Subscribe to DSC Newsletter

Reckless analytics: financial suicide or public disaster

Here we discuss two types of widespread analytic handicaps: individuals affected by poor analytics, committing financial suicide but impacting very few people, and government or corporate analytic failure negatively impacting millions of people.

Case #1: Public Disaster

Here are a few examples that illustrate the consequences of bad analytics made by government or corporate entities:

  • Walla Walla, a popular wine tasting destination in Washington, hoping to compete with Napa Valley, but closing major roads for hours every day in the middle of the high season, also closing local roads that lead to top attractions in the middle of the high season, issuing many tickets to tourists driving on its highways, issuing many parking tickets to tourists parking in town and spending $$$ locally (though there is no sign about parking rules - actually I told my wife this could be a fake ticket, it really looked like one, only $10, no reason why we were fined).
  • Washington DMV wrongly deciding how to allocate HOV lanes, pretty much getting it all wrong all the time and creating giant, permanent traffic jams on all highways around Seattle. I actually wondered how they can get it wrong 100% of the time. If their decisions were made purely randomly, they would be wrong only 50% of the time. I've lived in several states in US over the last 20 years, and WA is by far the worst, when it comes to highway management. Part of the problem is that they make decisions one day, and it stays permanent no matter how traffic patterns change over the next subsequent 30 years.
  • The city of Issaquah (where we live) paid $9,000 for a statistical study about building a roundabout close to a school where traffic is a nightmare. They hired a statistical consulting firm that ended up designing a roundabout based on July traffic, when the school is closed. Not only this was a terrible design of experiment (making inference based on just a few days of traffic analysis in July), but they could have asked us to do the analysis correctly in the first place for $3,000 only. It ended up costing hundreds of thousands of dollars to the community.
  • In our city, we have many traffic lights that turn solidly red for 200 seconds when you approach them (at 20 miles per hour), when there is no traffic. I'm sure the city has purchased very expensive analytic-based systems to manage traffic (it's a wealthy city) but have no clue how to make this system works. Again, it is an example where they get it all wrong all the time, when a perfectly random traffic light management system would work much better. The worse part is that they paid $$$ (via taxes) to have an advanced analytic solution to handle traffic lights. It is a very sad story for analytic people like me, because they were sold an expensive efficient system, but are not only unable to use it efficiently, but instead turned it into something worse than what we had before the advanced analytics era.

Case #2: Financial Suicide

These are examples of individuals (entrepreneurs) killing themselves through terrible analytics:

  • Many restaurants are doing a terrible job at optimizing their finances. We were at the Vine restaurant (pictured above), a very remote place 25 miles away from the closest little town, and accessible only by a dirt road. Beautiful mansion, fantastic food crafted by a fabulous unknown chef, and great atmosphere eating in a gazebo with a nice 95 degrees dry heat outside, at a price 50% below what it should be, for one of the best 7-course dinners we ever had in Washington. All in all, an hedonist paradise. Allan, the owner (who grows local orchards), a kind of modern American wild west cow-boy dressed in pink, was washing the dishes and serving tables. We learned that the mansion was abandoned for 12 years due to financial problems, before being restored and turned back into one of the great bread and breakfast in US. How long will this fabulous restaurant last? I think less than 12 months, unfortunately - not sustainable. For instance, their chef could be hired by a great NYC restaurant and offered a much higher salary. Also, none of the tables inside were occupied on a Thursday night, maybe due to poor marketing, being too far  away from civilization, and local highways being blocked for hours making customers arriving 2 or 3 hours late, forcing them too cancel reservation (happened to us the first time we tried to eat there) or stopped by the police (happened to us the second time we tried to eat there).
  • Again a restaurant example. On our way back from Walla Walla, we book a table on OpenTable at a local restaurant in Snoqualmie Ridge - a wealthy community for professional families, many working for Microsoft 8 miles away. When we arrived, we were told that they did not accept children and we were turned down (there were no such restrictions when we made our reservation). They only had one table occupied on a Friday night, when all their competitors were full booked. Why do such entrepreneurs make these stupid, financially suicidal decisions - turning down local families that could generate a steady income flow, not mentioning the bad reputation they get for their poor judgment? We don't know. We could understand that an upscale wine bar in NYC don't want kids and don't care because they have an upscale clientele of singles spending $$$ and who don't want kids around, but Snoqualmie, WA, is in no way comparable to NYC.

In the case of Walla Walla, we won't come back. In the case of the Snoqualmie Ridge restaurant, we ended up eating at home, and we won't come back. This represents several thousand of dollars that we want to spend but will probably be never spent (we'll spend more time vacationing and eating at home instead). So it does have a negative impact on the economy: people don't spend not because they don't want to spend, but because spending your money comes with such big and absurd problems, caused by analytically-challenged decision makers. I'm sure many executives managing corporate budgets think along the same lines.

Related articles

Views: 841

Comment

You need to be a member of AnalyticBridge to add comments!

Join AnalyticBridge

Comment by Vincent Granville on July 31, 2013 at 9:51pm

Another example of reckless analytics is the housing bubble. All the people who bought houses that they could not afford, and failed to predict that the bubble was about to burst (a level 1 prediction, that is a very easy one). You can say that the banks were the crooks, but no buyers were ever forced to buy a house at insane prices. Smart buyers have not been financially wiped out - indeed I made some good money in the real estate thanks to smart timing, between 2005 and 2012.

I bought a house in a bubble-free market in WA after selling my house in CA for $575k (we bought it for $270k four years earlier, it is now worth $230k). The WA mortgage agent was very unhappy that we did not buy a $1MM house in WA and just settled for a $600k home (now worth $650k). But they did a mistake when filing the papers: they applied an interest rate for a 15 year mortgage but issued a 30 year mortgage as requested. No way it was an error, this must have been an attempt to steal money from us. Just looking at the terms (without even using a pocket calculator), I politely said "there is no way this is a 5.85% interest rate". They changed all the paperwork and we came back to sign the corrected paperwork a few days later. Without me noticing the "mistake", these crooks would have stolen more than $100k from us over the course of the mortgage. I'm sure plenty of buyers have been robbed this way, are still being robbed now and don't even know about it, because they lack analytical skills.   

Comment by Vincent Granville on July 28, 2013 at 11:28pm
Hi Dominic - I'm sorry you expected to read an article where analytic ignorance is quantified, maybe something like 50 million Americans are severely analytically challenged, resulting in 200 billion dollars losses each year (I made up the numbers).

Yes my article describes local examples, but I am sure everyone can come up with 5 different examples every day. The disease is widespread, more prevalent and more costly than diabetes.
Comment by Dominic Roy on July 28, 2013 at 8:16pm

I was attracted to read this post because of the title. But I'm sad to say that I feel the analysis was very shallow and somewhat motivated by some frustration in your neighborhood. Your experience in a specific situation shouldn't be generalized, we need facts. I'm sorry, I don't want to make any lesson to you. You're far more competent than me, but I'm not convinced by this post.

On Data Science Central

© 2019   AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service