It may seem a little strange doing a postmortem for Swarm without yet doing the Haven one, but it's up on Kong now so it felt right to give it a pimp shot in the arm, plus it should be a lot shorter for me to write as I'm just going to cover the specifics in this game rather than the whole thing.
What went right:
1) The huge advantage we had over Haven was the insane amount of feedback we had banked when making this one. Filtering out the noise with feedback is always tricky, but with so much and so many detailed ideas it was a lot easier. Because of that Swarm is a lot tighter in many respects than Haven. The movement is better, the indicator showing the direction of damage, just so many little almost unnoticeable things got tweaked that even though we finished this only a couple of months after Haven, it makes a lot of things in the first game feel dated to me.
2) The AI co-op is the most obvious stand out feature. I'm not sure how this came about, I mean it's an obvious thing to do, but I remember Lux and I talking at the start of the project and agreeing that we had to make the game stand up on it's on, it wasn't meant to be Haven lite, but a good game in of itself.
From memory I just spent a day seeing how it would play out, if it worked and worked as well as we would have liked then sweet, if not then we'd just drop it and write that day off. More by luck than judgement it did play so much better that it set the mood for the rest of the game. I'd say having Jameson as your wing man in this game makes it what it is as much as the aliens.
One last thing to tie a bow on this one, having the different colour schemes for the main sprites made a huge difference, something we really should have done in Haven.
3) The new tilesets / locations worked so well. We managed to re-use a fair few assets without hopefully feeling like we'd just lifted them wholesale ( I think we only added one new visual effect though, the smoke flares on the planet side levels. In our defence, we did make a lot of one shot visuals effects for Haven ).
From that basis Lux really came up with some beautiful stuff, the sewer level is great ( Unfortunately a bit of a CPU hog ) and the planet side levels used the terrain splatting which we developed for Knight's Quest at last. The out doors levels are just stunning, we're going back to visit that planet again in O2.
4) The game flow, from pressing Play to actually playing the first level, was tweaked a lot thanks to feedback from Miniclip. Normally amends just 'cause a groan in all coders ( And these did too to be honest ) but they removed a lot of clicks and helped you get into the game much quicker. Amends which actually improve a game ? Anyone who does just adver-games will think that impossible. It did have one downside though, but lets save that for the next section.
5) I really like the "pop-ups". Popcap nailed giving awards to players, and I've always been a fan of that, and the COD titles have just taken it and put it in an arcade / shooting context, which I really like, so we embraced that a lot more this game, and I'm really pleased with it. We're always flashing up some update ( The mode in the recent Halo games whose name escapes me right now also inspired me with this a lot ).
It helps ground the game I think, shows we're not playing a survival horror, but a shoot'em up with terror aspects.
What went wrong:
1) Time. We just ran out of time and budget to do all the things we wanted. We had hoped to add a new baddie type ( A flying bug type thing. I had a vision of them being like a swarm of gnats ) and I wanted to add the sentry weapon ( Which is something I think everyone wants ). I think the AI just ate into the time to do those things and it was a worth while trade off, as it brought more to the game than anything else we could have added.
I can't remember how much we over-ran the deadline by, 16 days or so, not a great deal, but there are always lose edges which need tidying up and adding major things like a new baddie would just have blown the deadline completely. As soon as you're past that deadline you're working for free and you can only do that for so long no matter how much you love the project.
2) After the whole Owlmen story running throughout Haven there was a lack of context in this. We did have a briefing in there, which we removed to improve the flow, but we had no story at all, no reason for this to be happening. I assume people who played the first game got it, but that's not really valid as it's meant to be a title which can stand up on it's own. I was toying with having info-cards really hidden in levels which would progress the story, but they would only have been relevant to players of Haven as we didn't have enough scope to explain the whole thing.
I regret that we didn't explain Swarm mode more. In my head it was always like a TV show being shown back on earth, Lee and Jameson are basically dead men walking at this point, there's no escape, so they'd go out fighting and all the money they earned from the bug hunt would go back to their families on earth ( These aren't stupidly brave men, they're just two guys caught up in a shit storm which they have to tackle head on ). Even at one stage I thought of having crowd cheers if you did something really cool, like you were watching this in a bar back on earth on a TV, but without any sort of context it would just have been mental.
3) Some of the levels aren't great, level 2 is too hard as it's so small and restrictive. Also we had to re-use a [ Tweaked ] level from the swarm mode in Haven as again we'd run out of time. Level design doesn't come easy to me, it's something you need to iterate over to get just right, and with that clock ticking I didn't really have the time.
Plus making an interesting level which is basically just a loop is tricky, with no clear objective, no going from point A>B, it's hard to make it interesting, and I failed on a couple of levels, with the medical lab being the stand out mis-step ( And it really doesn't help that it's only the 2nd level ).
4) We'd reduced the difficulty of the challenges due to the feedback in Haven, but like in the first game we struggled to come up with the unlockable awards. They're ok-ish, but we didn't do ourselves any favours by having such a limited pool to draw from, and they were very last minute and of course with that the panic of "Shit, we haven't got enough stuff".
Not a major failing, but something we need to think about more for the next game.
I'm sure more things went wrong, but they escape me now, but those things annoy me enough for 5 issues.
When it went on Kong the other day I thought I should really give it a go ( After voting 5 for it, it felt like the right thing to do ) and I'd forgotten I'd used the "They're here" sample on the intro which made me jump. Then I gave the first level a play, kinda half hearted as I have played it once or twice before, and I really loved it. There's a bit in Aliens after the sentry guns have run out and they're all in a room as the aliens are closing in, and then they all fall through the ceiling and it really kicks off. It felt like that. Fucking felt great to be honest.
So I think that's this about done. Not many regrets, lots of things to be pleased with. I wish in a way we could have done this one first before Haven, but things like that are impossible to second guess.
As always it was a pleasure working with Lux, and Rob and Theresa at Miniclip were a joy too.
We can't wait for the Owlmen to return, and it's going to be soon...