Decision Making with Rankings

Ever have a tough time making a decision? Well, you’re not alone.
At one point or another, we all have tough decisions to make. Some of us struggle with potentially making the wrong decision. Some of us fear we’ll regret whatever decision we make. And still others get frozen up and never make a decision at all.
Myself? I’m the type of person that does way too much research into something and then takes all that data and applies a very unscientific method: the gut check. I figure I’ve done all the research and applied as much logic as I can, and now it’s time to just pick something based on how I “feel.”
I did this when I bought my new phone a few months ago and M and I did this when we finally bought our new place.
But there’s another way of doing it that I’ve been playing with recently, all thanks to Malcolm Gladwell’s latest piece in the New Yorker, What College Rankings Really Tell Us. Gladwell focuses most of his attention on how universities get ranked by US News & World Report and how subjective and potentially misleading these rankings are.
Especially when so many people use them as a guide to figuring out which schools are the “best.”
My Decision
It got me thinking about a decision I’m currently trying to make.
I want a new netbook and I’m trying to make up my mind on which of these three choices to settle on:
- Keep the one I have now
- Buy a Mac Air
- Buy HP’s DM1Z
I’ll spare you all the pros and cons of each of these, but suffice to say that the first choice is the cheapest (no spending), the Mac Air is the most expensive (by far), and the third one feels like a nice middle ground.
I’ve done extensive research and I know what I want to do, but I figured I would run this decision through a ranking system and see what it shows. Honestly, I was simply looking for validation for my gut decision to buy the HP DM1Z.
Here’s what my experiment with rankings told me:
I should buy the HP DM1Z.
The first thing I did was rank each one of these machines on a one to three scale on a series of seven qualities (Price, Portability, Design, Lifetime Use, Cool Factor, Power, Battery). One being the best and three being the worst. Then I aggregated all the numbers and Excel spit out what I already knew: the Mac Air and the DM1Z were the winners.
The problem was, they were in a dead tie.
Gladwell’s article has these great moments where a small tweak to the weighting of the rankings completely change the end result. I was looking for some of these unexpected results, so I started doing some weighting.
Which qualities do I think are most important? By giving price a higher value (35%) vs design (5%), I got new results:

HP’s machine leaped to the front of the pack and the Mac Air dropped. Giving price a heavier weighting did the Mac Air in, especially since I dropped the value of design, where it clearly had an edge in my rankings (it’s by far the coolest machine out there).
I decided to simplify things a little and cut down the things I ranked to four (Price, Design, Lifetime Use, and Battery) and see if that would do anything dramatic. Without weighting these, this is what I got:

Nothing new here. What if I weighted just these four qualities?

Now the Mac Air is really getting penalized for the price and it can’t make up the ground anywhere else. Why? Because I just don’t care that much about the areas where the Mac Air stands out (even though it’s way cooler than the other two). The design is impressive, for sure, but not impressive enough for me to overcome the massive price gap between the machines.
You’re talking to a guy who ate the same sandwich for years to save money.
Finally, in an effort to artificially create a “Gladwell moment,” I weighted design 50% and price 10% (the Bizarro Me) and here’s what I got:

Sorry Mac Air, the best I could do for you—even by trying to manipulate the results—was a tie.
This is an easy call.
P.S. I created a spreadsheet that I’m tweaking that I’m thinking of sharing. It allows you to type in three things you’re trying to decide on and then gives you a place to rank and weight them to tweak your results. Anyone out there interested in this or should I stop spending time on it?
Image by GDS Infographics


