what beating sudoku can teach you about product development

Sudoku is a great puzzle game. A three by three grid of boxes, each themselves a three by three grid of squares, the object of the game is to fill every row, column, and box such that the numbers 1 through 9 only appear once each. It is offered at different levels of difficulty. The lever by which the creator of the puzzle administers the difficulty is the number of numbers they give you to start, and the placement of those numbers*. Sudoku can be confusing and, at times, frustrating.

The thing with Sudoku is that sometimes you make progress quickly and others you're stunted for longer than you'd like to admit. Occasionally, you might even have to give up entirely. But the sense of fulfilment that comes from the completed grid is what keeps you coming back. This is how I often feel about product development.

Going further, I actually think that product development can be similar to Sudoku. And if that's the case — what can we learn from Sudoku about product development? My personal learning centres around the importance of iteration.

The way to approach a typical Sudoku puzzle is to run through the grid and identify the certain numbers (CNs), which are the numbers for which there is only one possible number it can be due to the given numbers by the creator. Once you fill in the first CN, you go through the grid again, and a new CN has been revealed by the addition of your previous number. You repeat this, with the process getting faster very slowly at first, and then quicker, until this satisfying cascade of CNs at the end that results in success.

The process for taking a product 0 to 1 has the same dynamic. Once you've built your earliest prototype, you take it out to users. You show them, you uncover the first round of improvements. You implement them. You take it out again, and these new improvements have uncovered more insights, and so on. Then, eventually, things start to click. You find Product-Market fit, and a similar satisfying cascade washes over you. It all starts to make sense.

Not only are the two similar in their virtues, they're also similar in their pitfalls. Any novice Sudoku enthusiast knows that blindly following your hunches and trying to fill in the grid without validating first will often end in disaster. The game is too complex and demands more respect than that. I think often we have more patience with Sudoku than we do with product building (there's a sentence I never thought i'd say). If we can bring some of that patience and an understanding that jumping ahead of the numbers can end badly, we'll be well served for developing great products.

———————————————————————————————

*Fun aside that I'm making sure I keep out of the scope of the article, but the mathematics behind Sudoku creation are *very* interesting. The curve that charts the difficulty of a sudoku puzzle for each number of 'initial numbers given' is not only non- linear, but shows a very narrow window for creating an engaging game. Sudoku puzzles can't really be made without around 17 given numbers, which is the theoretical minimum number of given numbers for a unique solution. All the difficulty in Sudoku exists between 17 and 30 given numbers, which is quite a narrow window. I've had a go at charting this below.

The narrowness of the window for an engaging game of sudoku is just another reminder that long-standing games that we might assume are successful because of their place in tradition and culture, and perhaps not by merit, are often actually examples of the Lindy Effect. The places these games hold as long-standing forms of entertainment come by virtue of nailing something we often didn't even know made something fun. It's why they've existed for so long in the first place. This article by Oxford University interviewing the creator of Jenga speaks to that and is a great bit of follow-up reading for anyone interested. Great things might start by accident, but the secret to their success is always explainable.