Many mods often either try too hard to be innovateive and become cumbersome or not try hard enough and end up being unoriginal. I won't pretend to be an expert (being as this is my first mod project) but I will opine on the subject.
The first thing you need to do is study. Look at other mods, even at professional games, not for ideas necessarily but for what they did right and what they did wrong. For example I'll cite Half Life 2. I loved so much about that game (part of the reason why I'm going to the trouble of modding it), but I saw pros and cons. For one, many sections of the game felt like excess in places. Entanglement was a particuarly annoying chapter for me. Even though a major story point is made midway through the chapter, I felt the chapter itself (particuarly one huge gunfight where you must set up three turrets) was often overly difficult. However, I looked around in the level and began to think strategically on my next playthroughs and got better as I replayed HL2. (I've gone through the game on hard more than five times. I know somewhat what I'm talking about.) Valve's inner theory is telling a story but keeping it firmly within the player's control. So we really don't begin to learn all about the combine until Black Mesa East when Eli explains the newspaper clippings on the wall. The beauty of that sequence was that it was *completely optional.* I love it when a story can get the gist of it across to a casual player, yet still have huge amounts of story in the backweb for the more serious crowd (since I've been on both sides of the fence it's even more apparent to me.). Excellent examples of this point are also in Episode One's Kleiner cast, which again is completely optional and pops up at several points so the player can make the choice themselves to listen or not.
When I first brainstormed "5," I first thought of the AI in HL2 as compared to a game like F.E.A.R. Fear's AI is excellent- they dive for cover, they can go everywhere the player goes (climbing ladders, jumping through windows) and even some places that the player can't (climbing under pipes), and doing more intelligent things like moving things around to make their own cover (see the 18 minute "Project Origin" trailer to get a hands on look at this idea.). HL2's AI often felt mindless in places (rappeling in front of a water boat just to be run down is funny once, but after awhile it feels stupid, not to mention some of the more fortiuitously placed explosives in some setpieces).
Another gripe I had with both HL2 and F.E.A.R. is that sometimes the experience felt overly linear. MINERVA: Metastasis's opening level for example feels extremely non-linear. (And once you open the service doors to the center of the island, it can be approached from multiple entery points, earning non-linear points from me.) Parts of HL2's "Ravenholm" chapter were similarly non linear, because you had a nice set of rooftops and streets and back alleys to explore and shoot zombies in.
So taking notes from the above games, I attempted to craft a mod. My first attempt was a never-to-be-released attempt at nonlinearity that I tenatively called "Wasted."
The idea was that the player had been tasked with taking a small team of rebels aboard a combine transport over the Sahara desert and retrieve a specified cargo- the HEV Mark V. Ultimately everything went to hell, the player accidentally managed to strap himself in the suit, most of his team is killed, and he manages to port out of the transport as it begins to crash (the very beginning of the game). Under the direction of his Neural Interface (NI) named "Delator" (Roman for "informant"), he was to scout a huge desert locale for pieces of the transport, rip as much data as he could from these scattered terminals, avoid roving patrols of Combine soldiers looking for him, and attempt to track down any survivors from his crew. I ultimately abandoned the mod when the maps hit a huge technical hurdle and the map data was corrupted. In any case, "Wasted" died a quiet drunken death (A little humor!). Out of the ashes of Wasted came "5." Originally called "The Hunt" (name changed because of a mod by that name already existed), it has grown into the in progress mod you see on this blog.
Moral of the lesson: Revisiting old ideas with fresh eyes can give you some excellent refinements you might not have seen back in the day. Remember that hindsight is 20/20, and if you didn't release an old project than feel free to rip yourself off. (Just like I did!)
Progress ticker: Section 1 in progress. Naturally I wouldn't want to spoil what I mean by that.
