When to implement online multiplayer

So I've been poking at a lot of basics for a game I've had in mind. Intended to be multiplayer in the long run, a platformer with cameras set to individual players (around 4, maybe). Individual player-chosen upgrades and statistics stored for the round that affects all players. The question is, when should I implement this functionality? It's a major part of this project overall, but it seems like a daunting task to sort out. In fact, it's getting to me so much that it's been turning me away lately...

So, is it better to get that out of the way as soon as it's sensible or get to it later in development and focus on other aspects for a while?

Have you gone through the multiplayer tutorials? It is worth spending some time to familiarize yourself with it to have a basis to know what sort of work you will need to put in.

The multiplayer plugin may seem intimidating at first, but it is actually rather well done and very clean to use provided you actually go through the tutorials and understand what each event is for.

This will help you judge if you want to proceed with a multiplayer game, and it will let you design with multiplayer architecture in mind.

On the other hand, I always like to prototype my base game first to make sure it's something I want to keep working on. Even if you had to redo some of the events later for multiplayer, it will be much easier the second time around, and you'll be sure your base game works as intended first. It is much easier to diagnose bugs and problems if you focus on one thing at a time!

Have you gone through the multiplayer tutorials? It is worth spending some time to familiarize yourself with it to have a basis to know what sort of work you will need to put in.

The multiplayer plugin may seem intimidating at first, but it is actually rather well done and very clean to use provided you actually go through the tutorials and understand what each event is for.

This will help you judge if you want to proceed with a multiplayer game, and it will let you design with multiplayer architecture in mind.

On the other hand, I always like to prototype my base game first to make sure it's something I want to keep working on. Even if you had to redo some of the events later for multiplayer, it will be much easier the second time around, and you'll be sure your base game works as intended first. It is much easier to diagnose bugs and problems if you focus on one thing at a time!

Ah, this is a really good way of putting things! Thank you so much!

I am in the same boat for a project I have in mind and oosyrag makes some good points.

I haven't had enough time/energy/motivation/clear mind in one sitting to break down all the parts in the multiplayer tutorials/examples. It is high on my to do list though as I know I will want to use it eventually evin if not for the project I have in mind, so for me I know it is 'worth' doing.

If you make the game without the multiplayer functionality first you will most likely have to start from scratch when creating the multiplayer version. But I may make a single player version first anyway to focus on the game play elements and making the game fun. I'm sure making a bunch of changes to optimize game play will be significantly more difficult with all the multiplayer elements involved.

Good luck and I will let you know any tips if I get it figured out!

I hesitate to give a shortcut, because I really, really feel you read through all the multiplayer tutorials, but here is the main design consideration:

All game logic and mechanics should be handled by the HOST. All PEERS generally only send their inputs to the HOST. The HOST will act based on these inputs received from the PEERS, then SYNC the state of the HOST game/objects to all the PEERS, so that they all have the latest state of the game.

The other main stumbling block a lot of people have is how to get connected and the concept of SIGNALLING SERVER, GAMES, ROOMS, and who is a HOST/PEER. But this really is generally very straightforward and the same for most games, so if you follow the tutorial it should be very clear. I think people have trouble with it because they didn't really read the tutorial to begin with.

Regarding starting from scratch - that isn't true! We spend the most time when we run into a mechanics problem that doesn't work right away as we imagine. Once you've figured it out, you won't need to figure it out a second time. Also, you can copy/paste events. When your base game works, all you need to troubleshoot is the sending of messages and communications. It is much more frustrating when you have to debug multiplayer communications at the same time as your game mechanics.