Activity Report

November 05, 2013 On luck and condition, or in other words, statistics

Have you ever brought a friend to a restaurant that you highly recommended - “You have to try it, it’s the best!” - but ended up having to make excuses when the meal didn’t turn out as good as you expected? Well, its not faulty taste buds that are to blame for your bitter experience, but rather that you didn’t employ statistical thinking.

It’s a fact of life that chefs have their good days and bad days. Your health condition is also not constant. And as ice cream sales go up when outdoor temperature exceeds 27℃, climate and weather conditions affect your sense of taste. Considering these things, is it possible to properly judge a restaurant by dining there only once?

As we all know, things have their ups and downs. There are good and bad times for everything. Sometimes the wind blows in the direction we want to go, and sometimes it blows against us. Usain Bolt doesn’t always run 100m in 9.58 seconds. Everything has its averages. It’s when certain conditions happen to come together by chance that records are broken and set.

Statistical thinking, to put it very simply, is a way of thinking that factors in these fluctuations in order to draw the most reliable conclusions.

Going back to the case of the restaurant, let’s say you found another one that you thought was pretty good. You again want to bring your friend there to impress him. If you are to employ statistical thinking to avoid an experience like before, there are mainly two ways you can go about it. One is to dine at this restaurant repeatedly, and study the taste of the cuisine on both its good and bad days. If you find that on good days it’s absolutely fantastic, and even on the bad days it’s acceptable, you can assume that it’s pretty safe to bring your friend
The other way to go about it is to try to recreate the conditions of the time when you had that really fantastic meal there. For example, you’d want to go during the same season and on the same day of the week, at the same time of the day, and sit in the same seats, have the same menu, etc… By doing so, you have a much better chance of avoiding one of the restaurant’s bad days.

This kind of thinking is not limited to judging restaurants, of course.
Let’s say there was a project that went very well at work. What can you glean from it to make your next project another success? What can you do to keep the last project from becoming seen as a fickle success, just a matter of luck? Like with the restaurant example, there are mainly two ways to employ statistical thinking. One is to gather the results of many successful projects in the past and analyze the common factors among them that could have contributed to their success, and in the current project, try to incorporate those factors that are applicable. The other way is to completely copy the previous project; but of course unlike restaurants, there is no point in doing the same exact project twice, so that wouldn’t be feasible. Therefore, the realistic way to go forward is the first way, in which you extract the factors for success from numerous past examples.

Statistics is not just about looking over past data. The discipline is about extracting the vital essence of data and using it to make a reasonable forecast of the future. If performance is fluctuating, statistics can show what the main factors of the fluctuation are. If data is lacking, it can show by how much. Statistics can show what the chances are of having another wonderful dinner at a restaurant that you’ve only been to once, and it can calculate what the conditions are to have the best meal there.

At the KAITEKI Institute, we are currently trying to incorporate statistical thinking in our studies in the areas of healthcare and organic chemistry (and not restaurant selection, although we do highly recommend the KAITEKI CAFE). One real example is analyzing past cases of successful health management in order to develop a system for effective health care services, which are also personalized according to individuals’ needs. Another example is extracting the factors for success in the development of technology for designing safe and effective pharmaceuticals.

Statistical thinking is a kind of “futurism that utilizes long history.” By employing this approach, we hope to deliver more products and services that can make a difference for the better in people’s lives.