FanPost

Junk 07/02: Stat Geek Challenge

 

So, apparently Hedo is either a decent FA pickup or not worth his asking price. I've actually heard arguments for both sides of the coin, plus an opinion that it could be a smokescreen. Also, he was once seen eating pizza. It would be great if there was a place I could go to get 36 hastily-assembled takes on the situation rehashing redundant talking points, but that's obviously a pipe dream.

So, let's make the best of it and put our collective heads together on something I've literally been losing sleep over: a puzzling statistics problem on a project I've been working on.

The only way I know how to explain this is in an imaginary example, so bear with me:

Let's say I own a company, and the company has six main products. I have $600 I'd like to apply to advertising and instead of divvying it up equally I'd like to put more money towards products that are struggling than those that are not.

So I draw up a report for each product to see how each is doing. Unfortunately, I use an unusual reporting system that spits out numbers that are percentage-like but definitely aren't percentages. Here's a breakdown:

Product A: -1.9
Product B: 1.5
Product C: 4.0
Product D: 5.6
Product E: 6.2
Product F: 6.3

(negative numbers mean the sales fell short of the goal, positive numbers mean they exceeded it; a 20 would be a perfect score)

Obviously, Product A could use a lot of help while Products E and F are pretty well set. I can tell A needs a good amount more dough than any of the others, but how much more?

Well, this is kind of where things fall apart in my little accounting act. My instincts say that I should find the average for all products (here it would be 3.7), then calculate the distance of each product from that, trying to find parity between all six. Anything below the average gets its share of the $600 multiplied by its distance while anything above the average gets its share divided.

That works fine in theory, and even makes sense in a human-logic kind of way: Product C, for example, is only very slightly above the average, so its share would be right around the $100 average. The wheels come off for me, though, when I start attempting to multiply or divide by values less than one (Product C would be a +.3, which would have the opposite effect of what I intended) or accounting for the possibility of dividing by zero (which should be zero effect, or 1:1).

I kind of seize up at that point and try to imagine a negative number represented in a pie chart or try to wrap my head around dynamic ranges or whatever and then feel bad and decide I'm curious to see if someone's got a new take on Hedo. Then repeat.

So, any ideas where I'm going wrong? Or maybe there's just somewhere I've gotta go right that I'm just not seeing?

If you came here to talk about sneakers or balls or celebrities or all of the above, congrats on making it this far and you're officially absolved of all necessity to comment on this dilemma. If you're ridiculously smart with numbers and can see the fallacy in my efforts, though, please let me know - it'd really save me some grief!

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Blazer's Edge

You must be a member of Blazer's Edge to participate.

We have our own Community Guidelines at Blazer's Edge. You should read them.

Join Blazer's Edge

You must be a member of Blazer's Edge to participate.

We have our own Community Guidelines at Blazer's Edge. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9347_tracker