I write The Numbers Guy, a twice-monthly column on how numbers are used and misused in the news, business and politics, for the Wall Street Journal's Marketplace page. I also write a blog by the same name, with posts most weekdays, for WSJ.com. And listen to an interview with me, conducted by Rita Schulte for the University of Illinois' WILL-580. (Click here and scroll down to Wednesday.) Here's another, with NPR's "On the Media"; I appear on the segment that starts about 25 minutes into the audio file. Numbers Guy won second place in the online category of the 2006 National Society of Newspaper Columnists contest.
Countdown Clocks Offer Little Information, June 29, 2007
Count-up and countdown clocks, such as the ones that track world population or AIDS, pack information compactly into a compelling, even frightening, message. The trouble is that they're not very precise. Plus, a blog post and reader discussion about the column.
How to Measure Class Gap in Reading?, June 15, 2007
I dissect the questionable child-reading numbers for kids from different income levels. Plus, a blog post and reader discussion about the column.
Flaws in Measuring the Poor May Hinder Solutions, June 1, 2007
"Our dream is a world free of poverty" is a slogan at the World Bank. But to some economists, the World Bank's definition of poverty is flawed and arbitrary. I take a look at how the numbers are calculated. Plus, a blog post and reader discussion about the column.
The Odds of Sinking 10 Holes in One, May 18, 2007
Jacqueline Gagne, a 46-year-old Rancho Mirage, Calif., resident, has hit 10 holes in one in the last four months, over a stretch of just 75 rounds of golf. I examine her improbable featand estimates of just how improbable it is. Plus, a blog post and reader discussion about the column.
Figuring the Impact of Allowing Felons to Vote, May 4, 2007
The numbers used to calculate the voting impact of Florida's move to restore civil rights to felonsthe counts of people affected, the percentage likely to vote and their projected party breakdownare fraught with uncertainty. Plus, a blog post and reader discussion about the column.
Both Sides in Gun Debate Push Dubious Data, April 20, 2007
The Virginia Tech shootings have reignited the U.S. gun-control debate, and advocates on both sides are using questionable numbers to support their arguments. I take a closer look at some of these numbers. Plus, a blog post and reader discussion about the column.
I'll Take 'Probability' for $2,400, Alex: Odds of Three-Way Tie, April 6, 2007
The first-ever three-way tie on "Jeopardy" led the game show's distributor to label the probability 1 in 25 million. But would it really take 75,000 years for another three-way tie? Not likely. Plus, a blog post and reader discussion about the column.
A Few Sales Tricks Can Launch a Book to Top of Online Lists, March 23, 2007
Online book-sales rankings are an object of obsession for authors because of their "democratic" potential. But if this is the democracy of bookselling, vote buying is an optionand public-relations firm Ruder Finn says it's figured out how to do just that. Plus, a blog post and reader discussion about the column.
Odds of 'Lost Tomb' Being Jesus' Family Rest on Assumptions, March 9, 2007
University of Toronto statistician Andrey Feuerverger's calculation that there is just a one-in-600 chance that "The Lost Tomb of Jesus" didn't house the bones of Jesus of Nazareth and his family helped propel an already explosive story. Plus, a blog post and reader discussion about the column, my first for the print edition.
Coming Soon: 'The Number 24', February 23, 2007
"The Number 23," a new thriller starring Jim Carrey, shows how just about any number can appear to have eerie properties if you look hard enough. Such as the number 24. Plus, a reader forum about the column.
Farm Group Sows Questionable Stat, February 15, 2007
A trade group for farmers is promoting a holiday aimed at "celebrating the continued affordability of food." But the group uses some murky math to crunch the numbers on how much Americans spend on food. Plus, a reader forum about the column.
Nielsen's New College Numbers, February 9, 2007
Nielsen is expanding its TV ratings to include college students living in dorms, a group that had previously been excluded from the research firm's numbers. The students are already having a huge impact on ratings for some shows, but national stats are being based on a survey of just 130 viewers. Plus, a reader forum about the column.
Coke's Contest Takes Time (and Soda), February 1, 2007
A promotion from Coke offers big-screen TVs, expensive trips and other prizes to loyal Coke drinkers. But to obtain the loot, customers need to jump through a lot of hoops. Claiming some of the biggest prizes could require you to drink thousands of bottles of soda, and repeatedly visit a Web site over several months. Plus, a reader forum about the column.
Winning a Longer Life, January 26, 2007
Can winning a Nobel Prize or Oscar really add years to your life? Carl Bialik examines scientific studies that purport to measure the effect of status on longevity. Plus, a reader forum about the column.
Another Look at Murder Stats, January 19, 2007
Murder stats are making headlines around the country, as cities publish their 2006 figures. But those numbers may not be the best gauge of violent crime, or a city's safety. Here's a look at two theories from criminologists that question the way the numbers are used. Plus, a reader forum about the column.
Rethinking Mileage Estimates, January 12, 2007
Later this year, car shoppers will see a drop in the fuel-economy estimates posted on new cars. The EPA is recalculating the numbers in an effort to better capture real-world conditions -- an update critics say is long overdue. Plus, a reader forum about the column.
Counting Internet Users, January 5, 2007
The Internet and all its high-tech tracking tools should make it easier to count Web searchers, video viewers and videogame players. But different methodologies and a lack of data transparency can lead to some dubious stats. Plus, a reader forum about the column.
Forrester's Controversial iTunes Report, December 21, 2006
A research report suggesting a steep drop in sales at Apple's iTunes store rattled investors. It also prompted reports that painted a rosy picture of iTunes. The uncertainty shows what happens when different people measure different numbers, using widely varying methods. Plus, a reader forum about the column.
Internet Video Stars Are Hard to Count, December 15, 2006
Was the Internet video of the teenage boy pretending to wield a "Star Wars" lightsaber really viewed 900 million times? A closer look at a marketing firm's ranking of the most-watched Web clips finds more science fiction than math. Plus, a reader forum about the column.
The Numbers Behind Pfizer's Decision, December 6, 2006
Pfizer abandoned a potentially blockbuster cholesterol drug after some patients died during clinical trials. The number of deaths wasn't much bigger than the number of people in a control group who also died. But a closer look shows how small differences can have big impacts in medical trials. Plus, a reader forum about the column.
How Many Kids Have Autism?, November 29, 2006
A number of media reports and public-awareness campaigns have reported that one in 166 U.S. children has autism. But the number doesn't tell the whole story. Plus, a reader forum about the column.
Grading the Pollsters, November 16, 2006
Pollsters earned overall high marks in last week's elections, accurately predicting the broad Democrat victory. Some did better than others, and firms haven't hesitated to trumpet their results. I look at the difficulties in evaluating polls, and crunches his own report card. Plus, a reader forum about the column.
McDonald's Big Prize Has Long Odds, November 9, 2006
McDonald's has upped the stakes in its heavily promoted Monopoly contest by increasing the grand prize to $5 million, but players are more likely to win the multistate Powerball lottery than collect the four winning game pieces. Plus, a reader forum about the column.
Clocking Footballs, Baseball Style, November 3, 2006
Football announcers are using sophisticated new tools to tell viewers how fast quarterbacks throw passes on key plays. But why are they reporting the stats as if they were baseball pitches? Plus, a reader forum about the column.
Crunching Numbers on a Coach's Call, October 26, 2006
Did a bad decision by the Mets' manager in the bottom of the ninth inning cost the team a ticket to the World Series? A look at how some baseball fans are crunching decades of game stats to come up with some insight. Plus, a reader forum about the column.
Trying to Herd a Cat Stat, October 12, 2006
Can a single female cat and her offspring really produce 420,000 cats over just seven years? Media reports and animal-adoption groups frequently repeat the startling number, but no one wants to claim ownership of this stray stat. Plus, a reader forum about the column.
A Closer Look at Sports Miracles, September 28, 2006
Father-and-son amateur golfers ace the same hole in the same round. Four Dodgers hit consecutive home runs in the bottom of the ninth. Just how unlikely are these feats? Plus, a reader forum about the column.
How Random Is the iPod's Shuffle?, September 21, 2006
IPod users have long suspected that the music players "shuffle" function serves up songs in a predictable pattern. A closer look shows just how difficult it is to truly randomize information. Plus, a reader forum about the column.
Grading Weather Forecasts, September 14, 2006
Just how accurate are weather forecasts? A Web site has taken an inventive approach to trying to answer that question, by poring over data from major providers. One conclusion: Where you live has a lot to do with reliability. Plus, a reader forum about the column.
'Drunkest' Rankings Need a Sobriety Test, September 7, 2006
A survey on the "drunkest" U.S. cities got widespread media attention, with some newspapers lamenting their towns' high rankings and others arguing they weren't high enough. But the survey, like so many "top cities" lists, used arbitrary methodology.
Rethinking the Solar System, August 25, 2006
How a group of astronomers decided that diorama you built in grade school has one too many planets. Plus, more fantasy statistics about productivity lost due to playing fantasy sports, and more.
Why Kids, Booze and Surveys Don't Mix, August 18, 2006
A recent survey made headlines by reporting that millions of kids are buying alcohol over the Internet. But the study used questionable methods to arrive at that jarring number. Plus, more on the Floyd Landis controversy.
More on the Landis Controversy, August 11, 2006
A second round of tests found a synthetic hormone in Tour de France winner Floyd Landis's urine, leading to a new round of speculation about how it got there. See what experts say about the theories.
Floyd Landis's Alcohol Defense, August 2, 2006
Floyd Landis is trying to hold on to his Tour de France title after a failed urine test -- and studies suggest a drunken night before the test could help clear his name. But do the numbers behind the alcohol defense add up?
Why Executives Speak Out, July 28, 2006
How do market researchers get busy executives to fill out surveys revealing what products and services they might like to buy? By luring them with rewards such as free magazine subscriptions or video rentals. Plus: How to count medical errors; and reader letters.
Putting a Number on Happiness, July 20, 2006
A recent study suggested residents of the island nation of Vanuatu are the happiest in the world, but the dubious methodology warrants a frown. Also, a PR campaign from Microsoft touting its online-ad system puts a questionable spin on some research, and more.
Hospital Stats Warrant a Second Opinion, June 29, 2006
A recent study said 122,300 lives were saved when hospitals instituted six steps aimed at reducing medical errors and improving care. But the methodology of the study could have allowed hospitals to hide disappointing results.
Lightning Stats Are Partly Cloudy, June 16, 2006
The National Weather Service is gearing up for a campaign aimed at educating the public about the risk of death and injury from lightning. But it may want to start with the government, which can't seem to agree on just how deadly lightning is.
Digging Into the Ethanol Debate, June 9, 2006
Critics of efforts to use ethanol-based fuel as an alternative to gasoline argue that more energy is used to create the fuels than is saved by using them. But some of the most prominent critics appear to have stacked the deck against ethanol in their research.
Watching the Pollsters, June 1, 2006
The flood of online polls and questionable survey data making its way into the mainstream press has professional pollsters looking for a better way to spot and quickly respond to faulty research and bad poll reporting.
The Box-Office Code, May 26, 2006
How do studios arrive at numbers for movies' box-office take? It's a mix of real-time reports and guesswork shaped by a host of cultural factors.
Should Mom's Pay Be $134,121?, May 11, 2006
Flowers, a card and dinner won't cut it for Mother's Day this year. If you really want to show your working mom you appreciate how much she has done for you in the past year, write her a check for $85,876. If she's a stay-at-home mom, the bill comes to $134,121. Also appeared in the print edition of The Wall Street Journal.
Measuring Chernobyl's Fallout, April 27, 2006
How many people died because of the Chernobyl nuclear-reactor explosion, which spewed radiation across northern Europe? Twenty years after the accident, the death toll remains in dispute.
Measuring the Child-Porn Trade, April 18, 2006
Is child pornography really a $20 billion industry, as was claimed in a recent congressional hearing (and repeated in media reports)? A hunt for the number's source uncovers more questions than answers. Plus, loaded survey questions influence an online-gambling poll. (The following month, NPR's "On the Media" interviewed me about the child-porn number, and other numbers hijinks; I appear on the segment that starts about 25 minutes into the audio file.)
Fuzzy Math on Illegal Immigration, April 5, 2006
Two decades haven't softened the debate over how to address the growing number of illegal immigrants to the U.S. -- or stopped those debating the issue from throwing around numbers arrived at through decidedly imperfect calculations.
When a Billion Isn't a Billion, March 30, 2006
How many zeros are in a billion? It depends on where you ask. Despite a few hundred years of pondering such questions, there is little global consensus on the way to name very large numbers.
Picking the Perfect NCAA Bracket, March 23, 2006
Companies are offering millions of dollars -- and a whole lot of pizza -- to the sports fan that fills out a perfect NCAA tournament bracket. I crunch the numbers and show why it's unlikely anyone will have to pay up.
A Bad Meth Stat Lingers On, March 11, 2006
Alarming stats on methamphetamine use published by the state of Tennessee linger on even after they're retracted. Plus, why most citations of the "law of large numbers" don't pass the bar.
Ambitious Homeless Count Fills A Void, but Has Its Limitations, March 3, 2006
In cities around the nation, volunteers are scouring the streets in an effort to get a more accurate count of the homeless population. The ambitious effort has its limitations, but could shed new light on an understudied group.
The Limits of Exit Polls, February 21, 2006
As polls closed in last month's landmark Palestinian elections, pollsters who interviewed voters announced -- incorrectly, it would turn out -- that the Fatah ruling party scored a narrow victory over Hamas. How did they get it wrong?
Sometimes in Polling, It's All in the Question, February 7, 2006
What does the public think about the Bush administration's wiretapping program? As a handful of recent polls show, it depends on how you ask the question.
The Results Are In, January 27, 2006
Nearly 300 Numbers Guy readers submitted answers to the quiz in my last column. Judging from the responses, many of you would make excellent copy editors.
Monitoring the Numbers in the News, January 20, 2006
Rich Holden, a former editor at The Wall Street Journal, has spent years monitoring bungled numbers in news reports -- and teaching journalists to be more vigilant. Take a quiz and see how you'd fare as an editor.
The Plight of the Family Farmer, January 12, 2006
Are hundreds of family farms really disappearing each week, as a television spot from Ben & Jerry's implies? Plus, a look at how well hurricane forecasters did in 2005, a guesstimate on how many people have lived in the history of the world, and more.
Evaluating Political Pundits, January 6, 2006
A new book takes a unique approach to evaluating the accuracy of political pundits by assigning numerical values to their forecasts. The conclusion: You're generally better off handing darts to a chimp than trusting the musings of a professional prognosticator.
The Numbers Behind the Transit Strike, December 21, 2005
New York's mayor has suggested the transit strike could cost the city $400 million a day, and other estimates put the figure even higher. But while the media love a big number, they often ignore its shaky foundation.
How to Split a Shared Cab Ride? Very Carefully, Say Economists, December 8, 2005
How do three friends split the bill for a shared taxi ride when they each get off at different stops? If they're economists, the solution might involve application of something called game theory, references to Nobel Prize-winning mathematicians, or even a look at the ancient guidelines of the Talmud.
Holiday Sales Numbers Don't Add Up, November 30, 2005
Just how strong a start is the holiday season off to? Analysts have offered mixed estimates, due in part to widely varying methods. One group uses online shopper surveys to gauge sales; another relies on video cameras to count mall shoppers. Still another monitors the weather.
Ranking of 'Smart' States Doesn't Get Passing Grade, November 11, 2005
A recent study naming Vermont the smartest state doesn't deserve a passing grade. Plus, the elusive link between scholarships and family dining, and your letters.
Lessons From the 1918 Flu, November 3, 2005
President Bush and others have pointed to the 1918 flu outbreak as a warning about the potential threat from avian flu. I take a closer look at the numbers behind that pandemic.
Sounding the Alarm With a Fuzzy Stat, October 27, 2005
A prominent national group is using a questionable stat to raise concerns about the U.S.'s scientific competitiveness with China and India. Plus, refining Katrina's toll, Google's estimate for the size of all the world's information, and more.
The Link Between Dinner and Drugs, October 7, 2005
Can eating dinner regularly with your kids keep them from using drugs? A recent study cited a correlation, but a closer look suggests it could be little more than a coincidence.
Crunching Baseball's Numbers, September 29, 2005
The manager of the Oakland A's loves to use statistics when discussing his team -- he said a recent losing streak was a 512-to-1 shot. A closer look at that number shows the difficulty in computing such stats, but one thing is clear: The A's have had their share of bad luck.
Estimates for Web Search Results Are Often Wildly Off the Mark, September 15, 2005
Journalists and others have long relied on simple Web searches to gauge how much interest surrounds a particular topic, counting the number of "hits" a search engine returns. But a recent study suggests those search-result estimates are greatly inflated.
Coming to Grips With a Grim Count, September 7, 2005
More than a week after Katrina devastated New Orleans, estimates of the human toll have been based on little more than sketchy math and wild guesses. (Plus, an interview with Barbara Butcher from New York City's medical examiner's office about how to count death tolls from disasters.)
Outsourcing Fears Help Inflate Some Numbers, August 26, 2005
For years, outsourcing foes have been talking about the number of engineers China graduates from its universitiesas many as 600,000 by some estimates, far more than the number of U.S. graduates. But the news articles and speeches that repeat the number rarely cite a source, and the official figure is far lower. (In May 2006, Gerald W. Bracey wrote an article in the Washington Post about these suspect numbers.)
Just How Big Is 600 Trillion?, August 12, 2005
An advertisement for a travel-industry software firm claims the company's technology is responsible for pricing 600 trillion trips every day. Is that possible? Plus, crunching the numbers behind expanded daylight savings time, and more.
Counting the Civilian Dead in Iraq, August 5, 2005
Several groups have tried to calculate the war in Iraq's toll on civilians by counting deaths from war, terrorism, crime and even illness. But the methods differ from study to study, and so do the results. Estimates range from the low tens of thousands to more than 100,000 deaths.
Avocados, Gorillas and Other Favorites, July 28, 2005
For my 25th Numbers Guy column, I revisit some favorite topics, including suspicious numbers from sponsored research, an update on Super Bowl avocado consumption, and the latest weigh-in for that gorilla in the room.
When It Comes to TV Stats, Viewer Discretion Is Advised, July 21, 2005
If media reports are to be believed, 5.5 billion around the worldall but one billion of the global populationtuned in to the recent Live 8 charity concerts. But that figure is the latest in a long line of big TV viewership numbers that don't add up.
A Simulation of Sept. 11 Shows Many Details Remain Unknown, July 14, 2005
How many people would have died in the World Trade Center if the Sept. 11 attacks had come later in the day, when the towers contained far more workers? Researchers are trying to answer that question, but some calculations show there is still a lot to learn about events on that day.
Pondering the Chances of a Nuclear Attack, July 7, 2005
How do you predict the likelihood of an event that has never happened before? That's what a group of arms experts had to do when coming up with odds for the chances of a nuclear attack by terrorists.
A Korean War Stat Lingers Long After It Was Corrected, June 23, 2005
A decade ago, the Department of Defense revised lower its official military death toll from the Korean War. So why does the higher number keep turning up? Plus, a study of the effects of race on employment gets oversimplified in news reports, and more.
How Much Is It Really Costing to Comply With Sarbanes-Oxley?, June 16, 2005
Some estimates on how much public companies are spending to comply with the Sarbanes-Oxley law are as questionable as the cooked financial books that led to the bill's passage.
In Hurricane Forecasting, Science Is Far From Exact, June 8, 2005
It's hurricane season, and that means experts are circulating their forecasts for how many of the violent storms we'll see this year. But a look at their track records shows mixed results, and the number of storms may not be the most useful stat to focus on.
Measuring the Impact of Blogs Requires More Than Counting, May 26, 2005
There are millions of blogs on the Web, but many of them are dormant or unread. I take a closer look at the statistics behind the growing medium.
Killings From 90 Years Ago Haunt Turkey in its EU Bid, May 16, 2005
The Ottomans' alleged genocide of Armenians 90 years ago has become a tense issue for modern-day Turkey, which is being pressured by some EU members to acknowledge the genocide and open up its archives. I examine a major point of contention: the number of Armenians who died.
Is the Conventional Wisdom Correct in Measuring Hip-Hop Audience?, May 5, 2005
It's an article of faith that 70% to 80% of those who buy rap music are white. But is this bit of music-industry conventional wisdom a good estimate? Plus: Neither side has a monopoly on questionable stats in Texas fight over gay foster parents.
Debate Over Gay Foster Parents Shines Light on a Dubious Stat, April 28, 2005
While Texas lawmakers consider banning gay couples from becoming foster parents, supporters of the measure are circulating some dubious stats. Among them: that children in same-sex foster homes are far more likely to be abused than those placed with heterosexual parents.
Retirement Is Long Overdue for Some Aging Statistics, April 22, 2005
A look at the use of an outdated stat shows why it's important for all numbers to have an expiration date. Plus, a cliched gorilla has really packed on the pounds in recent years, and more.
Fuzzy Math With Salary Calculators, April 14, 2005
Free online salary calculators aim to tell you how much you need to make if you want to move to another city and maintain your standard of living. Unfortunately, with most of them, you get what you pay for.
Small Change by U.S. News Leads to New Controversy in Rankings, April 7, 2005
A small change in the way U.S. News computed its most recent rankings of law schools could have a significant impact on minority admissions. It's a prime example of numbers run amok.
When Considering Surveys on Business, Follow the Money, March 31, 2005
There's no shortage of studies out there examining workplace behavior, warning of productivity drains and seeking to identify holes in office policies. But many of them are paid for by someone trying to sell you something.
Both Sides Massage the Data in the Debate Over ANWR, March 24, 2005
As the debate heats up over drilling in the Arctic National Wildlife Refuge, a closer look at the numbers being tossed around by both sides reveals that there's a lot of massaging going on.
Is the 'Truth' Campaign Being Truthful in Its Stats?, March 18, 2005
The antismoking "Truth" ads being run on TV are attention-grabbing, but are they really leading to a decline in teen smoking? The group behind the ads says it is responsible for 300,000 fewer young smokers.
A Look at the Globe, 45 Years Out, March 4, 2005
The U.N. says the world's population will hit 9.08 billion in 2050. But plenty of unknowns could have a big impact on that bold forecast, including the possibility of wars, the uncertainty of birth rates and the spread of AIDS.
From Waffle Irons to Eating Habits, Researcher NPD Has Your Number, February 25, 2005
Market researcher NPD Group claims to know all when it comes to Americans' dining and shopping habits. The firm puts out stats on everything from how many chicken dinners were purchased last year to waffle-iron sales. But can the numbers be trusted?
Heartbreak of Hockey Fans Aside, Canada's Economy Will Survive, February 18, 2005
Hockey fans may be devastated by the cancellation of the NHL season, but Canada's economy will survive. Predictions about the economic fallout from the lost season are likely overblown, and illustrate some of the murkiness of economic stats.
When It Comes to Donations, Polls Don't Tell the Whole Story, February 11, 2005
President Clinton said in a television interview that one-third of American households donated to tsunami-relief efforts. But that estimate was based on a phone poll, where social pressures likely inflated results. Plus, how big a hole you'd need to bury 3.6 million nickels.
Super Bowl Productivity Update, February 8, 2005
Numbers Guy readers who wrote in about their lost productivity due to the Super Bowl weren't far off from a widely quoted stat on the subject. Plus, a look at online auction fraud, and more.
Football Fans Likely Don't Know League's Most-Coveted Stats, February 4, 2005
NFL teams love stats, though many are never seen by the public. While analysts focus on yards gained and completed passes, teams use a dizzying array of metrics to track performance. Plus, avocado consumption during the Super Bowl, and more. Plus, readers' letters, and my responses.
A Survey Probes the Back Seats of Taxis, With Dubious Results, January 28, 2005
Some 63,135 cellphones were abandoned in the backseats of London taxis over the last six months, according to a quirky survey that made headlines recently. The precision of that number should be your first clue something's amiss.
Online Warnings Mean Well, but the Numbers Don't Add Up, January 21, 2005
An alarming stat is circulating on billboards and TV commercials around the country: One in five children has been sexually solicited online. While the motivation behind the campaign appears to be sound, the stat at its center is misleading. Plus, readers' letters, and my responses.
Just How Deadly Is Bird Flu? It Depends on Whom You Ask, January 13, 2005
A world-wide flu epidemic could kill 100 million people, or hardly anyone at all, depending on whom you ask. Why the disparity? It turns out projecting death counts from such a bug isn't just an inexact science, it's more like educated guesswork. Plus, readers' letters, and my responses.