2600W monitors.
We analyzed all 763 replay challenges over the past two regular- and postseasons in the NFL. Of these challenges, 451 upheld the call on the field while 310 overturned it. (Two replays were inconclusive due to equipment malfunction.) That's a 40 percent rate of reversal.
Splitting the results by broadcast format, we found that 44 percent of challenges on 1080i-broadcast games resulted in a reversal, compared to only 38 percent on 720p broadcasts. Taking the different number of challenges on each network into account, a replay challenge of a game broadcast on NBC, CBS, or the NFL Network was 5 percent more likely to reverse the call on the field than one on other networks. (We were unable to separate out upheld calls that were "confirmed" versus those where "the call stands" because the NFL doesn't make a distinction.)
It's possible this is just noise. A chi-square analysis of the results suggested those differences had an 87 percent chance of being related to the video format, and a 13 percent chance of being random. Science prefers results that clear a 95 percent cutoff. But since we aren't submitting this study to an academic journal, let's consider what could account for the difference, if it is real.
Each format delivers a different amount of raw information. One second of the 1080i video—with more pixels but fewer frames per second—contains 62,208,000 pixels. That's 13 percent more information than the same second of video viewed through the lower-resolution, higher frames-per-second 720p feed. That 13 percent could include the "incontrovertible video evidence" the league is looking for.
Except most of that extra information never reaches the refs. The monitor used by replay officials is incapable of displaying a full 1920-by-1080 image. Despite widespread availability of inexpensive full-HD monitors at your local Wal-Mart, the NFL is using screens with a maximum resolution of 1366 by 768 pixels.
That means that the 720p format ends up delivering more information per second. The 1080i feed still provides higher-resolution images—but only 14 percent higher, rather than the original 97 percent.
So if the refs are looking at a replay where the critical moment lasts less than 1/60 of a second—a glancing touch on a punt, say—they're probably better off with the higher frame rate of the Fox or ESPN feed. But if they're looking at something that needs higher resolution, like whether a player stepped on the sideline, the other networks have more information.
We're not sure why the NFL splurged on its replay server then skimped on the monitors. It's absurd that a replay official at Cowboys Stadium would have a higher-resolution image to observe on the Jerry-Tron than his or her official replay monitor. As for our study, there's still a few possibilities as to why games broadcast in 1080i have a higher rate of reversal:
* 1080i networks have better camera-placement strategies
* 1080i networks have better camera operators who capture more information-rich angles
* 1080i networks use better cameras/lenses
* AFC coaches (CBS broadcasts games when an AFC team is on the road) are "better" at using challenges
Again, though, there's a good chance the divide is just random noise.