Wednesday, May 07, 2008

Work Boring Need Things to Think About

So my loyal readers who diligently read my drivel regardless of their care for the topic or any interest at all in me or anything I have to say; I have a real job. It involves sitting in an office, in a cubicle, doing a repetitive task for approximately 9 hours a day with around 30 minutes of lunch to break up the monotony. What I do and who I do it for are entirely unimportant and as such will not be published here, but what this has given me is an appreciation for a few things such as going to bed early, internet radio, and statutory holidays. Over all of these things though it has given me an appreciation for lunch.

This appreciation had led to an (over-) analysis of my lunch habits in order to optimize said habits. If it weren't for my scientific background and love of needless use of big math words this analysis might of been along the lines of a ranking or other ordering of the restaurants around my office, possibly in some sort of deliciousness per cost per unit distance metric.

Now one could do this and it would be needlessly mathematical, subjective, and most likely entertaining only to me while writing it and other people who are in the exact same situation as me in both locality and mentality (a few) who read this blog (none).

In my situation, I purchase a lot of lunches. This is mostly due to my own laziness for not packing my lunch because I certainly don't make enough money to justify the purchase of everything I eat downtown. I average roughly one brought-from-home leftover dinner style lunch every week. Also every second week my office does, 'pizza Friday' as part of the standard high-tech company fare of giving perks to employees to keep morale up. (I think I like my brother's company which on important dates, will bring in enough beer to run Granville Island Breweries OUT OF BEER.)

This leaves an average of 3.5 days per week in which I will purchase a lunch. For the sake of round numbers, we'll double this to 7 lunches in 2 weeks. The number disparity is because I only work 4 days a week (4 lunches - 1 lunch from home - 0.5 pizza Friday = 3.5).

Optimization of these 7 lunches is the key because I've ran out of entertaining internet. So there needs to be a ranking of restaurants, but because I'm cheap we need to scale it might their cost.

If we want to score (S) restaurants into a ranking, we should compare all their traits. Fundamentally there are only two units though, desirability (D) as a omni unit that encompasses everything about a restaurant BESIDES Cost (C). You can feel free to break D up into parts, but fundamentally there is a measure of how satisfied you are by your luncheon experience BEFORE the bill arrives (metaphorically speaking, where I usually eat, you pay first THEN eat, because I'm poor, so before the bill arrives/neglecting consideration of the cost of said luncheon/whatever.

Clearly this gives us a system for scoring/ranking restaurants using the handy formula, S = D/C

Given that desirability is a hand-wavy non-rigorous quality and most likely extremely challenging to quantify into a simple number, we have to confront of the challenge of establishing baseline scales. Arbitrarily defining a McDonald's combo as say, D[McD] == 1, and something else like, the Keg as,

D[Keg] == 30. We could then simply say, "I desire going to the Keg (on Thurlow) for lunch thirty times more as a lunch then McDonald's."

You can then consider then rank things in comparison to these two points, then use points to move them around, only these two are set in stone though, all others are merely in comparison to them.

Cost is easier since you can just use dollars as the unit of it and divide as appropriately. Although some scaling might be good. I think SUBTRACTING the minimum price (M) of a lunch you might expect, since it'll push the scoring up of places that are cheap significantly. While moderately good places that are expensive will fall. The more you subtract, the more you favor the more expensive restaurants. You can just neglect any scores that come out as negative as "below" your level of consideration. This updates our score formula to, S = D/(C - M)

Then you can use the scoring to choose which restaurant you should go to! But clearly you shouldn't just go to your highest scored restaurant on the list every day! That would probably be both unhealthy and (unless you're my brother) drop the desirability considerably! I know I couldn't eat Butter Chicken every day regardless of how delicious it is. Anything you eat repeatedly gets boring after awhile (again, unless you're my brother).

So using some math, you can translates scores into a frequency space representation of how often you should go to each restaurant! If you update the desirability of each eatery you visit every time, your scoring system will more and more accurately represent how often you'd 'like' going to a particular place.

So yah, evidence suggests that workers on salary or hourly wages with mostly unsupervised access to a computer with internet access will waste on average 5 minutes of every hour fooling around on the internets. I'm glad I could share mine with you.

Anyone have any ideas for things for me to think about that aren't discriminatory tesla coils? I mean, just things for me to consider at work. Men-in-Hats problems. Math problems. The feasibility of turning my room into a ballpen.