It was a Tuesday night, maybe nine-thirty, and I was sitting on the floor of my oldest's room with my laptop open and a half-eaten sleeve of crackers next to me because I'd skipped dinner again. We had just pushed an update to the game — a new area, a couple of new items, a mechanic my kid had been sketching in a notebook for three weeks — and the concurrent player count was sitting at eleven. Eleven real people, none of them us, none of them John.

That doesn't sound like much. But for context: our previous personal best was four, and two of those were my youngest on two separate devices to see if the spawn point worked. Eleven felt like a stadium.

It held for about seventy-two hours. Then it stopped, the way these things always stop — not with a crash, just a slow bleed back to zero that you keep refreshing hoping is a lag in the dashboard. It wasn't a lag. That's the whole story, really, except for the part where it almost wasn't.

How a Notebook Full of Sketches Became a Tuesday Night Emergency

We've been building this thing for a while. The first version went nowhere in a way that was almost instructive — we launched it, nothing happened, and we sat with that for a few weeks before deciding to try again instead of quitting. That part I'm actually proud of, even if the revenue column remains depressing.

The update that almost worked came out of a Saturday afternoon where my kid spread about twelve pages of notebook sketches across the kitchen table and walked me through a whole new biome they wanted to add. I'm a software engineer by trade, but Lua is not my natural habitat and Roblox Studio has a way of making me feel like I'm debugging someone else's dream. I said yes anyway, partly because the ideas were genuinely good and partly because I had nothing better to do that didn't involve the grass, which I was not mowing.

We spent three weekends on it. My kid handled the design logic and kept correcting me when I implemented things wrong, which was often. I handled the parts that required actual scripting and Googling "roblox studio tween service not firing" at eleven p.m. like a man with a plan. By the time we pushed it, we'd put in probably eighteen hours between us, spread across a month of evenings and one long Sunday where we ordered pizza and didn't talk to anyone else.

The Seventy-Two Hours the Dashboard Actually Moved

The first sign was a comment in the game's discussion thread from a kid named — I'm going to say Dylan, because that feels right — who said the new area was "actually fire" and asked when the next update was coming. I read it three times. My oldest read it over my shoulder and said, matter-of-factly:

"See? I told you the mushroom part was good."

They had, in fact, told me the mushroom part was good. I had said it seemed niche. I was wrong about the mushroom part.

Over the next two days the concurrent count bounced between eight and fourteen. Not viral, not even close, but alive in a way the game hadn't been before. A few players left ratings. One of them was negative — something about a respawn timer — and my kid took it seriously and fixed it in about forty minutes, which is faster than I fix bugs at my actual job. We were getting organic visits from somewhere, probably a Roblox discovery algorithm doing something we didn't ask it to do and couldn't reproduce. I know better than to count on an algorithm I don't understand. I counted on it anyway.

By Friday the count was back to two. Saturday, one. By the following Tuesday, the dashboard looked exactly like it had before the update, which is to say: quiet. The kind of quiet that's hard to distinguish from broken.

I went back and looked at what might have triggered the bump. Best I can figure, the update hit during a window when the algorithm was surfacing new content in our category, and we caught a small wave that had nothing to do with us specifically. This is the part that feels familiar from other things I've tried — you get traction, you can't identify the cause, and therefore you can't repeat it. The traction meant something. The meaning is just not actionable.

Eighteen Hours, One Cold Pizza, and a Respawn Timer Argument

In pure money terms: nothing in, nothing out. The game has made some Robux across its whole life, and I genuinely don't want to convert that to dollars because the number is small enough to be embarrassing and I've already written about what this thing actually earns, which is approximately the cost of one medium coffee if you're being generous about the exchange rate.

The real cost was time, and the time math is not flattering. Eighteen hours across a month of evenings is not nothing when you're also running a day job across time zones and occasionally doing a DoorDash shift when the grocery budget gets tight. There were two nights I stayed up past midnight scripting something my kid had asked for, and I was useless in standup the next morning. My team is used to me being quiet in standups, so this may not have registered, but I registered it.

I also spent probably four hours after the drop-off just staring at analytics trying to reverse-engineer the bump, which is time I will not get back and which produced no useful conclusions. That's the part where I was most visibly a fool — not the building, which was at least fun, but the post-mortem on a three-day spike in a free Roblox game, conducted alone at midnight with cold coffee. If someone had filmed that, it would not have looked like a guy with a strategy.

What We Kept, What the Algorithm Kept, and What My Kid Already Knows

The mushroom biome is still in the game. The respawn fix is still in the game. Dylan's comment is still in the thread, and my kid checks it occasionally in a way that I find both sweet and a little heartbreaking, because I know what it's like to check something and want it to have changed.

What I dropped was the idea that we could time an update to catch another wave. We had one conversation about it — me suggesting we try to push something on the same day of the week, same time, see if the algorithm liked it — and my kid looked at me with the particular patience children reserve for adults who are missing the point and said we should just build the next thing we actually wanted to build. Which is correct. I was trying to reverse-engineer luck, which is a thing I do across several different hustles and which has never once worked. It's the same instinct that makes people watch highlight reels and try to copy the frame without understanding what built it.

What stuck for real is harder to quantify, which is annoying to say because it sounds like I'm about to get inspirational, and I'm not. I just mean: my kid fixed a bug faster than I can fix bugs, and they did it because a stranger on the internet said something was wrong, and they took that seriously. That's a real thing that happened. It doesn't pay anything. It happened anyway.

The game is still up. The count today is zero, or maybe one — the dashboard has a delay and I've stopped trusting it. We've got a new area half-sketched in the notebook. I haven't opened Roblox Studio in two weeks because the day job has been loud and the evenings have been short. Probably we'll get back to it. Probably the next update will also spike for three days and then go quiet. I'm not sure that's a reason not to do it.

Note: Everyday Hustle Jeff is the written side of a digital series. These accounts are composite and dramatized. Any figures are what they were on the day, for one person, and are not a projection, a track record, or advice.