judgment (when statistics get in the way of a good decision)


“Statistics are no substitute for judgment.”

Henry Clay


ah. “When statistics get in the way of a good decision.” Let me get this out of the way upfront. I like numbers. I have an Economics undergraduate and accounting and statistics  accounted for several of my <of the few> good grades in college.

I like that if you weave your way through numbers they can tell you things that can inspire the ‘real’ thought.

I like the fact that numbers can sway an “I think” based opinion to a “here is what I know” based opinion.

That said. I purposefully used Henry Clay (so think maybe 1800 as to date of the quote) so that some contemporary statistical wonk doesn’t come out of the woodwork saying something along the lines of “statistics have only evolved in the last 20 years” or something crazy like that. “We have never had better data to make decisions from than today!” <with the exclamation mark> is a statement that was as true in 1800 as it is in 2000’s and as it was in 100 BC.

This is an eternal issue.

People have looked at statistics since the time good ole Adam started calculating how many apples fell out of the tree to figure out how often he was gonna get laid. Henry Clay just had the luck to be quoted on it.

So before I begin my rant let me say, yes, I get decision-making is a cognitive process where the outcome is a choice between alternatives. And that numbers can play a role. I also get that people have different preferences as to how to approach decision making and that there will always be a varying degree between thinking and feeling and numbers and experiential. And I do believe all decisions, at least the worthwhile ones, have to incorporate some sense of logical decision-making. Logic in that we seek to exclude <or marginalize> emotions <as well as personal biases> and try to use only rational methods <perhaps even mathematical/statistical tools> with the intent to isolate what is typically called the decision utility.

I get all that.

Oh. And, by the way, I hope no one tries to dump the whole your view is driven by whether you are “left brain/right brain” mumbo jumbo on me because science has already proven that is an urban myth. There is no right brain left brain.

And. There is no “numbers are facts” crap.

Yeah. On that last one.


“Torture numbers, and they’ll confess to anything.

Gregg Easterbrook


Sure. Numbers don’t lie. But they also don’t tell you what to do.

Now, in saying that, let me suggest why I believe this statistical ‘torturing numbers’ issue has been an issue for eternity.

The thought: many people don’t want to make decisions. Okay. Maybe people just get nervous with accountability and use statistics to make the decision; not inform a decision.

Why? There are boatloads of reasons, but suffice it to say that without using numbers you are getting paid (or at least judged) not just on decision-making skills, but on your judgment skills. That means accountability is solely on you (the person).

Think about that. But also think about this (as you get judged). The following is an explanation on decision making using statistics:

Decision Making Under Uncertainty: Statistical Decision Theory

I’d like to start today’s lecture with a reminder about something I said a long time ago when we finished our survey of population viability analysis. Population viability analysis is best seen not as a way of garnering precise predictions about the fate of a population but as a way of ensuring that all relevant life-history variables have been considered, that they have been considered efficiently, and that we have a reasonable sense of the trajectory that the population is likely to follow if current trends continue. It provides a way of structuring our thinking about the problem. That’s precisely the way I think we should regard the approach to decision making that I’m about to describe.

One of the most difficult tasks facing conservation biologists, as I have emphasized repeatedly, is that decisions must often, perhaps usually, be made in the face of woefully inadequate data.

<ba bla blaaaaaaaaaaaaaaaaaaaaaa>bla bla bla little

So. From that incredibly dry mind numbing analysis of statistical decision theory they dropped this little bombshell in toward the end of the ‘how to use statistics’:

There is the recognition from statistics that there are two types of errors we can make in evaluating an hypothesis:

–          We may say that something is happening when it isn’t (Type I), or

–          We may say that something isn’t happening when it is (Type II).

Say what?


“The capacity of the human mind for swallowing nonsense and spewing it forth in violent and repressive action has never yet been plumbed.”

Robert Heinlein

(I wish I had written this in response to the statistical theory thingamajig)

Ok. What that means (to those of you solely dependent upon statistics). You may use the statistics to prove something is or isn’t happening and it may not be happening or it is happening <anyone now wonder why statisticians are avoided?>.

Anyway. I will back off the ‘accountability through numbers folk’ for a second. Trying to give statistical-using people the benefit of the doubt, let’s think that maybe when you are stressed out, frankly, any option seems pretty good — especially one which seems factual (numbers seem more factual to people despite the fact that once you start combining them they become less factual). I imagine it is like someone dying of thirst and drinking from whatever looks like the safest pool.

Uh oh. But some pools are poisoned.

And, unfortunately about this stress theory of mine, when the adrenaline from the stress wears out, you realized that the statistics you leaned on for the decision YOU made — were all bullshit (or someone points out they were bullshit when you actually invest some energy trying to explain them later).

And you are screwed. Screwed because of statistics (although people will inevitably try the “but the numbers told us what to do”).

Ok. Here is the good news (relatively speaking). You can do something about the stress decision making leaning on numbers thing.

Most people, given enough experience, become aware that stress can do a number on your decision-making skills.

How do I know? Well. Of course someone has done a study on it. Scientists have some statistical based thinking about exactly how stress screws up your ability to make decisions. According to ScienceDaily, psychologists Mara Mather and Nichole R. Lighthall (who completed a review of the literature on stress and decisions) they found that, even though you’d think being stressed would turn people into pessimists and therefore more careful, stress actually makes us focus too much on the upside of our decisions. Says Mather, “Stress seems to help people learn from positive feedback and impairs their learning from negative feedback.”

Uh oh. That sucks <maybe you cannot do something about the bad stress decision making thing>.


I guess my point in bringing up the study is that maybe under stress it is easier to grab on to statistics to make a decision <all the while thinking positive thoughts> and therefore avoided the judgment call on your own.

Uh oh (again).

Look. I was wrong. You can do something about this judgment thing. Judgment isn’t easy, but at some point you are accountable <or you should be> and hiding behind statistics just won’t hack it.

As Yoda would say “the answers are within you.” The key to making a smart decision is giving yourself the time to gather all the information you need <and, yes, that can include statistics> and move forward with whatever proactive thinking method approach you have some confidence in and make a decision.

A daunting decision doesn’t have to put you in an analysis paralysis death grip.

Use a logical decision-making method to help you evaluate your choices and pull the trigger.

And make a decision.

And not let statistics make the decision for you.

Here is the net on statistics: It helps us formalize and categorize our thinking to make sure that we have considered all relevant possibilities.

Quantitative analysis should be viewed as explorations of possibilities; not hard predictions.

I believe being able to use numbers, and statistics, to explore possibilities is truly a skill <or an art>.

Not everyone can do it. Ok. Well. That’s not true. Anyone can do it, it’s just that not everyone can do it well.

Knowing what to do with the numbers is an art.

In fact, just to circle back to the main topic of this post, let’s call it, well, judgment.

Yeah. Judging numbers. This means learning how to weigh the importance of one number versus another as well as learning which numbers are unimportant. To be honest, not many people have mastered that art.

But. That doesn’t mean everyone should get bogged down in statistics and numbers because if you do, yup, you can torture any decision you want out of numbers.

And, frankly, you are lying to yourself if you believe that is a decision. That is simply being a coward (in the decision making world). You have deferred decision to ‘numbers.’ And inevitably you are deferring accountability.

Sound harsh?

As harsh as this?

“I notice increasing reluctance on the part of marketing executives to use judgment; they are coming to rely too much on research, and they use it as a drunkard uses a lamp post for support, rather than for illumination.”

David Ogilvy


Sound like truth? Yup. Suffice it to say everyone should use numbers and statistics to inform decisions, but, inevitably, all good decisions demand some level of judgement. Ponder.

, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,
Written by Bruce