“Figures don’t lie, but liars often figure.” – Carol D. Wright

peimei

As a testament to how twisted my brain is, I LOVED statistics courses in college. I still do.  I had a 4.0 average (as I recall anyway), and immediately began applying it to everything around me.  Yes.  I’m that messed up.

So, why bring this up now?  Because statistical references play into everything around us.  From politics, to marketing, to work, to budgeting, to well, everything.  In this instance, because so many tech vendors use statistical claims to bolster their marketing charms, I decided to call them out.

Case 1 = “Fastest Growing ____!” claims

When someone says “such-and-such is the fastest growing __ on the planet!” what does that really mean?  Here’s what it means:

Divide the delta (value that denotes the change from state 1 to state 2 for the given time period) by the total number of state 1.  For example, going from 500 to 800 denotes a 62.5% change.  The rate aspect relates to the time period.  For example, if you said you were traveling at 80 MPH, but left off the H (hour) it wouldn’t mean very much.

In the same realm as velocity, a “fastest” claim denotes a velocity.  It implies “rate of change” or “relative change over a given time period”.

So, by itself, without a quantifier, it means very little.  In fact, it’s very often intentionally misleading.  That’s right.  Liars figure.

If a vendor says “Our product is the fastest growing on the market today!“, ask what the total counts (before/after) are.  If their product went from 1 license sold, to 2, that’s a 100% increase.  If their competition went from 100,000 to 150,000, that’s only a pitiful 66% increase.  Obviously their product is better than that pathetic 150,000-seat competitor, right?

Case 2 = Margin of Error

Another area that seems to get little attention (okay, zero attention) is the margin of error.  Often shortened to just “margin” or “MOE”.  This is the score that denotes uncertainty in the findings.  Put another way, the MOE denotes how far “off” the numbers can be, without violating the overall result.

This matters when the comparison between two or more items differs by less than the MOE.  In that case, it means the comparison is a wash.  That’s right.  The difference is so little that it should be considered meaningless (unless you’re aiming to prove indifferentiation).

For example: “Product A is favored by 52 to 48 over Product B” and MOE is 6.  That means it could be anywhere withing each being 6 units/votes/people/etc. higher or lower than the numbers stated. Pay attention to this when you see news or marketing (okay, typically the same thing) pitching something at you.

And now you know.

PS. Statistically, this article could be off by as much as 50% (MOE) 🙂

Advertisements

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s