Alamo Drafthouse Slaughter Lane
Questions About the Restaurant Poll
RECEIVED Thu., May 14, 2009
Dear Editor and Virginia Wood,
I've been thinking of these questions off and on for some years now, and with the release of the May 15 issue, I've finally remembered to sit down and ask them.
As regards the release of the current edition of the 2009 Austin Chronicle Restaurant Poll, my questions are these. How many people voted in total? How many people voted for each winner and subsequent place? Did they receive 31 votes? Three hundred eleven votes? Three thousand one hundred and eleven votes? How many votes did they win by? Two? Twenty-two? Two hundred twenty-two? I ask these questions for a number of reasons.
First, all of the fine folks who win these polls love to show them off in your paper and in their establishments. Great. But how big of a win was it? Do half the people in Austin think they're No. 1? Or is it really just a few hundred who took the time to vote? And, though I don't know how you'd know, was it a real win or some degree of ballot-box stuffing by employees and friends?
Second, was the win a clear win or an exceptionally close win? On paper, a win looks like a win. Was it?
Third, no doubt many readers look to these poll results to determine what new-to-them places they may want to investigate (hence the desire by restaurants to have the bragging rights), so why doesn't the Chronicle ever print the actual vote totals so your readers have a little more clear idea of where they may choose to spend their hard-earned money?
I've lived in other communities with papers similar in format to the Chronicle and with similar annual polls. I've never seen polling results printed. What gives? Any chance you're afraid to risk advertising dollars from folks whose results don't show so well?
I've been thinking of these questions off and on for some years now, and with the release of the May 15 issue, I've finally remembered to sit down and ask them.
As regards the release of the current edition of the 2009 Austin Chronicle Restaurant Poll, my questions are these. How many people voted in total? How many people voted for each winner and subsequent place? Did they receive 31 votes? Three hundred eleven votes? Three thousand one hundred and eleven votes? How many votes did they win by? Two? Twenty-two? Two hundred twenty-two? I ask these questions for a number of reasons.
First, all of the fine folks who win these polls love to show them off in your paper and in their establishments. Great. But how big of a win was it? Do half the people in Austin think they're No. 1? Or is it really just a few hundred who took the time to vote? And, though I don't know how you'd know, was it a real win or some degree of ballot-box stuffing by employees and friends?
Second, was the win a clear win or an exceptionally close win? On paper, a win looks like a win. Was it?
Third, no doubt many readers look to these poll results to determine what new-to-them places they may want to investigate (hence the desire by restaurants to have the bragging rights), so why doesn't the Chronicle ever print the actual vote totals so your readers have a little more clear idea of where they may choose to spend their hard-earned money?
I've lived in other communities with papers similar in format to the Chronicle and with similar annual polls. I've never seen polling results printed. What gives? Any chance you're afraid to risk advertising dollars from folks whose results don't show so well?
Do tell,
Peter Remington
Peter Remington
[Food Editor Virginia Wood and special issues Editor Kate Messer respond: We do not share polling numbers. When you share polling-number data with the organizations that have an interest in winning and that wish to have an edge in "gathering" votes, any undergrad who has taken a statistics class could easily figure out how many votes would set them off with an advantage. We cannot control one vote per one person like in municipal and national elections, so this is the best policy. The information we share is this: We get thousands of votes in each of our three annual polls. A percentage of those are determined to be legitimate. We employ some statistical theory to determine and resolve some issues regarding suspected stuffing. Also, a different number of votes makes up each sampling group for each award. For example, someone who votes for Best Sushi may have many empty nonvotes in other categories. Statistically, only a very small sampling is needed for credibility. The overall number of votes in each of the three polls is more than comfortably above that number. The Chronicle has never written about a restaurant solely because they advertise with us and has never ignored a food establishment because they do not advertise with us. While many winners do use their poll wins (banners, certificates) as promotional tools and some advertise with this paper, many others do not. Would knowing how many votes Uchi (a nonadvertiser) received really enhance your enjoyment of its sushi? Would Hut's Hamburgers (a longtime advertiser) actually taste better if you knew how many people voted for it?]
