Enemies!
Even if most sucked, we did some cool shit! In usual ToG style there were grand ideas, but enemy ideas were a lot more realistic, nothing was too far out there.
We did
Junkbots and they rocked. They roam around and pick up and equip junk pieces - arms, heads, and legs. Washing machine legs vibrated them forward slowly making a rumbling sound, tire pile legs had them bouncing in a slightly uncontrolled way, and there were more. There was a propeller head which let them fly/jump around a bit. They had poison gas arms, rocket launchers, chainsaws, and more that I forget. With the right combination of garbage they were legit threats. I think they were Barkley 2 enemies at their best. With garbagious garbage they were garbage.
Cybergremlins were our first real enemies, I think. Their looks and behaviour were randomized. Small, fast, weak, and hit you with wrenches. Ended up being the 'cookie cutter' enemies that most others were modeled off, for better or worse.
They were also symptomatic of Barkley 2 combat issues. Just run backwards and shoot. In theory they were dangerous if you got cornered, but in practice with open ended maps that could never happen unless you messed up. When we tried to make them more dangerous, they became murder machines if they could get close in group, hit stunning you until death. We could never find a good balance, they kept alternating between lame ducks and MvC2 Magneto.
At some point a walking spider tech demo turned out awesome and became the
Crab-o-tron. They were Fucking Baller, would walk and jump around, with flamethrowers and also would litter the area with static fields to avoid. Legs were shields so had to hit from right angle. To this day they were the most fun Barkley 2 combat I have played.
I made a Crab-o-tron boss prototype, ten times bigger size and you would run between the legs and it would spawn minions. I'm not sure if anyone else tested it.
They taught us the lesson that enemies were best designed by just trying something out in game and seeing what works. Unfortunely we didn't really learn this lesson as a team and start working this way.
The opposite way - have an idea, write it down, create sprites, and then try to make it work in the game - led to garbage like
Weresnails. By now you should be able to spot their problem instantly by just watching that clip, where I am intentionally playing bad, and repeating "Open Ended World" ten times.
By the time we made
Catfish we had got better, but they and the way they were made still had all the above issues. We tried to make them work by casting nets to trap you in place. It helped a bit but all the core issues were still there. They saw a lot of polish, or in other words, time spent on something that wasn't great instead of talking about and fixing the game's issues. This was 2016/2017 and by now I was giving up hope we would ever really honestly discuss and fix combat.
Aliens were my attempt to make a faster enemy. They were strong, and faster than you when running straight, so for a first in B2 you couldn't just run away. I think they showed the way to good B2 combat. Unfortunately I could never give them enough polish to prove it to myself or others. It turns out it's really hard to focus and polish something when you're working free on late evenings an hour or two at a time.
Swamp Monsters and Kobolds also happened. That clip shows me dodging like a moron trying to make the clip looks interesting. In reality you would stand at a distance and calmly gun them down.
The swamp area had a nice mechanic to try to force combat into high gear: Mosquitoes. You got more mosquitoes surrounding you over time, fast, and would start losing health. Only getting close to a certain type of plant would make them go away. Some of those plants were Mimic monsters. It could have been really cool but was also hard to get to work well and fun in practice, because 1) 'poisonous floor' areas are rarely anyone's favourite areas and 2) swamp never really came together with fun combat because see above swamp enemies.
What could we have done differently to avoid this massive garbage void of time and work? Fucking Playtesting. We should have playtested everything from day 1. We should have had as a standard workflow to implement one thing, playtested it, and got feedback before even dreaming of doing something else. We should have playtested once a month. All our problems would have become instantly obvious. After two years when it became apparent that this was an issue, I brought this up thousands of times, to deaf ears, there was always some fucking reason why we couldn't playtest at that specific moment. Gun's breeding wasn't complete so the player couldn't see the whole experience. We needed affixes first. We needed more polish first. Everything needed to be connected first. Who cares? We need a good core gameplay before anything else matters. This is the one big learning I take away from this project into all future projects, and you should too. This was also a chicken and egg problem, we never playtested, meaning the game was never really playable by an outsider, meaning we could never playtest, meaning it was never... on and on. It was a huge effort to get demos ready for conventions, and while the game was really liked it also needed someone there to explain it (from what I hear - I never went to the conventions, sadly) When we came back from cons we had lots of feedback, but never really did anything with it combat wise.