Here was Derek Smart's predictions for CitizenCon 2015:
http://dereksmart.com/2015/10/star-citizen-the-endgame/
- They are going to parade all the A-List actors – all of whom cost a ton of money – for SQ42. This despite the fact that the stand-alone game is reported to be almost fifteen months away from completion.
- They are going to show the opening cut-scene sequence for SQ42 with Gary Oldman giving his speech, someone flying and landing, NPCs waving etc.
- They are going to show Star Marine. Again. It’s still a mess – and nowhere ready for production release. Reports tell me that if it gets released before year end, it would be a miracle.
- They are going to show multi-crew. Again. Problem now is that, as of the last report I got, it doesn’t run smoothly at all. And so they are now running the demo on super computers with 16 cores, loads of memory etc. Again, not production release ready.
- There is nothing new being released to backers, other than another AC patch apparently.
Seems like he was pretty on the money to me.
Note how he he was telling people that Illfonic's Star Marine was DOA at a time when CIG themselves were saying it was
"weeks, not months" from release.
There's also the matter of CryEngine. In Derek's first blog he wrote:
Right from the start in 2012 when they said that they were using CryEngine3 as their baseline, I was skeptical. But if they kept the scope, and scene sizes manageable, I felt that it was totally doable.
Once the feature creep and increased scope started to unfold, I knew they were in trouble.
Remember what I said earlier that there is no game engine on the planet that would power the game I wanted to build, and that I’d have to build it? Yes, same thing here. That is precisely why all the top-tier developers build their game engines around the game they’re making. And even those who end up licensing middleware game engines do so based on the fact that they are making a game that fits within the framework of the engine they’re licensing. Nobody is going to license UE4 to develop a flight sim.
That CE3 engine is, first and foremost, a first-person engine. Second, it is geared toward small scale session-based games. Now imagine using a level-based engine, suited for first-person games, and trying to build an open persistent world with it. That’s like me trying to outfit my Tesla with the engine from a Prius. Bad things can and will happen.
This quote has really stood the test of time when you look at the last year of development. From CIG admitting that they've had to retool nearly half the engine, to the still incomplete "Item 2.0" system made in order to prevent the engine from having to render every object on the map, the completely inadequate player count per server for an MMO, and the endless array of glitches and hiccups that make Star Citizen a very pretty but very broken demo. In fact, nearly a year after Derek wrote that blog, Chris started sounding a lot like "Smart Speak":
http://scqa.info/?show=10FTC&episode=85&qid=4
And so essentially the core of CryEngine was never built to simulate this many items and objects, and we've had a long term refactor that we've been working on, the Item 2.0 that keeps on getting mentioned is one of the core ones … part of it, but we're basically refactoring the way the whole entity system updates; we're refactoring the fundamental …
So we've done a lot of top level changes on top of the CryNetwork but it's just not going to be good enough handle systems this big and this much data going back and forth. So we've long had a full rewrite of everything in CryNetwork on the table and we've been working on it and now with what happened, and what you could see on 2.4 and the framerates and everything it's just clear we're at the limit of what the current engine will handle without these new systems that we've been working on for quite a while coming online.
But let’s move on to more recent stuff. For instance, there’s the meat and potato’s prediction that
3.0 and SQ42 would not be coming out this year at a time when both had been projected for a 2016 release by CIG.
Then there’s the recent Chris Roberts interview where he states that
with the current level of funding, they would only be able to complete the single player portion of the game (but not SC). This means that Chris's projection of $65 million to complete both SQ42 and SC with all stated stretch goals was far, far off the mark. In his very first blog on the subject Derek wrote:
http://dereksmart.com/2015/07/interstellar-citizens/
There isn’t a single publisher or developer on this planet who could build this game as pitched, let alone for anything less than $150 million.
And here we are at around $150 million, still without the money necessary to complete the game.
For my money though (and I mean this literally), the most important "prediction" Derek has made is that if you tell CIG you will be informing the AG and the BBB about their conduct, that you will get the refund that you deserve without much fuss. That’s info that’s worth it’s weight in virtual spaceships many times over.