Faster, Smarter Way to Build in The Sims 3
We have important news to share about the future of modularity in our projects for The Sims 3 — and it comes straight from months of testing, observing the game’s behavior, and a lot of technical investigation.
Since I decided to bring the OldSunset project to a close and embark on this journey of building Tiny Worlds, my goal has been not only to create more optimized worlds where gameplay feels more meaningful, but also to use the “performance headroom” these worlds provide to push detailing even further.
Although my first modular project emerged as a way to inspire this concept in inZOI, as I had the opportunity to learn more about the modeling process, I was also able to recreate cities like Norge in The Sims 3. That experience made me realize that, with the right modeling skills and enough creativity, it’s possible to reproduce virtually any place in the world — fictional or real — inside the game.
That said, I also had to face a few dilemmas along the way. After exploring the limits of the TS3 engine a bit more deeply, I reached a clear conclusion:
modular doesn’t need to be complicated to be powerful.
What’s changing?
We’re adopting a new approach that makes modular construction:
More intuitive
More fluid
Lighter on the game
And, most importantly, more player-friendly
It has been genuinely rare to see people using my pieces, and I attribute part of that to the complexity of assembling buildings as if they were made of small LEGO pieces. While this approach offers immense freedom, that very same factor seems to discourage builders from adopting it, since it becomes just as labor-intensive.
What we discovered along the way
By working directly with the internal behavior of TS3, we realized that:
Fences follow a much more stable placement logic than decorative objects
They reduce the number of persistent objects on a lot and provide a faster, more predictable, and more consistent building flow
This allowed us to rethink how and why modular pieces exist, clearly separating:
what is structure
from what is detail
The result: less friction, fewer manual adjustments, and more time spent creating.
What this means for you
In practice, you can expect:
A more organized catalog
A system that’s easy to understand from the very first use
Large constructions completed much faster
Creative freedom without the feeling of “fighting the game”
This is a long-term change. A new modularity standard that respects TS3 for what it is — not for what we wish it were.
There are, of course, advantages and disadvantages, but this new standard doesn’t invalidate what we’ve done so far. On the contrary, it allows us to adopt an even more intuitive and faster construction workflow.
In the example above, it takes me roughly 25 seconds to fully fill the front section of the building. In theory, that’s already a relatively short time. In practice, the more details a building has, the longer it would take to finish.
Observation note: I don’t really know how to make GIFs properly, so if this looks too slow to you, please take into account the time I described in writing. I genuinely tried to keep the GIF in real time, but well… it seems it ended up slower than it should have been. Please consider the written information instead.
On the other hand, using this new approach, I can fill the exact same area in less than 1 second. In other words, a 99.998% time optimization.
Even though this approach may seem extremely logical — and might prompt questions like “How didn’t you think of this before?” — the answer is simpler than it sounds: I didn’t know how to make fences. 🤣
After dealing with some performance issues, such as a small lag whenever I exited Build Mode and entered Live Mode, I realized I needed to rethink my workflow in order to build something sustainable in the long term. Over the past month, I took three weeks to completely rebuild all of my modular pieces, optimizing them with clean geometry and eliminating unnecessary vertices. Today, most of these pieces have fewer polygons than the game’s low-LOD version of a lamp.
Even after such extensive optimization, there was still a small delay whenever I exited Build Mode. And no matter how short that loading time was, it genuinely bothered me. I began to notice that the game seemed to be recalculating every single piece — a problem that obviously doesn’t happen with fences.
The difference between using individual modular objects versus fences (combined with objects for corner connections) is a waiting time of 7 seconds vs. 0.38 milliseconds.
Of course, this means redoing part of the work I did last month: reallocating some modular pieces and turning them into fences, as well as changing the base object from a lamp to a pillar. It’s a significant amount of work, but I’ve already invested more than 20 hours just between yesterday and today to find the correct positioning for each piece and achieve perfect alignment. I also ran multiple stress and performance tests to validate my theories on how the game interprets and loads objects.
If there’s interest, I can share more of this process in the future. But for today, I’ve written enough — and now that the pieces are finally snapping correctly, it’s time to rest. Tomorrow will be a full day.
Once again, thank you for your support. I hope you enjoyed this new approach, and I’d genuinely love to hear what you think about all of this.
Your support is what allows me to spend so much time creating and making new discoveries in The Sims 3. So, thank you so much. Thank you for being part of this process and for supporting development that prioritizes clarity, performance, and thoughtful creation.
The best is still yet to come.