Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Happylisk]

I think corruption is a smart mechanic that's handled well in the game.  Once you become moderately corrupted you begin to suffer heat spikes, misdirections, parts fall off, and energy drains.  Mostly harmless fluff, but in the right (or wrong) circumstances this can cause a fight to go south.  You also have weapons failing to cycle or even worse firing randomly.  You can't play cogmind without quickly learning to never carry a launcher around equipped while corrupted.

The one aspect of corruption that I think has some problems is data loss.  The worst offender, maprot, has already thankfully been removed.  Data loss as it remains essentially takes two basics forms.  One, you forget what parts are.  As a mechanic I think it works.  After you play long enough, you know what parts enemies are dropping after a fight.  What this mechanic really does is it forces you to make a decision when you encounter stockpiles.  Maybe those unrecognized devices are advanced force fields, maybe they're field recycling units.  You can strap one on to find out and restore the loss data but that will cost matter and, more importantly, time.  It forces the player to make a strategic decision with food clock ramifications, so I think it works.

My big problem with data loss  is that it can destroy data that's effectively irreplaceable.  Some of the best rewards in the cave branches are the derelict logs.  A log that gives you a branch entrance, zone layouts, or index(machine) for a future floor is huge.  The problem is, it's so easy to lose this data, even in the face of extremely trivial corruption. 

An anecdotal example:  I was in the proximity caves on -3, looking for the exit to research.  I found a log giving me the entrance to Testing on -3.  That's a huge bit of data!  I bumped into a programmer who gave me 2% corruption.  I slapped on my system restoration module and made for the exit.  In that interval I still managed to suffer two "data loss" corruption events.  On -3, I did not get the intel reveal for testing, and had to find it the hard way.

Maybe that was exceptionally bad luck, but I feel like it's way too easy to lose data from derelict logs due to corruption.  The whole point of branches is that they're high risk / high reward scenarios, and if the rewards can be that easily lost, they're much less desirable.  I know that the derelict logs aren't the main rewards in each cave branch, but they're certainly big reasons why I go in in the first place. 

While there are anti corruption devices in the game, I don't think that solves this particular problem this given how easily data loss can occur.  A system restoration module can take hundreds of turns to get rid of corruption, and walking around with a corruption screen to block corruption effects 1) is not guaranteed to work, and 2) handicaps your combat abilities.  They're certainly useful devices, but they're not going to prevent you from suffering the occasional data loss effect.  As it stands, even with anti corruption utilities, as little as 1% corruption in caves can result in you forfeiting extremely valuable intel.

tldr: People are inherently loss adverse.  They'd rather not lose something they already have than gain something of equivalent value. (See https://en.wikipedia.org/wiki/Loss_aversion).  It's a bummer to lose a branch reward, even if a minor reward, due to an RNG mechanic.  My suggestion: Put a corruption threshold before data loss occurs.  Even 5% would do it.  If you're running around caves super corrupted that's one thing, but if it's extremely minor corruption it'd be nice to avoid corruption effects that are in essence irreparable. 

[zxc]

I didn't know you could lose derelict log data through corruption. That sounds harsh.

I like the idea of thresholds for different corruption effects. This adds an even greater meta knowledge component though which I don't like. Perhaps, as you get more corrupted, you experience worse corruption effects a greater proportion of the time (as well as experiencing effects more often).

Another thing that can bug me is how even 1% corruption completely trashes your sensors if you don't have a signal interpreter. I think it could be more gentle than that. Plus, rare false alarm ?s on the map are probably more interesting and confusing than just littering the map with them all the time, but you'll never experience it in that way, only the spammy way.

The other thing is accidental firing, which can be EXTREMELY HARSH in the wrong places, and can happen with just 1% corruption. If that started happening past 5% corruption say, it would not be as ridiculous.

[DDarkray]

I always thought the data loss corruption loses only your item memory. I'm suspecting the Derelict Log you retrieved is showing you the exits found within Testing branch, not the exit to the branch.

[zxc]

I think it tells you the location of the exit leading to Testing as Happylisk says.

[Happylisk]

Derelict logs simply give you a hack selected from a list applied to a future level (the hacks being zone layouts, component/prototype stockpiles, security intel (ie sentry locations), machine index, branch/main access/exit points, etc).  In the same way you can lose data obtained from hacking a terminal, you can lose data obtained from these logs (ie future free hacks).  Easy to replicate.  On a factory level pull off a hack like Index(Terminals) and then let yourself get corrupted.  Eventually a data loss corruption will erase that terminal intel.   It's quite a bummer.

I've been mentally keeping track of the interplay between derelict logs and corruption for awhile.  I am 99.99% sure that corruption can erase derelict log info.  If I'm wrong Kyzrati will chime in.  It's also worth noting that if there's some confusion among veteran players like us about this mechanic, you have some opacity here that needs addressing.   

While I proposed a threshold before data loss occurs, zxc's weighting proposal works too.  As it stands, it seems like all corruption events have the same chance of occurring, and higher corruption merely 1) increases the frequency of corruption events, and 2) decreases the chances of successful hacking .  One solution to this problem is to make the worse corruption effects (data loss and FCS firings) have a lower chance to occur at low corruption levels.  Regardless of whether you use a threshold or weighted chance dependent on corruption level, you can avoid meta knowledge problems by 1) mentioning this in the manual, and/or 2) having a friendly robot in Zion mention this mechanic. 

zxc brings up a good point about the shooting corruption effects.  My first serious attempt at a research branch ended because of a weapon triggered by corruption around you know what.  I was barely corrupted, so that is quite the harsh result.  Yet another reason to address the worst corruption events happening at trivial corruption levels.   

There's a third solution besides weighting or thresholds: remove data loss as a corruption event except for the loss of item memory.  I didn't include this in my first proposal, since I thought it might be too extreme, but I do thing that intel lost is a not fun mechanic.  It reminds me of item destruction in DCSS: a not fun experience that was rightfully removed.  No one likes losing stuff because your number came up on the RNG lotto.  Note that item destruction in Cogmind is completely different: the entire game revolves around item destruction, and it clearly works as a mechanic.  I agree that data loss is very flavorful, but flavor should always take a backseat to mechanics and gameplay experience, and once a feature ventures into "not fun" territory it needs looking at.

Upon reflection, I realize I said something in my OP that was not correct.  The loss of derelict intel is not irreplaceable.  Because these logs are just free hacks, you could replace any loss intel (with the exception of the unique zone layout logs) by simply pulling off the hack on the next level.  The thing is, a pure combat build is not going to easily be pulling off top tier hacks like Access(Main) or Index(Machines), especially once you're in the research levels.  That's why losing the intel from these logs is so depressing!

[DDarkray]

The Data Loss message is a very vague term, and the only way I can clearly see its effect is from seeing "Unknown" parts lying on the floor. However, losing intels is very difficult for me to notice, which is why I never knew it could do that.

One suggestion I have is to make the Data Loss info more transparent. I think it would be great if the log message could specify what kind of data I'm losing.

[zxc]

Yep, however if it simply said 'item data lost' without specifying the item, that would be appropriate. It's kind of interesting not knowing what item you lost intel for.

Happylisk: discord when??!

[DDarkray]

Yep, however if it simply said 'item data lost' without specifying the item, that would be appropriate. It's kind of interesting not knowing what item you lost intel for.

I agree. I was thinking somewhere along the lines of:
System Corrupted: 10 Item Data Loss
System Corrupted: 1 Testing Exit Data Loss
System Corrupted: 2 Terminal Indexes Data Loss

[Happylisk]

Another random thought: if you have to have intel data loss in the game, you could add a function to system restoration modules giving a slight chance to restore loss data per turn. 

[Valguris]

System Restoration Modules are good enough already, but changing the effects of System Guards (12% chance to ignore a corruption effect) or adding new utilities that help us deal with corruption would be nice. Restoring lost intel is one idea. I thought about reducing effective corruption, i.e. item that reduces effective corruption by 5% would make 6% corruption be only as harmful as 1%, while any corruption of 5% or lower be harmless... as long as you spend a utility slot for it.

I, too, did not realise that 'data loss' can mean something other than parts ID loss.

[zxc]

That sounds cool. Like an item which halves effective corruption. So it might make 2% -> 1%, or 40% -> 20%. Percentage based rather than flat reduction would be cooler I think, and makes it more desirable when you have a ton of corruption (while if you have a little bit you might prefer a system restoration module).

[Kyzrati]

Happylisk's original assumption was incorrect. "Data loss" only loses part information, period. That's all it's ever done, and that has not changed. You cannot lose derelict log data, or terminal data, or anything else. (Perhaps Happylisk lost it some other way and misinterpreted what had happened, e.g. CR.)

In terms of other effects, I shied away from having a flat threshold for various effects if only because the existing system already handles it based on a percentage of your current corruption. So certain effects like misfires will already be more rare than data loss simply because they have a larger divider (current corruption/25%, checked at most every 15 turns, and can be blocked by other more common corruption effects--multiple effects are not allowed within 15 turns of each other).

[DDarkray]

Happylisk's original assumption was incorrect. "Data loss" only loses part information, period. That's all it's ever done, and that has not changed. You cannot lose derelict log data, or terminal data, or anything else. (Perhaps Happylisk lost it some other way and misinterpreted what had happened, e.g. CR.)

Glad the intel-rot effect isn't true. I was a little worry back there. 

In terms of other effects, I shied away from having a flat threshold for various effects if only because the existing system already handles it based on a percentage of your current corruption. So certain effects like misfires will already be more rare than data loss simply because they have a larger divider (current corruption/25%, checked at most every 15 turns, and can be blocked by other more common corruption effects--multiple effects are not allowed within 15 turns of each other).

Very interesting. I was wondering if it's based on your current corruption% or not because it seems like a lot of players feel that 1% corruption would mean that all types of corruption effect have equal chance of occurring. Perhaps this type of information could be made as part of an in-game lore?

[Kyzrati]

Well, I could just put all the numbers in the manual, but I'm not sure this matters too much. What really matters is the final in-game result and how it impacts players. (I personally don't find corruption much of an issue, basically ignoring it throughout the entire run and letting it accumulate however much it wants to, but different players have different approaches.)

[Sherlockkat]

I dislike being corrupted, even a little bit. Mostly because I always feel uncomfortable when I don't have full control of my char. That being said, corruption consequences (not effects) are very black and white. It's complete non-issue everywhere except in research branches. Then, the consequences are fairly drastic (oh, misfire? here, have another food clock)

I also dislike most of the corruption purge utils. They are somewhat grindy. They are pretty slow (except maybe the really high end that I never seem to find) that their only real utility is in caves (or chutes, if you are GJ). That might be the design intention, to have a hard to remove status effect. In which, the high end corruption utils (sys backup modules are fine due to their caves utility) can be rebalanced to be more interesting. Corruption based robot hacking capabilities might be interesting. Interfacing with a corrupted system should be detrimental in some sense, right?

[Happylisk]

Egg on my face!  I was pretty sure about this, but K's the only one who knows exactly what's under the hood.

Upon reflection, I think I know why I got this wrong.  Some logs gives you level/branch exit data - they tell you where an exit is when you enter a level.  Other logs give you entrance data.  I incorrectly believed this meant you would find this specific entrance on your map when you entered a level, which led me to believe that intel loss was occurring when i wasn't finding these entrances.  I'm now guessing that entrance logs act like a signal interpreter - they don't show you a staircase, but if you find it you know where it goes, just as if you had used a signal interpreter on it.   Can you confirm K? 

Knowing that corruption only effects item knowledge is a huge relief.  I've been playing for months with the belief that corruption would cause schematic and unaware analyses to be lost - good to know that's not the case.  K, might want to tweak the "data loss" corruption message to "component data loss" or something like that so players understand that data loss only effects item knowledge. 

[Valguris]

at most every 15 turns

So that's why the first few corruption % make a huge difference, while later ones do not! 1% corruption would mean ~once in 115 turns. 2% -> ~65 turns... 10% -> once in ~25 turns. 20% -> once in ~20 turns. 99% -> once in ~15 turns. There's barely any difference between 10% corruption and 99% corruption. But there is a huge difference in the first few corruption %.

Perhaps higher corruption should allow for more common occurences of side effects? Change the formula for minimum number of turns between corruption effects or remove this threshold altogether. I want to feel my 40% corruption actually be a problem rather than an annoyance.

[Kyzrati]

K, might want to tweak the "data loss" corruption message to "component data loss" or something like that so players understand that data loss only effects item knowledge. 

Done!

That text was written back in 2012 for the 7DRL when there were multiple types of data loss and they all just used the same message to keep the true effect cryptic, so it's about time to change that.

I'm now guessing that entrance logs act like a signal interpreter - they don't show you a staircase, but if you find it you know where it goes, just as if you had used a signal interpreter on it.   Can you confirm K? 

While yes there are both entrance and exit logs (two different things) and it's possible to confuse the two, exit logs technically do give the location before even seeing one.

That said, notice that there are also limits to the number of entrances/exits you get! For example you might learn of one exit to Research, but there are actually three on the current floor. So you can still come across exits you didn't know about that lead to the same place. (These numbers are shown in the log both when you gain the data and when it's applied to your current map knowledge, even if these two events occur in different locations as they often do.)

I want to feel my 40% corruption actually be a problem rather than an annoyance.

Everyone mob this guy

[Happylisk]

Adventures in corruption being a problem rather than an annoyance! 

Yesterday I'm in a research branch and I fry a lone grunt in a hallway.  A previously unseen "you know what" pops into view, turned an ominous shade of yellow.  I guess their vision is a pinch longer then yours?  I flee, round a corner, and put on my best weapon.  It's a special type of thermal cannon that does up to 100 damage, backed up my a particle charger.  "Not today buddy," I smugly think to myself.  The "you know what" rounds the corner and is at point blank range.  I fire!

A corruption effect triggers and the weapon fails to cycle. 

The "you know what" does to me what it does best.  Next turn my weapon goes off and the little bastard is one shot, but by that point too little too late.  ~roguelikes~

[Kyzrati]

Failing to fire at the wrong time can definitely be a bad one, and that happens quite a lot at high corruption!

Although corruption is meant to be another kind of clock, I don't think I'd want to escalate the effects to a true serious problem at the higher amounts as Valguris suggests. It would maybe put a bit too much emphasis on avoiding and clearing corruption? (If that were desirable, an alternative simpler approach would be to increase the amount of corruption you take from various sources like EM damage, so that the 100% corruption threshold would be that much more scary. Right now it rarely kills anyone.)

[DDarkray]

Although corruption is meant to be another kind of clock, I don't think I'd want to escalate the effects to a true serious problem at the higher amounts as Valguris suggests. It would maybe put a bit too much emphasis on avoiding and clearing corruption? (If that were desirable, an alternative simpler approach would be to increase the amount of corruption you take from various sources like EM damage, so that the 100% corruption threshold would be that much more scary. Right now it rarely kills anyone.)

I guess you can double the rate of receiving corruption for Cogmind, and then halve the chance of the corruption effects to occur to compensate. How often do you see players with 50% corruption anyway?

[Kyzrati]

It could be scaled up but I think that would end up overemphasizing the meta of corruption mitigation, which would also disproportionately affect combat over flight. As is it's something that sometimes throws little wrenches into plans and forces players to react, but isn't usually outright deadly.

In any case, the next release does make some changes to corruption, including a minimum threshold for a number of effects (most notably misfires).

[Happylisk]

That'll be a very handy change for dealing with certain branches!

Will there be an ingame way to see what those thresholds are?  Or will the manual note which effects have thresholds the point at which the effect is possible?

[Kyzrati]

I haven't added that yet, no, although I did/am considering it. Kinda sucks, to be honest, because in the interest of transparency I should make all the numbers known, but ugh, more numbers... Under the earlier rules this wasn't a thing.

Also, it'd probably have to be in the manual, where then I'd have to explicitly list all corruption effects...