unity3d mistake YOU don't want to make

Don't let your unity3d project fail because you made the same mistake.

[RobLang]

My top tip: don't upgrade your unity version just because it's come out. Wait until you reach the end of a feature, make a source control branch (or back up), install the new unity version, upgrade your project and see what's broken. Unity upgrades nearly always bring breaking changes. If you're near the end of your project, you're unlikely to see much benefit from the new version, so it's cool to hang tight on an older version of it means your game is going to get completed.
(Unless you know for certain there is something pivotal you need)

Great video, thanks Jason, keep them coming!

[aleksp]

Thanks for these tips, I'm new to unity and sometimes I really feel stuck and here you are coming from the heavens

[zat]

I really like that last tip about prioritizing the creation of the core game loop. I have started a few projects, thinking ahead. Wanting to add this feature and that feature, this mechanic, this asset. But nowhere along the way are you able to press play and even properly experience what I want the game to be. However, my next project I didn’t even start with coding. I wanted to build a maze puzzle game where you manipulate blocks to open up paths to the end. I prototyped a level with wooden tiles with a few obstacles I wanted and played it like a board game. I brought it around with me and, after getting positive feedback on the engagement, I started building it with Unity. I started first with a maze and a player who could move around it. Then, I added the basic block that was a foundation for the blocks, one you could push around. Just having two blocks in my way, I completed the loop by resetting the level when you get to the end. It already felt fun to play around in. I think that’s what “finishing the loop” taught me the most. You can have an idea, that sounds neat in thought. But may once actually played. The scariest part about putting too much time aside for expansions in the core loop before making it is finding out all that work leads to a game that’s not good. It took me a week (in my spare time) to make the other obstacles, and it’s taking me more time to make the UI and menus. Imagining doing all that first prior to having a playable level and then finding out it wasn’t gonna work sounds awful and disheartening. Not to mention, hard to sit through and work on. I’m not the biggest fan of UI making, if I was doing that and still didn’t have a game to play, it would be hard to stay motivated. Maybe once I’m done with this one, I’ll pull out those old projects and finish up their loops quick. I may end up finishing those too.

[InexperiencedDeveloper]

My brother based his entire first project on a single goat asset he downloaded 😂

[HoboKnight]

#7 speaks to me. In highschool I spent weeks on level editors for games I never finished. More recently I spent a couple months on an AI system and never actually got the rest of the game really working. Proud to say that I have learned from past mistakes, and also learned to better scope my projects, and am closing in on completing my first non-jam game after less than 2 months. Pretty much just polish at this point, but I definitely now understand that old game dev saying that the last 10% of the project is 90% of the work.

[erikdewhurst]

Yeah, man. This was validating. I recently used your suggestion to use a separate sandbox project for testing new assets. I felt free to fully play around with the asset and didn't fear breaking the final project.

[michaelmcveigh]

importing assets into a temp project first is really great advice. I actually think I might use this strategy when creating new prefabs too. Build it in a temp project importing the assets I need and then exporting to my main project. Thanks Jason

[BrettBernier]

Started with a medieval character with great animations and the ability to hot-swap armors and clothes.

Unfortunately I couldn't match the art style myself and the artist moved on to greener pastures.

Was stuck with this beautiful set of characters in a poly world.

Still working on the project though haha

[jonathanlorenz-indiedev]

Once I found a spider with procedural animations. When I saw it climb on everything, in my head the game was 75% completed. You're welcome!

[timotheusschuchner]

I have been buying a lot of stylized assets and focusing what I am building around those as well as making a big world without making the fun parts first. I needed to hear all of this

[TegridyMadeGames]

2: yes I'm very guilty of this but I think if you choose a smaller asset with simple code that you can understand theres no problem with it. My most popular game was made with a 5$ wave shooter that i modified (and fixed) but the code was simple enough for almost anyone to understand.
Of course making one from scratch is always better and just pulling a few scripts and models always works best.

[SentinelDevelopment]

Great video, as always. Keep up the great work Jason. You really are by far the best game dev youtuber.
Maybe make the number transition thing a little shorter the next time... but apart from that: great video. Thank you!

Oh and that "Dont open"-Box is tickling my brain

[krissloo]

Apparantly it's only me who never created a game based on assets :) although some assets inspired me to make certain games which I never made eventually

[InfallibleCode]

Great video, Jason. Thanks for the tips!

[cptant]

For 6 - "Polishing systems before you have a gameplay loop." I think that there is also another side to this. Sometimes things just don't feel right until you polish them.

[onurpekdemir]

Thanks so much Jason! Your advice is gold as usual ... Would love to hear your opinions on how exactly to split and separate core systems of a basic game (e.g. some arcade style project and how to set up and where to hook up systems for a simple UI, a main menu, some settings, load/save systems, main gameplay loop, etc). I tend to end up with monolithic blocks of code. Some general advice about when and how to use things like namespaces, partial classes, events over if-statements or general thoughts about organizing source code (e.g. where and how to declare fields for best performance) would be very much appreciated as well. Keep up the good work! You're one of my favorite YouTube sources for a long time now (although I could not check out every single video). Best wishes!

[reptilianaliengamedev]

Synty Studios, Ork Framework, Game Kit Controller, Gaia, Horror Kit Fps, Opsive Character Controller. Pretty much just start from stratch when I hit a wall with all these.

[supertenchoo]

am a newbie this is my second year learning unity and waste a lot of time learning and re-learning now i realize some time i should try to build a simple game on my own coz start making on your own game is the best way of learning

[ogulcanzorlu]

I've done all of these. Thanks for sharing Jason, great tips.

[TheGrooseIsLoose]

A trivial mistake to make that can waste time and possibly break things: if you upgrade a package or install a new package, especially something that interacts with other things like the Input System, and then something isn’t working, don’t panic and start upgrading other things or rapidly changing things to try and fix it. Stop, restart the editor, and then if it still doesn’t work (but it probably will work), take your time figuring out what is broken and what needs to be fixed.