(Something a little more lighthearted. Maybe.)
Every morning I “play” at least one Sudoku puzzle. I have a book of 3000 puzzles and they’re listed as “medium” to “hard” and I’ve graduated to “hard”. I can usually get one done in about 5-10 minutes and it kind of helps me wake up with my coffee and frame the day.
Yesterday I had to redo one twice, eventually erase it, give up, and move on to the next one in the book, promising myself to return today. This had not yet happened in my Sudoku history, and I wasted (?) probably 45 minutes yesterday trying to prove to myself I could do this puzzle. I had to complete the next one to “validate” that I still have “got it”, but it didn’t quite sit right.
This morning I finished the puzzle, but I had to do something I really hate doing: I had to guess. In most Sudoku, there comes a time where you are looking at 2 of the 9 slots in a line or a cube and you *know* that it there’s two numbers that fit into those two slots, and you know which numbers they could be (e.g., Slot A is a 5 or a 9, slot B is a 5 or a 9). This allows you to know that all of the other slots in the line are not those two numbers, but it’s not like they’re interchangeable: those two slots also impact their cubes and impact their perpendicular lines. Guessing, even educated guessing, is a gamble.
The gamble increases in scope when you make that guess early on: say, when your puzzle is only 20% complete instead of when it’s 80% complete, because the downstream impacts of your guess permeate through almost every other number choice. So, in a puzzle of 81 numbers if you’re 30 numbers in and you guess, then 50ish more number choices are impacted. Granted, the stakes are low: if you’re wrong, which you will discover with about 10 numbers to go, (or hopefully earlier), you grab your trusty eraser (may I recommend a Pentel Hi Polymer, I prefer mine to a Pink Pearl, my earlier go-to) and erase the crap out of the sudoku and start over. The only thing you’ve lost is time (and maybe a little pride).
We are, in my organization, coming out of RIFs, entering a planning season, and doing the logical thing: we’re looking around at what has to be done and with what resources (human and otherwise) we can do it. We’re also trying to predict the future: what will we need 3, 6, 12 months from now; where will we need to be in 2 years, etc. In some cases, we know pieces of the puzzle, and in some cases, we have to guess. We have pretty PowerPoint slides and impressive Excel pivot tables and neatly combed backlogs when we are done, but ultimately, there will be a time in this process where we have to say we think Slot A is a 5 and Slot B is a 9. Even in my relatively steady area of Fundamentals and Operations, there’s an aspect of predictability that is missing; while not nearly as bad as my Product Manager friends have it (their customer base is legit millions of people; my customer base is some <1000 engineers), the human component represents a blank slot on the Sudoku board and we have to put in a (hopefully very educated) guess as to what goes there. The price of erasure and starting over is and can be disruptive and is not brushed aside by a “never mind, we’ll get to it tomorrow”. It happens, and we absorb it, but it’s not terribly fun, either to explain or execute. In our case, this un-fun piece is mollified by an evolving process and an attentive release team that keeps us on track and has a system in place for when we identify that we indeed guessed wrong, and that Slot A was supposed to be a 9. It doesn’t take the sting completely out, but it does help.
And then we go on to the next puzzle.
Bobbie Conti 