I Hate Mountains

Two weeks ago, we talked about the manor level and why we decided to do things the way they are right now. This week, the focus continues on the third level, the underground and on the overall campaign difficulty.

When Nicolas wrote the game design document about the campaign, one of the key elements was to build a rather long campaign with changing environments and gameplay. The goal was to provide players with a non-monotonous campaign that would work perfectly well as a whole, but also when playing each level independently. The global path had to be logical, but the gameplay type (open space/close combat) and the environment (exterior/interior) had to change beetween each map, following this approximative pattern :

As you can see, when you combine the different elements that make a Left 4 Dead level more or less difficult, we get the overall difficulty of each level. In game, this is pretty much correct for the moment, when testing we feel that each map is as difficult as it should be, and the third map is really difficult. Much like every official campaigns, we tried to balance the difficulty around the third map. The campaign starts slowly, progressively reach a crescendo at the third map (No Mercy’s sewers, Death Toll’s church, Dead Air’s construction site, Blood Harvest’s bridge) and slows down until the final arrives. But our third map is really hard.

At start, the underground level was supposed to be a catacomb-ish level, with crypts, coffins and all the extra cliché things you can think about. But when the prototyping time came, it was clear that it would never work the intended way. No one would ever want to walk in an endless network of catacombs and crypts during half an hour, plus that didn’t seem very realistic from the start. That’s why we decided to mix several underground environments in the same level. This choice was not really popular among the team until we all actually saw the result when the level reached a decently polished state. The different environments blended together perfectly well and if you didn’t pay attention to it, you didn’t even notice the changes until the end. For the moment, the level is composed of four different underground environments and architectures.

Iterative screenshots of a part of the underground level

The underground level was entirely made in a really short time, I’d say three weeks. Prototyping, gameplay tests, architecture and prop flooding were all completed in a matter of weeks. The only problem with this level is its length. For the moment, this is the longest level I ever played on Left 4 Dead. Taking the shortest path, it takes 3 minutes to run from the beginning to the end, while it only takes 2:25 on the longest Left 4 Dead level (No Mercy’s sewers, which is already a quite difficult level). For the moment, we believe the map is good as it is, because we’re able to finish it using the expert difficulty, but we are not against the idea of cutting a whole part if it’s really too long and difficult.

One of the key elements of making a level more or less difficult without interfering too much on its length and geometry is to tweak the item/weapon spawning. Unlike Left 4 Dead official campaigns, we decided to give more than 4 first aid kits in this level. Because the crescendo event is located in the middle of the level, you often reach it in an already bad shape… thus, you leave it in a terrible shape (if not dead), and there’s a whole half of the level left to do. That’s why we decided to give more than 4 first aid kits, because we know every player isn’t the best player in the world. The trap here is that we’re not really giving 4 extra first aid kits just before the crescendo event, so there won’t be enough for everyone… We believe it will force players to play efficiently and work as a team (or they will just insult each others and keep their first aid kits for themselves).

That’s all for the underground level, we are against spoiling everything before the release, so you probably won’t get more details before the release. On a side note, a few people asked us if we could give an approximative completion percentage for each level, so here it is:

• Forest: 90%
  What’s left: soundscapes, items/weapons tweaking, nav tweaking
• Manor: 60%
  What’s left: exterior, sounds, event, items/weapons, nav
• Underground: 95%
  What’s left: soundscapes
• Lumberyard: 50%
  What’s left: detailling, event, sounds, items/weapons, nav
• Lakeside: 15%
  What’s left: everything except prototyping and ambient

Stay tuned for more developer banter until the release.

Yesterday, we learnt that some people managed to get their hands on some private-exclusive-early access to the Left 4 Dead SDK. Which means a few things :

1. First, that there is a SDK. That’s not as bad as some people thought after all. Even if it’s five months late, it’s still better than never.
2. Second, that Valve is testing it. Therefore, we can assume that they got it right this time. Hopefully, when the “new” SDK comes out, it shouldn’t screw everything during several weeks this time. But seriously, we’re not really this confident, and chose to prevent the automatic SDK update in Steam anyway.
3. Third, we know that someone at Valve’s headquarters is reading this blog, Google Analytics is tracking you! What are you waiting for to offer us an early access to this SDK? Come on! Mail us at team@ihatemountains.com, dont be shy!

Anyway, since I developed a little tool to handle the custom content, sort it between Left 4 Dead content and external stuff and embed it automatically in the map files, it’s not much of a problem anymore. BUT STILL, we’d really like to see the new functionalities, if any.

A few days ago, we unveiled our new project I Hate Mountains and this website was already visited by more than 3,000 people. This is quite impressive and it motivates us for the next part of the development. We’d like to thank you all for your support and your confidence. It’s now our job to try not to disappoint you, and believe us, it will not be the case.

Talking about the campaign, here’s some fresh news.
In our first level, the forest, a big part had to be reviewed and redone after the release of Left 4 Dead. At the beginning, it was quite an impressive and vast open-space: fun to play, hard to optimize and finally… it had some problems to fit in the Left 4 Dead gameplay.

Making a Left 4 Dead level is not only a question of level-design and geometry. You can’t build it and play it straight out of the box like most of Valve’s games. You have to complete another important task, the nav mesh creation. The goal is to mark the paths for the game, these nav meshes contains a lot of informations for the artificial intelligence. Where can survivors and infected walk ? Where is the event ? Where can they respawn ? and stuff. The Director also needs it to build the gameplay during the game.

Problem is, when we got our first playable version of the level and tried to build the nav meshes, the process took three hours to reach 1%. Finally, we never managed to get a fully working version of the level. The game didn’t knew how to compute the path from the spawn point to the first safe room because there was way too many combinations in this vast open-space.

Conclusion ?
We found out that Left 4 Dead was not really suitable for open and vast environments. Something had to be done. This is why I spent the last weeks trying to transform this open-space level in a corridor-like version. This is not something we wanted to do, but we had to resign, and after a lot of talking and babbling with the team members, we found several ways to transform the level without loosing this precious open-space feeling.

Today, this first level is mostly finished and we believe the creation process recently reached 75% (unless the late release of the SDK messes everything up, wouldnt be such a surprise).

Concerning the other levels of this campaign, our progress is constant. The third level is already beeing tested right now. We’re trying to push the director to the limits, forcing him to make mistakes and telling us what’s wrong with the level and the nav meshes.

Stay tuned for more developer banter until the release.
PS: The screenshots page is now open.