Relative versus absolute, trend versus deviation; these distinctions are key in economics, and altogether far too lacking in media. Two recent ESPN articles that detail some record numbers produced by Super Bowl XLVIII are guilty of these errors, and more fundamentally, the sin of omission.
For the fourth time in five years, the Super Bowl has set a record for the most-watched television event in U.S. history, drawing 11.5 million viewers even though the Seattle Seahawks 43-8 victory over the Denver Broncos wasn’t really competitive.
Catchy! A record! Wowza! The Super Bowl is setting a new viewership record nearly every year, how about that? The headline “The American TV audience continues to grow slowly over time, increasing the viewership of different viewings across the board” is just not as catchy. But that is more or less what is happening. The percentage of the television audience watching the game actually declined from last year’s 48.1/71 rating to this year’s 47.6/70.1 Super Bowl XLVIII’s viewership was an absolute gain but a relative loss. The share rating at kickoff–44.5/70–actually was the highest of all Super Bowl kickoffs, suggesting that the blowout did indeed reduce viewership significantly.
The number of Super Bowl-related tweets was up from 24.1 million last year.
The same consideration applies here. There are millions more users on Twitter now than there were a year ago. Recently Twitter has averaged roughly 16 million new, active users every month.2 ESPN presents these figures as if the only thing that is different is that this is a new Super Bowl with different teams. But things change over time, and the state of the world, sans Super Bowl, is not the same today as it was a year ago.
Gamblers wagered a record $119.4 million at Nevada casinos on the Super Bowl, allowing sportsbooks to reap an unprecedented profit as the betting public lost out in Seattle’s rout of the Peyton Manning-led Denver Broncos.
While an interesting figure, do not take the gambling sum for anything particularly groundbreaking. Sports gambling has been increasing for a while, nearly doubling from a decade ago.3 This represents an increase, but the increase was foreseeable. The record numbers are not a significant deviation from the trend of annual growth.
Casinos paid out at 8-to-1 for the safety. Fans who bet that the first score would be on a safety cashed in at 60-to-1. It was the third year in a row that sportsbooks have been hit on the safety bet.
Though not a knock on ESPN, there are a couple interesting things in play here. Almost undoubtedly some people some where will remember this as if safeties are “getting hot” in the Super Bowl and think that a Super Bowl safety is more likely now. While significant rule changes (say, to intentional grounding or offensive holding) would affect the odds, without such things it is likely to pure dumb luck. Vegas knows this. Bookmakers might be a little ticked at the moment, but they may be able to take advantage next year, if people think safeties are suddenly more likely when, in reality, they are still quite rare.
On the flip side, perhaps after three consecutive safeties (which is ridiculous, if random), people will underestimate the chances of a Super Bowl safety next year. This is the Gambler’s Fallacy: a run of rare safeties does not mean that a safety is less likely; without changes to the state of the game (rules, etc.), a safety has a constant probability. As the sample size grows, the runs and streaks even out, eventually regressing to the mean. But in one particular game–say next year’s Super Bowl–the chance of a safety remains the same.
“The betting public certainly had their moments in this game. Some of the props that they bet on every year came through and they were rewarded,” said Jay Kornegay, who runs the LVH sportsbook.
This is great. Mr. Kornegay runs the sportsbook; of course he remarks that the gambling public had their moments! He is the one profiting from the gambling public! In other news, Lucy has remarked to Charlie Brown, “Hey, you’ve had your moments buddy! Tomorrow could be the day you finally kick the football!”
ESPN is terrific. But some of what makes their content interesting is what they fail to say.
- A program’s rating share is given as #A/#B. #A is the percent of all individuals with televisions who are watching the program. #B is the percent of all individuals watching television at the time who are watching the program. Of the American viewers who were watching something on television during Super Bowl XLVIII, 71 percent of them were watching the game. And do click here to verify these figures. ↩
- An active user is someone who logs on at least once a month. It varies, but generally between one-third and two-third of monthly active users log on daily. Click here for source. ↩
- Click here for source. ↩