Stats Made Easy

Practical Tools for Effective Experimentation

Sunday, September 30, 2007

Statistics so simple the only a child can do it

My youngest daughter, a senior in high school, asked me to help her with homework the other night: “Dad do you know anything about Z-scores?” :)

I am always happy to help with math or stats. However, it often proves challenging to not just get the correct answer, but also to follow the latest educational methodology.

My colleague Tryg passed along this ‘number-hummer’ by mathematician/satirist Tom Lehrer:

Hooray for new math,
New-hoo-hoo-math,
It won't do you a bit of good to review math.
It's so simple,
So very simple,
That only a child can do it!


If you can spare about 4 minutes, view this clever math/music video of Lehrer singing this song, including a verse on base 8 (“just like base 10…if you are missing two fingers”).

After many years of seeing my good deeds for helping with math homework not go unpunished, my oldest daughter admitted once that I had explained something “way better than the teacher or the book.” :)

As the British gamesman Nubar Gulbenkian said “The worse you play, the better you remember the occasional good shot.” This works well for me in golf and also math/stats – fields where I frequently fare poorly, but every so often, hit the sweet spot.

Tuesday, September 25, 2007

Fantasy football – addicting to a sports fan who loves stats


Let me say from the start that I am working on this blog late in the evening after a good day’s work at Stat-Ease. However, I must confess that I’ve been hooked on fantasy football since being introduced to it about 25 years ago – at least a decade before it became popular throughout the USA. Back then one would eagerly await the Monday morning sports pages for the Sunday stats. Now fantasy-football players can get instantly updated via websites such as that maintained by my son Hank. Unfortunately for their employers, web-surfers seeking scoring stats do some of this on work time. The executive search firm of Challenger, Gray & Christmas (CG&C) estimates that addiction to fantasy sports could cost the nation's businesses $36.7 million daily. The CG&C calculation stems from research indicating that 14 million people play fantasy football and that they typically spend 10 minutes of every workday managing their team. That seems conservative to me (not that I ever took a look at Hank’s website when I get to work first thing Monday and checked how my team did over the weekend).

PS. While searching the internet on this topic, I came across a promising site offering the “world's most comprehensive football statistics.” However, the pictures looked funny to me – a round ball rather then the prolate spheroidal shape that I’ve grown to love for being so unpredictable when wobbling through the air and bouncing on turf. It turns out this website is run by the British Association of Football Statisticians. I’d say they are far sicker in their fascination of sports stats than any Americans!

Thursday, September 13, 2007

Assessing the risk versus benefit of wearing seat belts


While lamenting the recent bridge collapse in Minneapolis, my barber admitted to being so phobic about careening into rivers (we live along the banks of the Saint Croix) that he routinely unbuckles his seat belt so as not to be left hanging as his car slowly sinks below the water’s surface. The fear is that his body weight will keep the locking mechanism in place. My theory is that without the belt restraint any occupants of the car will be bashed about and lose consciousness, which could be prove fatal when going underwater. These differing views come down to my assessment of risk versus benefit relative to the odds figured by my barber.

This morning I heard a report on the local TV news that two teens, cousins, died in a two car collision where none of the occupants wore seat belts. The local sheriff told the reporter that in his opinion, no one would have been killed had they belted up. However, a relative of these unfortunates said that when they were children a family member had perished in a rollover in which the deciding factor was being trapped by a seat belt, thus they eschewed these restraining devices, thinking them unsafe. The irony of this assessment of risk versus benefits, proven wrong by this tragic event (or were they just unlucky?), is heart-rending.

Here is a statistic that tells me it's best to buckle up: More than four-fifths of people in Minnesota wear seat belts, but about half of all people killed in car accidents went unrestrained. Think about it.

My awareness about preventing injury from car crashes is heightened at the moment due to me having just completed a defensive driving course. Minnesota law requires insurers to provide a 10 percent discount to anyone doing this at least once every three years. I’ve gone through the course twice now and found it helpful each time – getting good reminders of things I already knew and learning about a few changes in what the experts advise for certain situations. Odds are that anyone who drives regularly (I am averaging 20,000 miles per year), will be involved in an accident about every 10 years, according to MetLife Auto – my insurer who sponsored the course on defensive driving. Being a fan of statistics, I do have one beef with them – this quote puts me off:

“By learning to drive more defensively, you can reduce your chance of becoming the next statistic.”

I would think one would want to become a statistic because statistics are a good thing! However, I am certain that this attitude is held be a very small minority of the population, most of whom would much rather risk their lives than be subjected to statistics.

Anyways, be careful out there. Watch out for the other people. Put on your seat belt and drive defensively. That’s my assessment of risks versus benefits and I am sticking to it.

PS. This photo pictures the remains of what was once one of my favorite cars – a ’96 Ford Contour that I passed along to one of my daughters. She got hit by a Cadillac Escalade. Both drivers wore their belts and got protection from air bags, so although they were ambulanced to the local hospital, neither one suffered any apparent injury and walked out under their own power.

Friday, September 07, 2007

Jokes about scientists and statisticians who are powerfully driven to repeat bad things

Physicist Randall Munroe posted this insightful cartoon showing how the behavior of a scientist differs from normal when confronted with shocking consequences of a controllable behavior. It reminded me of the anecdote about a research lab burning down accidentally. Afterward a physicist and chemical engineer (my profession) discuss the mass transfer of fuels, exothermic reactions and the like. Meanwhile a statistician stands there quietly listening. Finally the other two ask "What do you think?" The statisticians says that all of this technical talk matters little, action should be taken immediately. What is the recommendation? Burn down another lab! (Knowing how important sample size can be for the power of statistics, I suspect that this kept going until the entire complex went up in flames.)