Forging Ahead: Ship It

When I was making my first round of changes after the play testing, one of the things I struggled with was figuring out why people were doing what they did on Crossroads. I had to sort out behavior and feedback that stemmed from a lack of familiarity with the map, versus actual flaws. Using the side man cannon for the rocket rush, for instance, didn’t happen because people didn’t know where the rockets were. But it was also partly because the man cannon was badly placed: it wasn’t visible and it was too far from the base, and people were just hopping into vehicles instead of wandering over to it. One was because the map was new, the other was bad design.

Similarly, after the third round of play testing I got some feedback that cut to the heart of the map, and I wasn’t sure whether it was a lack of familiarity with the map or a real problem. The general observation was, Crossroads was too big and if you weren’t on a vehicle, and took too long to traverse.

The reason I hesitated to do anything about it initially was I knew there was probably a mix of perception and reality at play. Back when I put the core layout down, I spent some time running the various routs between the bases on foot to see how long they took, in part to make sure I had the right mix of shortcuts and vehicles for players to use. And what I found was traversing Crossroads on foot  was comparable to other successful, large BTB maps in the Halo series. Valhalla, Avalanche and others took a long time to traverse even when using man cannons or teleporters. I figured, since Crossroads took comparable time to cross, it was just perception due to how open the map was.

Because Crossroads was a wide open map, players might perceive distances between locations to be larger than they were. It’s sort of like how moving at top speed in Monitor mode in Forge seems slow when flying the skies of Forge World, but the same speed can feel lightning fast when skimming the ground. In this case, because Crossroad’s major landmarks were visible from most areas of the map, approaching them on foot seems to take a long time. And that can make you feel like the map is larger than it actually is.

Bungie-made BTB maps utilize several tactics to break up their size. They place major land marks along the way, and generally create shorter sight lines so large maps are broken into sub-sections. So instead of trekking across the entire map, players move from one key juncture to another, which are not very far away from one another. Avalanche wraps around a curve, while Valhalla’s central hill bisects the map cleanly. Both use man cannons (and in Avalanche’s case, teleporters) to help move players across open spans. Terminal has a series of buildings to work through, while Rat’s Nest features curved hallways and compartmentalized rooms. The maps are big, but the major segments are less so.  But rarely do players on foot have visibility from one end of the map to the other along a route they have to trek. And when they do, such as in Rat’s Nest, it’s an optional one. But Crossroads didn’t have many sections of the map which were fully occluded from one another. The cover created segmentation of the firing lines for infantry, but the mix of elevation changes, long roadways, chasms and tall landmarks meant players could see one side from the other at almost all times.

So, I initially dismissed the criticism to mere perception. I was on the cusp of leaving it at that until I started to think about the problem differently. I realized that I had been focusing on how long it took players to get from base to base, straight across the map, and comparing that to various other Halo BTB maps. But more often than not, players aren’t trying to make a straight shot cross the map. They’re going after power weapons and vehicles, securing tactical locations and otherwise crisscrossing the entire span of the map during combat, even in objective games such as CTF. And I hadn’t spent any time evaluating how players on foot moved through Crossroads from landmark to landmark other than between the bases.

What I did was fly up as high as I could in Forge, looked down at the layout, and looked at the elements of the map that helped infantry move through it quickly. From base, players could take the man cannon down to the pyramid, or head up to the central towers for a man cannon launch to the far side. Over at the quarry is a Mongoose so players could move from that corner in either direction quickly. Meanwhile, the ramps and grav lift below the base move players from the bottom level straight up to the base area.

But there was a theme for the routes in place so far: they all helped players get around the sides of the map or from one end to the other. There was nothing in the middle areas to help players get out of there quickly.

I hadn’t given thought about how mobile players out in the center were once they were there. And it turns out, they had to take a long hike before they reached anywhere important. From the center area, it was a 30+ second trek back to the base from rocket spawn. And if you head down below and didn’t want to take the grav lift – say the enemy base was well defended – it was nearly as long a hike to the ramps next to the base. It was a long, open hike for infantry.

In the image above, it’s clear that getting infantry from the central ramps back to base is far too long of a slog. I needed to add some routs to chop it down, adding some movement loops to speed players through the center.  Avalanche had wonderful one: hit the teleporter, snag the rockets, then do the big vault down to the center for the laser rush. Then, loop back over the central man cannon and cut over to the base. The map was mammoth, but it was a brisk hike through thanks to the movement assists.

To get the rockets on Crossroads, you hit the man cannon, snagged the rockets and then….walked, for a very long time, back to base. I had added the tunnels along the lower bridges as cover, but that didn’t get people out of the center when they wanted – just protected and hid them as they went.

I decided to work the issue at both corners, starting with the middle ramp. Luckily, I had two man cannons and the budget to go with them. I put them down in the corner so they could skip the walk around the gap the path wraps around. It would hopefully be useful for loops like the rocket rush, and for infantry to close the distance on the base quickly. It might even motivate some additional usage of the middle now that a more viable attack route was present. To prevent vehicles and players from hitting it by accident, I used a wedge to push the man cannon to the inside of the corner and out of traffic lanes.

Players land gently right in front of the base. Vehicles make the trip over as well, though their landing is not quite as graceful, to put it mildly.

Down under the base, I worked in a ramped platform up to the top level next to the base. Now players below the base had three options, and didn’t have to make a long hike if the grav lift was defended. This also helped fill in all the “dead space” I identified last week, and gave that corner of the level a purpose.

With those two additions, I couldn’t come up with any routes from one key location on the map to another that was too long.

Since I still had a few hundred in budget left over, I did a final polish pass. I had noticed on the playtest before this update that I had violated one of my own rules about marking off ledges. The corner in front of the base was very tough to judge because the walkway had such a low profile.

I decided to block off the edges in a way that would also help infantry to close in on the base a little easier. Currently there was one lonely tunnel section in the center of the approach.

To add some additional color to the base approach from center, I added a second tunnel brace, and moved them to the ends of the bridge, rather than one in the center. Now both corners were properly marked off, and players had somewhere to duck in from center and next to the base. It also added a few places to safely house some additional spawn points.

And finally, I added a couple of pillars to the walkway that the grav lift was on. It was always too exposed but I didn’t have the right objects left in the budget to do much about it. With the base rebuild, they got freed up. Just having a block or two on the platform adds a lot to the space, as attackers or defenders have something to dance behind, rather than an open field shootout. The map had enough of those as it was.

And that was it for the geometry changes on Crossroads before the final play tests. I had the map configured for CTF, Slayer, Assault and Stockpile. But one of my favorite game types was Neutral Assault, and I’d just built a map that had a giant hole in the middle. Not exactly helpful for neutral game types. So I put the bomb (and one of the Stockpile flags) here:

I knew it would be either hilariously awesome, or hilariously bad, but it would be hilarious either way. Fortunately, the play testing was a lot more of the former than the latter. The predicament it places is wonderful: go for the bomb, and you land on the other team’s side of the map. Mwahaha.

The past two weekends, this version of Crossroads got put through a few play tests, and they went well. Not perfect, but well enough that I’m satisfied that the map is about as good as I can make it. People were familiar enough with the map that all the intended routes got used, and the new features were obvious enough that they didn’t require much explanation or experience to find. Feedback after the tests was positive, and I had no other nagging issues that I wanted to iron out. And also, budget was at $0. It’s all on the map.

It was a mix of gratification and relief going back over the final play testing films and seeing some fun, crazy stuff happening. It seemed to capture some of the spirit of Halo’s BTB battles, with long and mid range DMR fights, vehicular clashes and epic, hilarious moments. Such as this one:

So, here’s the link to the file set, which I submitted to the Forgetacular contest. There are still a couple tiny things I would like to do – and there was a report of a Ghost launch – but in reality I could probably tinker away at the map forever. At some point, it’s time to walk away and ship it, so to speak. And the deadline for Forgetacular is coming up fast.

On a programming note, it wasn’t my intention to hold back on posting a link to the map until the very end. But the timing of the blog posts meant I was always writing about a version that was about two weeks old, and I thought it would be unfair to post what I knew to be an inferior version, and one for which I would be ignoring community feedback. But starting with map #2 – still in the primordial stages – I’ll try to make this series more real time and interactive. As soon as I get a playable version up and running, I’ll put it out into the wild, and then try to incorporate feedback into later versions. I haven’t been responding to comments on the site thanks to a cookie issue with my browser which has been blocking my responses (in many ways, I am not good with computer), but I’ve been reading and appreciating all of them and should have the issue sorted out soon.

Thanks for following so far.


Click here to return to the Forging Ahead Index


  1. Josh
    November 17, 2010

    hey I have been reading on your progress for a while and have been looking forward to playing this map. Just downloaded it and will play it tonight.

  2. November 17, 2010

    Wow, it really looks good 🙂

    Thanks for all the help received and really hope you get far.

    I’m getting all my mates to play on it tonight 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *