Author: Brendan

I’ve played Connections like four times now, which of course you can understand means that I am ready to issue a wise verdict from the top of Game Design Mountain. The verdict is: this game’s interface and its incentives are at odds, which irritates me, but there are simple changes that could make it more fun and less frustrating without changing its difficulty.

I became aware of the game to begin with because part of the NYT Games team’s marketing strategy is working. Finishing a given day’s puzzle yields one of those little block patterns of color-square emoji, the ones that look like 🟩🟨🟦🟪 instead of 🟩🟨⬜️⬜️⬜️. It’s enough like Wordle to make one curious, but different enough to be distinguishable even if you have reduced color vision, and it takes good advantage of their ownership of Wordle’s IP and its most potent promotional idea. It’s also a big bet on the long-term appeal and viability of the game, because you only get so many uses of the shareable grid before it loses its brand value. But actually playing Connections has given me a new appreciation for the elegance of Wordle’s quality-of-life support, because Connections does not support life.

In one abstract rendering, both games could be reduced to “solve a puzzle in X educated guesses from a fixed domain of discourse.” Per Wyna Liu, the game’s editor and credited creator, the placement of the initial squares in its grid is deliberate, a choice designed to lead you down false paths before you sort out the right ones. The editors clearly enjoy coming up with grid items that fit more than one potential answer, and in my experience it’s easy to come up with multiple groups that are conceptually closer than the ones they chose (DOG and HEEL belong with JERK and SNAKE instead of FETCH and SIT? Really?).

That means that the actual game pitch is “solve for all four answers at once, and don’t get greedy by jumping to conclusions!” Which, fine, whatever, there’s nothing wrong with that in isolation. Liu herself, in the brief interview above, expresses support for the “pen and paper method” of working on solutions. Which is again fine, if we were talking about a game that the NYT printed on paper. But it’s not! It’s a web app! If the editor believes that the right approach is to sort out all four groups and check them through before submitting your answer, why doesn’t the user interface allow you to do so?

I’m not an interaction designer, but even I can come up with a simple solution here: make the tiles draggable, and allow the player to move them all into a solution grid before hitting submit. Instead, the only ways to move the tiles the developers have offered are a random shuffle button or a jump to a conclusion. The tiles shift themselves to the top available row when you make a successful guess, which is telling the player “go ahead and take your shot, I’ll reward you by reducing the complexity of the board if you do.” But that’s a contradiction of the editorial strategy laid out above. If you do fall for the temptation of the UI, your wrong guess might yield a cue like “one away!” that is of very limited utility, and that then vanishes, leaving only a record of your punishment. The effect is more like a taunt of “skill issue lol” than an encouragement. That’s reinforced by sharing-based marketing gimmick and the way many players use the pen-and-paper attack to solve the puzzle with no record of wrong guesses, so any mistakes at all come across as failure rather than tactical work.

This is what I mean about Wordle’s comparative elegance. Wordle gives you more guesses, more information per guess, and a persistent record of those guesses in both the solution rows and its keyboard. That means you can walk away from the game and retain your context when you come back to it later. Perfect solutions are possible only through a rare stroke of luck (or cheating), so if you’re competing against your in-laws on a group chat, it’s easy to evoke a sense of mingled competition and camaraderie. Even if sometimes everyone gets grumpy about HOMER or __GHT, there’s a real difference in the play experience between “ah, that was my fault” and “oh fuck off, you smug jerk/dog/heel/snake.”