@kuligs2 @JusTiCe8 You both are right.
Bots are probably the area that needs the most work right now. Implementing bots and the single-player mode was a much bigger undertaking than I anticipated. The game was originally designed for multiplayer only, meaning everything had to be designed with networking in mind. Single-player is different. So basically, I had to create an "extra game" for Super Score, which took a lot of energy to get done quickly. So in that sense, yes, bots and single-player are a completely new addition to the game that started recently.
JusTiCe8 ID Software back in the day as well as Valve have gigantic talented and dedicated people, the two examples you give are among the most successful games back in the day.
And just to add, they were also well funded, and those titles were also considered AAA for their time. So, being a solo dev with a bootstrapped shoestring budget, one shouldn't underestimate the hard work, talent, and money that went into getting the results of titles such as Half Life and Quake 3. Just because those games were done a long time ago doesn't mean the topic got any easier. We might have better tools today and more help from an engine, but the concepts to learn and the problems to solve are still as complex today as they were then. It's like math, learning calculus takes as much effort today as it did before. We just got more access to information and tools that help us along the way, but math itself hasn't changed.
That said, there are probably concepts I can learn from those titles shared over the years.
kuligs2 lso im not gonna sugarcoat anything. if it sucks im gonna say, it sucks.
As long as it's constructive, I welcome it.
kuligs2 Last i heard he was using AI neural nets to train bots.
This is something I have put aside for now (might revisit it later). Showed a lot of promise in the small tests I did. But when I started to scale it, this method started to break down and was very hard to control. It's probably mostly because I didn't know the technology well enough to produce good results at a larger scale, rather than because it's a bad technology. I will keep reading and experimenting more on the topic. For now, the bots are created in a more traditional way, with a state machine and nav mesh.