Welcoming Our Less Precise Programmer Overlords

We've had over a hundred X1 runs uploaded since it came out earlier this month, in the meantime having also released a couple updates to that build as reported via additions to the original release post. The first patch was mainly a pair of important bug fixes, the second however was a modified version of the tracking experiment we've been playing with to somewhat increase long-term pressure while still avoiding direct tracking by extermination and assault dispatches.

I also did a couple of my own full runs, one of which I finished streaming this week (1, 2, 3). In it you can see some of the new legs, too.

Wanting to just see how everyone would react without any prior details, I haven't said much myself about how the changes work, though I've certainly been watching for any feedback, and so far it seems like there's support for our new experiments, both part disruption and the loss of tracking.

The latter probably goes without saying--players are practically always on the side of nerfing enemies :P (that said, I wouldn't say this is necessary an outright nerf more than simply a desirable change, but we'll get to that later).

And for disruption... well yeah it's a drawback, but not necessarily a major one except in at least one particular spot which was the primary impetus behind the desire to make that change.

Time for details!

Disruption

We'll start with this one because it's easier.

So again the main reason I wanted to remove Cogmind's part disruption immunity is actually due to a specific end-game encounter with a particular major NPC equipped with relatively significant disruption capabilities.

Balance-wise it may have been okay back when Cogmind had access to fewer tools, but in the years since their introduction, they've come to feel a little more underwhelming for it, so I wanted to up their power and use X1 to test out the possibility of doing it with a change to this mechanic.

The fact that Programmers are also capable of a low chance of disruption is less important, and at the same time it seemed like a decent change to pair with the reduced tracking capabilities since, in theory, you might be able to dodge even more Programmer squads if you're smart about it and/or have the means.

Another good reason to remove disruption immunity is to restore another facet of parity between Cogmind and other robots. In a roguelike you ideally want mechanics to work the same for everyone, avoiding special exceptions for the player, and while it's necessary in a few cases, it's overall not common practice in Cogmind, so being able to remove this immunity would make me a happy designer.

The main issue is of course whether doing so is too overbearing for the player. So far it seems like the answer is: not really. Your core is still immune, after all (oh no, partial exception retained :P), parts are not exactly disabled for that long, and it's not all that common. If it does happen it's also probably not super debilitating.

It can definitely lead to some hairy situations (I had one such instance in my own run force me to give up a particular plan when Programmers caught me in a bad spot), though that's essentially what playing a roguelike is about--dynamic problem solving.

That said, in X1 this change was only minimally applied, whereas once fully implemented there will be ways to negate disruption if that's important to you, including some ways that might directly factor into parts you're already using. Part shielding could offer immunity, for example, as could Dynamic Insulation, which might get more folks to actually carry and use that on top of its extant benefits.

As some players have brought up, it might very well be a challenge of the annoying type should the aforementioned major NPC (*spoilers, hence the lack of a name drop) maintain their current build while gaining the ability to disrupt Cogmind's parts, and I agree, so I can see combining this mechanical adjustment with a simultaneous adjustment to their build, either removing half of that capability or replacing it completely with something else. We'll see about that.

Overall the disruption thing seems like a good move.

Tracking

So... if they aren't fed your precise location, what exactly do exterminations and assaults do now? Essentially, they know your "general area."

Previously their tracking was enabled via the normal robot tracking behavior system, although when dispatched the squad leader would get a free record of your current location, one that would ignore their normal memory limits and last until they see you, at which time their standard memory takes over and you could lose them, as described by the in-game lore that explains this to some degree.(You could also destroy the leader who has the record, thereby forcing any followers who hadn't seen you yet to just wander around if they hadn't already been passed along local targeting info about you.)

Now these dispatches actually have to find you first. They do get a little help in that regard, as stated given the general location where they'll look around for you. You may very well be gone already by the time they get there. Over time, however, they will expand their search radius and continue looking, but they don't have any other details and rely on their own sight, or reports from their nearby allies, as usual. So depending on the layout they could easily head off in a direction further away from you, giving you a better chance to escape (even better if you have knowledge of their position, of course! sensors good).

This new behavior may or may not be a good thing, depending on your build and overall strategy (and again the layout as well). Whereas before you could more easily set up an ambush for an inevitable incoming attack, now you would have to get their attention first, and also can't otherwise be assured of which direction they'll come from, or when. Removing some predictability from their behavior introduces a little more possibility for chaos, and if you do manage to evade them once, maybe the situation is worsened if and when they come by again, so you may have a new choice in whether or not to be proactive about taking them out, or trying to leave the whole area as quickly as possible.

The original X1 build kept the system as simple and lenient as possible, just having them base their search around your position at the time when they were dispatched, though on playing with it for a while it felt like we needed a bit more pressure than that, so in the last patch (which I was using in the last stream of my run) while they do expand their search radius, it does still occasionally recenter on your area (while maintaining its radius), so there's always at least some chance you'll see them until you either deal with them or leave the map. Sensors, sufficient speed, and good planning can of course delay that possibility for a long while, barring other unexpected complications.

As with disruption, this feature was not fully implemented, just enough to test out what it feels like and whether it's desirable, and it seems like we probably want it, yeah? That being the case, there are quite a few other adjustments that will need to be made as a result:

• More randomization of extermination dispatches

• Assaults will be a little more aggressive, but right now they're using mostly the same approach as exterminations

• On that note would certainly need to take a look at high/max sec

• Qs will likely get some interesting tracking updates (don't worry if you don't know what Qs are)

• Reconsider usefulness of the Triangulator

• A number of other little tweaks to related systems and behaviors

• Related lore updates, naturally, including an NPC that's all "Back in my cycle, Programmers used to... I wonder what happened to that?" Kidding.

There's actually a good bit of work to do in there! The X1 experience is already pretty representative of what the majority of a run would be like, though, if you want to continue testing it out.

A few of the general implications of this new system as lifted from my notes:

• Exterminations/Assaults are both less predictable and harder to ambush (including with traps)

• Fast builds have an even easier time dodging them when using infowar (especially sensors)

• Easier to distract them with allies and drones

• Triangulator does not help in this case (hm, maybe it could actually still pinpoint these squads specifically? that would be neat and pretty useful)

• No more player complaints about perfect tracking :P

Next

I'll get to finalizing these features later and be waiting for any more feedback as I continue to work on Beta 15 content, in particular up next we have the PROTOFORGE (not at all ominous, nope). As of this week, Scraptown is now completely integrated into the plot and I'm ready to move on to the late-game UFD elements.