[rss] [contact] [forum]

What's this all about?

I'm interested in and write about a wide variety of topics - economics, psychology, marketing, music, etc. I prefer writing long articles to short posts and don't update very often.

Recent Activity

Popular Pages




Decision Theory

Decision theory was always something that interested me, and it's something I used when figuring out what to buy for my home studio. Here are a few models to help confuse your further.

In simple terms, decision making involves figuring out the problem you're trying to solve, generating alternatives, and choosing between them. Each step has it's own challenges.

Problem Definition

It's often easier to focus on the effects of a problem, or get caught up in something you think might be a solution than to actually figure out what the problem is. I often make fun of this consumerist culture by making fun of the guy who goes out and buys a new, expensive guitar so he can learn to play... He could've learned on his old guitar, or could've spent that money on lessons, but he somehow thinks buying a guitar will make him want to play.

Maybe I'm biased, and I'm certainly not making judgments because I've done this myself and will do it again, but I think he's displacing his desire to make music with a desire to own musical instruments. We all do it... "if only I had ___ I could ___." However, we're not solving the real problem.

This involves stepping back and being like Elmira from Tiny Toons and asking yourself "why? why? why?" over again until you can't answer the question any more... My favorite business-friendly version of that question is "What is it you're trying to solve?" I've actually been able to shorten a lot of meetings by asking that question.

Another problem is conveying a situation to other people. In business we call this "gathering requirements." Most people will give a laundry list of steps. "Go to the store and buy me some milk and eggs." So you get to the store and there are 3 kinds of milk (whole, skim, and part skim), and no regular eggs, but egg beaters. What do you do? If, instead, you were given the instructions "I want to bake a cake, but I need milk and eggs. Can you go to the store to buy me some milk and eggs?" Once you got to the store, you could make an informed decision - whole milk and real eggs.

Without a complete set of (a) a description of the end state, and (b) a description of the current state of affairs, it will be difficult for anyone to think on their feet should complications arise. Whenever you're giving or getting instructions, be sure that both of these are clear.

Generating Alternatives

Generating alternatives can be easy and it can be hard. In academia the brainstorming/idea generation process is easy - you sit around and throw out ideas. Write them all down and see which ones stick, or get fleshed out. In the real world, brainstorming tends to work only over dinner for fun, or at design agencies.

When I was building my home studio, it was easy to find products that more or less fit my needs - there are entire magazines dedicated to bringing them to my attention and telling me how great each of them is. Countless thousands of dollars are spent doing this. Even the little 'underdog' has a group of people (sometimes called evangelists) who want me to know why it's better than the others.

In high-stress situations, however, alternatives become more difficult to generate.

Multiple Regression: Pros And Cons

Here's a scenario from when I was building my studio.

Okay, I'm looking for a studio monitor (speakers to everyone else) that's are cheap, say less than $500. Reasonably flat response with decent bass extension, and will allow me to create mixes that will translate well to other systems.

That's pretty much the typical request for information I see on the pro-audio boards though usually in the form "What's the best microphone under $200," and that's exactly what I was looking for when I bought my monitors.

This request acts as a filter - 1 no monitors over $500, 2 no monitors without spec's for frequency response and bass extension, and 3 they must allow me to make decent mixes that will sound good a wide range of stereos.

One way to make a decision with only the published spec's and reviews is to weight the pros and cons, and I did quite a bit of this... Mostly because I wanted to reinforce to myself that I was making good decisions, and because I couldn't actually afford everything yet, so weight pros and cons over and over was the next best thing. Sort of like a teenage girl writing her first name with her boyfriend's last name over and over again.

Product Price Spec's Notes
Event 20/20bas $850 35Hz - 20kHz, ±3dB (-2dB @ 38Hz), One of the of the most popular self-powered monitors in the world... Excellent published spec. More expensive than I wanted.
Behringer Truth B2031 $400 "Ultra Linear Frequency Response from 50hz to 21khz" Behringer has a worse reputation than Event. Published spec's are sketchy (ultra linear?). In my price range.


So let's say I narrowed down my choice to these two. I can create the proverbial two columned list of pros and cons for each speaker and write out what I think the pros and cons are.

If I want to get real fancy, I can assign a weight to each decision point like this:

This is called Franklin's Rule, or Multiple Regression. In a nutshell, it's a list of weighted pros and cons. It's plusses are that it includes lots of options so I know already what the strong and weak points are, on the downside it takes a lot of energy to do, and how do you know when to stop adding criteria? "Oh, and they have to be stylish."

A simplified version of this is Lexographic, which chooses the one criteria that is the most important and selects the item with the highest score in that criteria. For example, "I chose the Event 20/20bas because it had the best reputation."

A similar & well known is given the name Dawes Rule. This is the simple "let's make a list" decision method everyone knows. You know, "divide a piece of paper into two columns and write Option 1 on one side and Option 2 on the other and then write down all the pros and cons of each." Dawes Rule simply adds up all the pros (and cons) and picks the one with the highest score.

Heuristics: I'll Take It!

When I go on Amazon and read reviews of books, the first thing I do is sort the reviews by the "most helpful first." While the most helpful review might not be the best one to help me make my choice, many of the better reviews get voted to the top, so if I start with the most helpful and work my way down, I often get a good overview very quickly.

I found one site with reviews of products that allowed you to sort the products by the itmes with the most reviews, the items with the highest rating, etc. I found the "most reviews" sort was better than the highest rating sort because three people might give a product a 5 and it will have a perfect score, but 200 people might give a product an average of 4.7... I'd trust the 200 people more than the 3 people.

My decision heuristic was this: Find the product with a lot of reviews AND a high score. Say, in the top 10 most reviewed products, and with a score over 4.5. I figured that if a lot of people reviewed it AND rated it well it's probably a good choice.

This is the Well my broker is E. F. Hutton and E. F. Hutton says... decision making model. You choose one trustworthy source or criteria and choose the first one that comes along that matches it.

Heuristic: Of or relating to a usually speculative formulation serving as a guide in the investigation or solution of a problem: “The historian discovers the past by the judicious use of such a heuristic device as the ‘ideal type’” (from dictionary.com)

One sample Heuristic is "Take the Best." This means you find the one criteria that's worked for you the most times in the past, and use it to make your decisions. "I always trust the Zagat Guide." "The Gap always has jeans that fit me." "I read everything by this author."

A number of my friends use this method, especially those that hate shopping. "This one matches my criteria, I'll take it." In fact one guy I used to know used "I'll take it" as a sort of catch phrase. No comparison shopping, just take the first good one that comes along.

So I'd take the first item that had both a lot of reviews & a decent rating without weighing other factors. They would choose the first nice looking restaurant, or pair of jeans that were of a brand they always chose and were in their size.

Recognition Primed Decision Making: Experience Counts

Being under pressure to make a decision causes stress. It doesn't matter if that pressure is being in a burning building or a deadline that's 3 months away, the pressure to make a decision and act causes us to think differently.

In situations like this, we don't have time to generate alternatives and weigh them against each other, though if we're inexperienced that may be something we fall back on. The more experience you gain, however, the more clearly you can see what needs to be done and can do it.

Joe Friday from the "new" TV series Dragnet is like this. In every episode there's a scene where he's briefing his staff on what's going on and giving orders. He knows exactly what he's doing. Then comes a moment where he's stuck and the case introduces something new and unfamiliar. At that moment, he asks his staff for suggestions and alternatives.

As you gain experience, you build a mental image, a roadmap of the ideal way things are done, as well as of any pitfalls you've encountered before. When you encounter a new situation, you compare it to your mental image and without realizing it you assess the situation. This mental model is so ingrained in you, that you can often sense something's wrong without actually knowing why.

So how does someone using this model choose from various possible paths? Experience. They come up with a possible solution, one that their experience leads them to believe will work. They quickly run through it in their minds to look for flaws. If they don't find any, they act. If the do find one, they abandon the idea and look for a new solution that works. They don't, however, compare alternatives.

Of course, nobody chooses to use the Recognition Primed Decision Making Model (RPD), this model was formed by watching people in high-stress situations - fire commanders, military leaders, nurses in an emergency room - and interviewing them afterwards.

You can teach Multiple Regression & Heuristics as decision making methods almost as mathemtical models, how do you teach RPD? Since RPD is based on experience, you figure out how to give people experience, or enhance their mental models. One way is to put them in situations that are similar to the ones they'll be in. Simulations, basically. Another is to tell stories. Rather than give a laundry list of what to look for, by telling a story you can give your experience to others.

So How Do I Choose?

I know people who epitomize each of the above ways of making decisions, and at times I've used them all. Some people are naturally slow to make decisions and love to weight options. Others naturally make snap decisions, and still others make decisions based on experience.

I suspect that there's a scale from Multiple Regression through Heuristics and RPD. That we all use each in different situations. Those of us who have the luxury of time and enjoy making comparisons or want to know we've made the right choice in an area where we have little experience will use Multiple Regression. Then when we've gained some experience and knowledge, we'll use Heuristics. Finally, when we've become veterans in a given situation, we can sense when it's right or wrong and use RPD.

  1. Multiple Regression
    More information than experience, takes a lot of time, for when you don't know what you're doing and want to know you've made the right decision.
  2. Heuristics
    Some experience, faster, for when you kind of know what you're doing and are willing to take a chance.
  3. RPD
    Lots of experience, happens subconsciously, for when you know exactly what you're doing.

Interestingly, each of these can be used against us as well. Oddly enough, this traces the history of advertising. Product manufacturers can give us a lot of facts & figures to help us feel like we're making the right decision "by the numbers" (Multiple Regression). They can also tell us that someone we know and trust (Michael Jordan, 4 out of 5 Dentists, etc.) uses the product (Heuristics). Finally, they know we're on to them, so they have commercials where they make fun of themselves, letting us know "we know you know commercials are stupid, well so do we," throwing off our fine tuned sense of  sophistication (RPD).

Tossing a Coin

I've suspected for a long time that most of us make decisions on a gut level. We're drawn, for whatever reason, to a particular situation, product, person, etc. for highly personal reasons. Then, once we've made our decision, we use logic to back it up, and not the other way around. So even while I'm weighing the numbers, I may design the scale to tip one way or the other. This is sort of like tossing a coin to make a decision. If it's the option I want I go along with it, if it's the option I don't want, I realize it and choose the other one.

However, there are times when we need to choose between two unequal but foreign choices. I hope this guide helps you do it.

Anyone interested in further reading is encouraged to check out these books:

I wrote some more about decision theory in this article on The Apprentice "Scoop Dreams" episode.

page first created on Sunday, April 06, 2003

© Mark Wieczorek