Stealth and Risk

For combat stuff that doesn't fit into any of the other forums.
Acarin
Posts: 121
Joined: Fri Sep 04, 2020 10:49 pm

Re: Stealth and Risk

Post by Acarin » Mon Feb 15, 2021 1:45 pm

So for the sake of making this easy, I've made some assumptions here and will explain as I go:

First, I make the assumption that the person searching is rolling the median value within their range each time... I know this is a huge assumption but didn't want to have to work with 2 variable rolls here... so someone with 100 perception skill would have a max roll of 200 and always be rolling 100 for the purpose of this calculation. Next I assume that p(being seen)=1-[p(not being seen based on a single roll)^(1+ absolute value of reroll number)].
The most simplistic situation is the following:

Stealth of hider and perception of searcher are equal (percentage is approximate chance of being seen on one roll):
0 rerolls: 50%
-1 reroll: 75%
-2 rerolls: 88%
-3 rerolls: 94%
-4 rerolls: 98%
-5 rerolls: 99%

0 perception versus 500 stealth:
0 rerolls:9%
-1 rerolls: 16%
-2 rerolls: 23%
-3 rerolls: 29%
-4 rerolls: 35%
-5 rerolls: 41%

300 perception versus 600 stealth:
0 rerolls: 29%
-1 rerolls: 49%
-2 rerolls: 64%
-3 rerolls: 74%
-4 rerolls: 81%
-5 rerolls: 87%

Ok... so again, each of those is the probability of a single median roll on the search end resulting in a stealth character being spotted (or the result of a single stealth check). As a note, in order to get into position to ambush with 0 rerolls against something with like perception, the chance of achieving this is approximately 0.50*0.50*0.50 = 12.5%... so in setting up an attack against someone of equal perception, we're seen 87.5% of the time and induce paranoia. I assume that there is usually more of a gap for mobs though and that wechuge perception, for example, is lower than mine... this would be the result if I was trying to ambush black, however... I would have about a 12.5% chance of not being seen. Pretty low given our skill is approximately equal.

Ok so... risk accumulation. Going to provide a straight progression based on 3 and 5 sec rts... haven't logged in to check the actual rt yet.

3 second rt:
Start of search: 12 risk, 0 seconds elapsed
Search 2: 12 - 3 +12 = 21 risk (-1 reroll), 3 seconds in
Search 3: 21 - 3 +12 = 30 risk (-1 reroll), 6 seconds in
Search 4: 30 - 3 +12 = 39 risk (-1 reroll), 9 seconds in
Search 5: 39 - 3 +12 = 48 risk (-2 rerolls), 12 seconds in
Search 6: 48 - 3 +12 = 57 risk (-2 rerolls), 15 seconds in
Search 7: 57 - 3 + 12 = 66 risk (- 3 rerolls), 18 seconds in
Search 8: 75 risk, (-3 rerolls), 21 seconds
Search 9: 84 risk (-4 rerolls), 24 seconds
Search 10: 93 risk (-4 rerolls), 27 seconds
Search 11: 100 risk (-5 rerolls), 30 seconds

5 second rt:
Start: 12 risk 0 seconds elapsed
Search 2: 12 - 5 + 12 = 19 risk (0 rerolls), 5 sec
Search 3: 26 risk (-1 reroll), 10 sec
Search 4: 33 risk (-1 reroll), 15 sec
Search 5: 40 risk (-2 reroll), 20 sec
Search 6: 47 risk (-2 reroll), 25 sec
Search 7: 54 risk (-2 reroll), 30 sec
Search 8: 61 risk (-3 reroll), 35 sec
Search 9: 68 risk (-3 reroll), 40 sec
Search 10: 75 risk (-3 reroll), 45 sec
Search 11: 82 risk (-4 reroll), 50 sec
Search 12: 89 risk (-4 reroll), 55 sec
Search 13: 96 risk (-4 reroll), 60 sec

Ok... so no one is clearly meant to stand there for close to a minute while someone is searching for them...

But let's now calculate the probably of being caught on each search using the numbers above...

We can start with someone with perception at half of our stealth and a 5 second rt and 0 starting risk and calculate the probability of being seen after X searches.

Search 1: 29%, 0 seconds
Search 2: [1 - (.71*.71)]*100= 50%, 5 seconds in
Search 3: [1 - (.71*.71*.71)]*100= 64%, 10 seconds in
Search 4: [1 - (.71*.71*.71*.51)]×100= 82% 15 seconds in - this is where the reroll kicked in
Search 5: 91% chance of having been seen, 20 seconds in
Search 6: now we have added another reroll - so [1- (.71*.71*.71*.51*.51*.36)]*100 = 97%, 25 seconds in.

This means that even if you have twice the skill of your opponent, you are still near guaranteed to be seen before 6 searches. That means that 600 stealth against 300 perception will essentially allow you to be found within 25 seconds. I will do the math for the other scenarios in a bit... but the point is that even with a massive skill gap, I'm likely to be noticed VERY quickly and nearly guaranteed to be noticed by 5 or 6 searches.
You reach toward ((DEV Rias)) ... Pull(d225([1]x)):214 vs Mark(d1100):714 = -500 (-222%)
You notice ((DEV Rias)) glance your way, causing you to quickly withdraw your hand from his wool
drawstring pouch (open).
Roundtime: 5 seconds.

Acarin
Posts: 121
Joined: Fri Sep 04, 2020 10:49 pm

Re: Stealth and Risk

Post by Acarin » Mon Feb 15, 2021 3:07 pm

Ok. A few more...
For 0 perception against 500 stealth with 5 second search rt:
1: 9% chance of being seen, 0 seconds elapsed
2: 17%, 5 sec
3: 25%, 10 sec
4: 47%, 15 sec (risk kicks in)
5: 56%, 20 sec
6: 66%, 25 sec
7: 74%, 30 sec
8: 80%, 35 sec
9: 86%, 40 sec

So with the current system, someone with no perception can spot someone with 500 stealth in about 15 seconds (4 searches) and is 80% likely to have found this person in 35 seconds.

Without risk rerolls, the progression would look like:
1: 9%, 0 sec
2: 17, 5 sec
3: 25%, 10 sec
4: 32%, 15 sec
5: 38%, 20 sec
6: 44%, 25 sec
7: 49%, 30 sec
8: 53%, 35 sec
9: 58%, 40 sec

Note that in the current risk scenario, it takes 15 seconds to have around a 50% chance of finding a highly skilled stealth user with no skill at all... without risk, it takes about 30 seconds.

A 3 second search scenario with risk (0 perception versus 500 stealth) would look like:
1: 9%, 0 sec
2: 17%, 3 sec
3: 31%, 6 sec
4: 42%, 9 sec
5: 51%, 12 sec
6: 63%, 15 sec
7: 71%, 18 sec
8: 80%, 21 sec
9: 86%, 24 sec
10: 91%, 27 sec
11: 94%, 30 sec
Note that in this scenario, someone will no skills spots someone with 500 stealth over half the time in 12 seconds and almost all the time in less than 30 seconds.

Progression with 5 sec search rt when both stealth are perception are equal looks like:
1: 50%, 0 sec
2: 75%, 5 sec
3: 87.5, 10 sec
4: 97%, 15 sec (risk has kicked in - normal would be around 93% - maybe not a big deal since you've been seen anyway)
5: 99.3%, 20 sec

Note that to have any reliable chance of an ambush against someone, your stealth MUST be far greater than their perception but even people with no perception can see you very quickly of they search. The reroll penalty does dramatically increase this problem.

My point about nightblades was that being drawn out in these scenarios adds a combat penalty on top of being found... but let's ignore that for now and just discuss these rates.

If I invest 500 ranks into a skill, should someone who has not invested be able to find me more often than not in 12-20 seconds? Should I also only have a 12.5% chance of sneaking up on someone with equal skill? Should they near ALWAYS be able to find me within 4 searches (but most of the time before). Maybe the answer is yes to all of those. To me, it seems that to reliably use stealth in the past, skill needed to be disproportionately high compared to perception (the other game too - I have mentioned this before) and even then, you're likely (not just able) to be seen quickly.

Just throwing these numbers out there because risk compounds the problem. If stealth rerolls started at 1 or 2 or abilities supplemented them, this progression would seem less concerning to me. I dont know though... maybe all this is intended. Would live to hear what others think.
You reach toward ((DEV Rias)) ... Pull(d225([1]x)):214 vs Mark(d1100):714 = -500 (-222%)
You notice ((DEV Rias)) glance your way, causing you to quickly withdraw your hand from his wool
drawstring pouch (open).
Roundtime: 5 seconds.

Acarin
Posts: 121
Joined: Fri Sep 04, 2020 10:49 pm

Re: Stealth and Risk

Post by Acarin » Mon Feb 15, 2021 3:13 pm

Just to add... I get that I'm not supposed to be standing there letting someone search. The question to me is less about whether I'm actually going to do that thing and more along what it would look like if I did (and does that seem reasonable). Someone could definitely get off a search or 3 before I disengage though... and there's no signal when paranoia ends so might be hard to judge after trying to disengage from open combat.
You reach toward ((DEV Rias)) ... Pull(d225([1]x)):214 vs Mark(d1100):714 = -500 (-222%)
You notice ((DEV Rias)) glance your way, causing you to quickly withdraw your hand from his wool
drawstring pouch (open).
Roundtime: 5 seconds.

User avatar
Rias
DEV
Posts: 952
Joined: Sun Sep 03, 2017 4:06 pm

Re: Stealth and Risk

Post by Rias » Tue Feb 16, 2021 12:59 pm

As a note, in order to get into position to ambush with 0 rerolls against something with like perception, the chance of achieving this is approximately 0.50*0.50*0.50 = 12.5%... so in setting up an attack against someone of equal perception, we're seen 87.5% of the time and induce paranoia.
Passive chance for discovery is not the same as chance for discovery when being searched for. Stalking a target has base 2 rerolls in favor of the stealther. Sneaking into a room also has a base 2 rerolls for the stealther. Risk is still a factor.

I also was mistaken about Mark having a discovery chance. I could have sworn there was a chance for getting a creepy feeling when someone uses Mark against you (and inducing paranoia), but I was wrong: There's no skill check at all for initiating Mark. (I was probably thinking of The Other Game.)

Here was my experience in Risk accumulation with the searcher always searching at 3 seconds. I am stating the Risk and Rerolls as calculated for that search's skill roll (recall that risk ticks down during the roundtime before the searcher can search again), as the Risk is applied after the skill roll and will therefore apply to the next search, not the current one.

Search 1: 0 Risk, 0 Rerolls
Search 2: 9 Risk, 0 Rerolls
Search 3: 18 Risk, 0 rerolls
Search 4: 27 Risk, -1 reroll
Search 5: 36 Risk, -1 reroll
Search 6: 45 Risk, -2 rerolls
Search 7: 54 Risk, -2 rerolls
Saerch 8: 63 Risk, -3 rerolls
Search 9: 72 Risk, -3 rerolls
Search 10: 81 Risk, -4 rerolls
Search 11: 90 Risk, -4 rerolls
Search 12: 99 Risk, -4 rerolls
Search 13: 100+ Risk, -5 rerolls
This means that even if you have twice the skill of your opponent, you are still near guaranteed to be seen before 6 searches. That means that 600 stealth against 300 perception will essentially allow you to be found within 25 seconds.
I will say first that I'm planning on adding a reroll to the hider, so they have a base +1 and don't get into the negative as quickly. That said, the idea is still that a hider will be discovered eventually, and within a fairly short time, if they remain in a room while being searched for.

It might help for me to know what kind of scenario you're imagining where a character remains in hiding at length while someone is searching for them, without taking any action that either removes them from hiding (like an attack or maneuver) or removes them from the room that is being actively searched (sneaking away from the people searching for them). I did see a mention of talking to people from hiding. In that case, I think the increasing Risk is particularly justified - the searcher is 100% certain someone is hiding in the area, and is likely following the sound of the hider's voice.

Being discovered by a searcher shouldn't be considered Game Over. I also suggest leaving the room when people start searching if the plan is not to make an ambush (in which case I would say just start your attack early - refrain from an "all or nothing" approach). Stealth is not invisibility, so sometimes stealthers will need to reposition or retreat. One can always sneak back in after letting their Risk tick down, and hopefully by then the searcher is satisfied there's nobody to be found and has stopped searching. That's additional Stealth movement checks, but it's much better than remaining passively in the room being searched and taking that Risk accumulation which will hurt rerolls and indeed make it highly likely the hider will be discovered.

On the flip side of all this: I do want to discourage people from just constantly searching all the time. Like having a conversation and just searching the entire duration because you can talk during roundtime, so why not? I may add some kind of sanity penalty to excessive searching, or some other meaningful discouragement. Hmmm ... maybe something as simple as increasing roundtime as more searches are performed within a short timeframe would do the trick.
Note that to have any reliable chance of an ambush against someone, your stealth MUST be far greater than their perception
I really feel that logging in and playing and getting a feel for the actual in-game experience would be beneficial. I've played my own Nightblade a lot over the past couple days as a result of this discussion, as well as played yours, and this just has not been my experience on either of them. Stealth setup favors the stealth user with rerolls. If anything, it feels quite easy. I've got a very simple routine with no special tricks or inside info. Go to the combat zone, hide, sneak around until I find a mob (or one wanders in). Stalk twice, mark, attack when prompted by Mark. The majority of my attempts are one-shot kills. I rarely get noticed or trigger paranoia in the mob. Log in, play around with it, and let me know if you have a different experience so we can pin down where things are going different between us. I feel this will be useful for seeing if the actual experience lines up with your theoretical concerns.
If I invest 500 ranks into a skill, should someone who has not invested be able to find me more often than not in 12-20 seconds?
If the unskilled person is searching and the hider is doing nothing to react, then I find this reasonable. As mentioned before, stealth is not invisibility. Don't get me wrong, I'm pro-stealth. My bias falls in favor of stealth users, as I almost always end up playing the rogue type in games. I love the stuff. But I don't feel it should be treated as an invisibility/immunity skill. It should give a significant initial edge for sure, and stealth users can do all sorts of fun stuff when others don't have reason to suspect. But when someone starts looking for me, I should have to take some kind of action, not just sit passively and assume my skill is going to keep me safe. And again, being detected in stealth is not Game Over for the stealther. They just need to adjust tactics at that point. They haven't taken any damage or injury. And as I've said, Nightblades are very capable at open combat for the times they're spotted. A Nightblade should be adaptable, rather than limited to being a one-trick pony who cannot be viable without a specific flawless stealth attack setup.

I'm willing to tweak - as I said, I'll probably put in a base +1 reroll for the hider to prevent them getting into the negative quite so quickly. But my experience has been that Nightblades feel not only viable, but especially so, even after these change to searches causing them to be detected in a shorter time.

Acarin
Posts: 121
Joined: Fri Sep 04, 2020 10:49 pm

Re: Stealth and Risk

Post by Acarin » Tue Feb 16, 2021 3:57 pm

Rias wrote:
Tue Feb 16, 2021 12:59 pm
Passive chance for discovery is not the same as chance for discovery when being searched for. Stalking a target has base 2 rerolls in favor of the stealther. Sneaking into a room also has a base 2 rerolls for the stealther. Risk is still a factor.
Thanks for this clarification. Helpful info.
I also was mistaken about Mark having a discovery chance. I could have sworn there was a chance for getting a creepy feeling when someone uses Mark against you (and inducing paranoia), but I was wrong: There's no skill check at all for initiating Mark. (I was probably thinking of The Other Game.)
Very useful info. I didn't factor that in to anything but have never been caught on a mark so figured this was the case.
Here was my experience in Risk accumulation with the searcher always searching at 3 seconds. I am stating the Risk and Rerolls as calculated for that search's skill roll (recall that risk ticks down during the roundtime before the searcher can search again), as the Risk is applied after the skill roll and will therefore apply to the next search, not the current one.

Search 1: 0 Risk, 0 Rerolls
Search 2: 9 Risk, 0 Rerolls
Search 3: 18 Risk, 0 rerolls
Search 4: 27 Risk, -1 reroll
Search 5: 36 Risk, -1 reroll
Search 6: 45 Risk, -2 rerolls
Search 7: 54 Risk, -2 rerolls
Saerch 8: 63 Risk, -3 rerolls
Search 9: 72 Risk, -3 rerolls
Search 10: 81 Risk, -4 rerolls
Search 11: 90 Risk, -4 rerolls
Search 12: 99 Risk, -4 rerolls
Search 13: 100+ Risk, -5 rerolls
This seems to be aligned with what I put up previously but we just have a different starting number. One accumulates 12 risk on the first search which ticks down by 3 before the next search is initiated. My table is a little different, but we are saying the same thing with you providing the risk value that is going in to the search and me provided the risk value post search. In the previous probability calculations, I did account for risk penalties occurring on the roll after that risk value was accumulated.
I will say first that I'm planning on adding a reroll to the hider, so they have a base +1 and don't get into the negative as quickly. That said, the idea is still that a hider will be discovered eventually, and within a fairly short time, if they remain in a room while being searched for.
I think this is a great idea.
It might help for me to know what kind of scenario you're imagining where a character remains in hiding at length while someone is searching for them, without taking any action that either removes them from hiding (like an attack or maneuver) or removes them from the room that is being actively searched (sneaking away from the people searching for them). I did see a mention of talking to people from hiding. In that case, I think the increasing Risk is particularly justified - the searcher is 100% certain someone is hiding in the area, and is likely following the sound of the hider's voice.
Most ranged stealth users will likely not make a single attack then run. I would think that while ranged should be balanced and someone shouldn't be able to fire 10 arrows without being spotted, they should be able to make it through a few combat rounds against someone with no perception (if their perception is high) before there is a high chance of being seen (see previous probability). With a 5 second rt for attacks, being seen by something with no perception 50% of the time before you can initiate a 4th attack seems a bit fast. I'm not really ranged though so doesn't matter to me. As I stated previously, I don't intend to stand and do nothing while someone is searching anymore. In the past, I would do this while trying to wait out a mark. Instead, I will walk into another room and sit around for a minute or two. Stealth shouldn't be invisibility or last forever but why should someone who has not invested in spotting me be able to do it reliably in a couple combat rounds (and also give me penalties at the same time). Stealth is definitely strong but requires an investment. What's the point of investing if someone fresh out of the generator can spot me in 15 seconds and drive my combat bonuses down as well. If I want to see people easily, I should invest in seeing them. If you want to see what I mean, you can test this on my clone by throwing a knife at a vagrant and counting the searches until you're found. It's pretty fast. No, it's not how I intend to play but it's definitely illustrative of the impact of the change.

I understand wanting anything to be possible with rolls and also limiting stealth but I'm unlikely to strike a greater wechuge in 5 rounds with no melee and dodge (I could but the outcome is really unlikely). Why should it be so dramatically different for perception?
Being discovered by a searcher shouldn't be considered Game Over. I also suggest leaving the room when people start searching if the plan is not to make an ambush (in which case I would say just start your attack early - refrain from an "all or nothing" approach). Stealth is not invisibility, so sometimes stealthers will need to reposition or retreat. One can always sneak back in after letting their Risk tick down, and hopefully by then the searcher is satisfied there's nobody to be found and has stopped searching. That's additional Stealth movement checks, but it's much better than remaining passively in the room being searched and taking that Risk accumulation which will hurt rerolls and indeed make it highly likely the hider will be discovered.
To me, this is an issue about the amount of time that it takes to accumulate these penalties. Lets assume someone searches and I attack immediately following the first search. I am jumping in to that attack immediately with a nightblade defensive reroll decrease as I also took 10 risk from the attack. This penalty is based only on the fact that the individual searched. Again, I did not say that I plan to stay in a room forever while someone searches, only that the rate of risk accumulation is too fast from the search action. You will probably try to address this situation and tell me why I'm playing it wrong or that it would tick down before he can attack again (aware) or so what if my roll isn't optimal (aware). I stated this because risk from other sources also accumulates quickly... that includes in open combat. The risk accumulation from a single search is also a bad start to open combat even I was not spotted on that search.. This means that I am going into the open with risk already accumulated and am not just waiting down that search risk. I tried this before... hurling a knife (10 risk), 2 search ticks through rt and moving in to melee, and then ambushed. My risk was over 50 by the time I made the attack, the attack adding 10... seeing my risk nearly hit 60 (and therefore negative rerolls) within a short period, I did decide to leave and run down risk. This is clearly not a realistic situation (not how I would fight), but with this search change, risk accumulation is way more rapid. What this does is pigeonhole a nightblade into only attacking if no checks are failed and running and sitting around for a couple minutes if someone is tipped off at all or you fail a stealth check. Going into open combat can work but getting to non-warrior rerolls if not done very quickly.. So yes... you can still be quite effective but a single viable option (succeed on ambush or run) gets real boring after the 1000th time doing it. And yeah... the rp of being a stealthy type (who typically does remain in hiding) is definitely also not very fun when anyone can spot you or penalize your combat with a few quick searches, regardless of how skilled that player is. But killing people without interaction/warning is definitely still viable so I suppose not a total loss.

Playstyle aside...

I really did try to make an objective case based on the numbers I presented as to why I was making the claim for the hypothetical situation to see if those numbers felt right to you. It had 0 to do with playstyle. I am not stating that nightblades are not viable or that we should be invisible forever or that this situation will be occurring frequently. As I very clearly said, the calculations were around whether the numbers look right to you. If they do and you think they're reasonable... that's great. I think the risk accumulation is very high (20% more risk per search than an attack).

Big potential issue as well... does risk accumulate from searches when you are not in hiding too? I think it does but haven't confirmed...
On the flip side of all this: I do want to discourage people from just constantly searching all the time. Like having a conversation and just searching the entire duration because you can talk during roundtime, so why not? I may add some kind of sanity penalty to excessive searching, or some other meaningful discouragement. Hmmm ... maybe something as simple as increasing roundtime as more searches are performed within a short timeframe would do the trick.
I have been in many situations where I cough and someone just keeps searching until they find me 20 or 30 searches later (a situation where I do not care if I'm caught). This seems excessive on the searcher end (not just on the me being able to stay hidden end). If stealth takes increasing penalties from searching, I entirely agree that excessive searching should also be discouraged with rapidly increasing rt or a cap to number of searches within a particular timeframe. There is also the CvC group combat in factions situation to think about... Someone could just stay back (guarded) and continually search to disable all stealth in the room through risk increases. Increasing rt for multiple searches would likely accomplish this.
I really feel that logging in and playing and getting a feel for the actual in-game experience would be beneficial. I've played my own Nightblade a lot over the past couple days as a result of this discussion, as well as played yours, and this just has not been my experience on either of them. Stealth setup favors the stealth user with rerolls. If anything, it feels quite easy. I've got a very simple routine with no special tricks or inside info. Go to the combat zone, hide, sneak around until I find a mob (or one wanders in). Stalk twice, mark, attack when prompted by Mark. The majority of my attempts are one-shot kills. I rarely get noticed or trigger paranoia in the mob. Log in, play around with it, and let me know if you have a different experience so we can pin down where things are going different between us. I feel this will be useful for seeing if the actual experience lines up with your theoretical concerns.
I dont have problems killing things with mark. This has never been an issue. I am, however, regularly seen by even very easy things as expected. If you are only being seen less than 5% of the time, we are having very different experiences. I will do some additional testing eventually though and give you actual non-"theoretical" results
If the unskilled person is searching and the hider is doing nothing to react, then I find this reasonable. As mentioned before, stealth is not invisibility. Don't get me wrong, I'm pro-stealth. My bias falls in favor of stealth users, as I almost always end up playing the rogue type in games. I love the stuff. But I don't feel it should be treated as an invisibility/immunity skill. It should give a significant initial edge for sure, and stealth users can do all sorts of fun stuff when others don't have reason to suspect. But when someone starts looking for me, I should have to take some kind of action, not just sit passively and assume my skill is going to keep me safe. And again, being detected in stealth is not Game Over for the stealther. They just need to adjust tactics at that point. They haven't taken any damage or injury. And as I've said, Nightblades are very capable at open combat for the times they're spotted. A Nightblade should be adaptable, rather than limited to being a one-trick pony who cannot be viable without a specific flawless stealth attack setup.
Well ok. Just because someone is waiting (standing there or remaining in the room) doesn't mean they're doing nothing. I could be trying to wait a few more seconds for mark to activate. I could be trying to move in to melee and wait til the target restarts rt. 12 seconds is a little over the time of 2 attacks and it's hardly a long time. If someone who has capped stealth shouldn't be able to remain reliably hidden from someone with no skill for a few combat rounds (again, 12 seconds is very far from indefinitely) and quickly starts taking penalties... then I just don't know what to say.

Not sure about the one-trick pony thing. That's kind of what nightblades are right now... maybe two tricks (mark/ambush and sweep/aimed shot).
I'm willing to tweak - as I said, I'll probably put in a base +1 reroll for the hider to prevent them getting into the negative quite so quickly. But my experience has been that Nightblades feel not only viable, but especially so, even after these change to searches causing them to be detected in a shorter time.
Cool. Good idea.
You reach toward ((DEV Rias)) ... Pull(d225([1]x)):214 vs Mark(d1100):714 = -500 (-222%)
You notice ((DEV Rias)) glance your way, causing you to quickly withdraw your hand from his wool
drawstring pouch (open).
Roundtime: 5 seconds.

User avatar
Rias
DEV
Posts: 952
Joined: Sun Sep 03, 2017 4:06 pm

Re: Stealth and Risk

Post by Rias » Tue Feb 16, 2021 4:45 pm

I understand wanting anything to be possible with rolls and also limiting stealth but I'm unlikely to strike a greater wechuge in 5 rounds with no melee and dodge (I could but the outcome is really unlikely). Why should it be so dramatically different for perception?
It's to put increasing pressure on the hider and push them into taking some kind of action. While hidden, one is unable to be targeted and therefore not at risk of combat rolls striking them and such, so I feel it's justified to put some extra pressure on hiders to act. Relevant note: The balance system was initially implemented to do the same for combat. It used to be that incoming attacks lowered balance and outgoing gained it back, so if a target idled only relying on their high skill numbers to keep attacks from landing, their defense rerolls would become increasingly worse until they inevitably started taking hits. It caused various other issues with Balance also being a resource that people were expected to spend on other actions so we shelved that aspect of combat, but I do hope to bring it back one day in a better way.
Just because someone is waiting (standing there or remaining in the room) doesn't mean they're doing nothing. I could be trying to wait a few more seconds for mark to activate. I could be trying to move in to melee and wait til the target restarts rt.
And that's fine. It's up to the ambusher to decide how much time and risk they want to take. Squeezing in a stalk to melee when someone is searching could very well be worth it. I wouldn't personally wait for a Mark to trigger as that's more of a random gamble than I'd prefer, but if someone else wants to, that's their choice to make. I just want there to be meaningful pressure to make those choices.

User avatar
Candelori
Posts: 30
Joined: Thu Sep 07, 2017 5:29 pm

Re: Stealth and Risk

Post by Candelori » Tue Feb 16, 2021 11:56 pm

This has been a very interesting discussion to read. I appreciate these updates, speaking as one of the blundering oafs who will only ever have a paltry helping of perception to his name. All the same, I can understand the frustration of someone investing heavily in a skill to see it be made so ineffective in so short a time. It is good to see that the stealth users will be given an improved starting position and have more time before the more serious negative reroll penalties are applied. I wonder if it might also be worth considering making the decline more gradual, to allow stealth users to enjoy their skill a while longer before being forced to leave their hiding spot. We may still add pressure for stealth users to act while giving them some little more time to consider their options or make preparations for adjusted strategies, or perhaps time enough to deliver a venom-filled speech from the shadows before their next action.

If a balance is needed in exchange for allowing stealth users to spend more effective time in the shadows, which seems to be the primary matter of concern here, the stealth attacks themselves might be made a touch less potent. I know I do have some of my own concerns at having my own very heavy and expensive skill investments in melee, shield use, and armor made inconsequential against someone with any talent for stealth and finding chinks in armor. I have yet to face such an opponent or experience such a situation in the game, however.

I should also say I would be very interested to see the similar system for combat re-implemented in some way, though that is a discussion for another topic at another time.

Acarin
Posts: 121
Joined: Fri Sep 04, 2020 10:49 pm

Re: Stealth and Risk

Post by Acarin » Wed Feb 17, 2021 10:33 am

Candelori, thanks so much for the input here and for helping me do a little testing last night.

Rias, I went in to wechuges last night for a little testing as well as you suggested. I'm seen by them approximately 1/4 of the time and coming out of hiding and running does not trigger paranoia so this really isn't an issue for hunting like level mobs as their base perception is high enough that I'm forced to react if seen... so not an issue. Yes, I am still an effective hunter.

To clarify further, I also checked how long it took someone at a lower perception value to spot me. I was found on 1, 2, 3, 3, 4, and 6 searches (median value being 3 searches out of this limited data set - allowed risk to reset to 0 before retesting). Perception was less than 1/3 of my current stealth. I have also replicated similar results on vagrants. Adding a reroll to prevent this immediate discovery would be very much appreciated. I do get that we should be forced to make a decision. In equal combat, we are forced to do that very quickly (usually after the first search) which is fine. Where I see the problem coming in, again, is for things that are way weaker. 2 scenarios currently exist: we're found quickly upon search, we're not found and risk rolls up until we are (potentially with a very odd combat disadvantage).

Potential suggestions:
1. As you have suggested, add a starting reroll so the range of potential rerolls is 1 to -4. This will give about another 10 seconds before being found but will make it harder for people to be found on an initial search (meaning lower starting probability at equal skill) and risk penalties to nightblade combat will still be present. I would be fine with that. Still pressure to move but just not quite as immediate... you can wait to see if this was a single search or they keep going before acting.

2. Scale risk accumulation to the ratio of perception versus stealth ranging from 8 to 12 (and maybe even 16 if something has twice your stealth in perception value). This would reflect the actual "risk" of being caught on a search (how much risk you are at based on how good the person is at finding things compared to how good you are being found) with risk accumulation being slower when someone has less chance of finding you in the first place. It also means that going up against something harder than yourself creates more risk and you take those penalties faster. So the actions that I feel it necessary to take are guided by our comparative skills but I still have to choose to do something quickly in most cases... equal perception versus stealth could be set at 12 risk accumulation as it is now. If the searcher has no perception at all... risk would be set to 8.

3. As you suggested, increase search rt for successive searches so that risk accumulates quickly to start on first few searches but allows more tick down time on later searches... the idea being that as the searcher continues to excessively search without finding you, they have to look harder (since they have exhausted the easy spots) and the stealthed person may begin to feel less overall accumulation of risk in being caught (since this guy has been searching but hasn't found me yet so clearly can't find things that are right in front of him)... and also to encourage the searcher to also take action of they feel they're in danger and someone is stalking around. This does less to solve the issue of immediately being found by low perception people but I do think it's important because, again, the stealth person should not be the only person forced to act quickly when they identify risk... someone searching 20 times in a row when they don't find anything also shouldn't be a possibility.

I would like to see all 3 of these implemented as this would optimize the dynamic and add more pressure on both sides but also make actually training in perception and stealth more valuable on both ends, preserving the idea that training in things actually makes you better at it.
You reach toward ((DEV Rias)) ... Pull(d225([1]x)):214 vs Mark(d1100):714 = -500 (-222%)
You notice ((DEV Rias)) glance your way, causing you to quickly withdraw your hand from his wool
drawstring pouch (open).
Roundtime: 5 seconds.

User avatar
Rias
DEV
Posts: 952
Joined: Sun Sep 03, 2017 4:06 pm

Re: Stealth and Risk

Post by Rias » Wed Feb 17, 2021 10:43 am

Acarin wrote:
Wed Feb 17, 2021 10:33 am
1. As you have suggested, add a starting reroll so the range of potential rerolls is 1 to -4.
This was just added in this morning. Huzzah!
2. Scale risk accumulation to the ratio of perception versus stealth ranging from 8 to 12 (and maybe even 16 if something has twice your stealth in perception value).
Hrm, I do like this idea.
3. As you suggested, increase search rt for successive searches so that risk accumulates quickly to start on first few searches but allows more tick down time on later searches
Yeah, I'm thinking this is a good way to discourage excessive searching and also increase the searcher's own risk (if not Risk with a capital R) because by putting themselves in more roundtime, they're opening themselves to a more opportune ambush. As you said, making them also have to weigh whether it's worth it to keep searching or take some other action to try and prepare for or avoid the suspected impending ambush.

Acarin
Posts: 121
Joined: Fri Sep 04, 2020 10:49 pm

Re: Stealth and Risk

Post by Acarin » Wed Feb 17, 2021 10:49 am

Thanks, Rias. This took a little back and forth but glad we're in more agreement now. I really like cogg (and stealth characters) and just want to see this weighted/optimized so it's fun for everyone and feels balanced. I appreciate all the great work you do and your consideration/testing around this... I know I've been a pain but it's been a productive discussion.
You reach toward ((DEV Rias)) ... Pull(d225([1]x)):214 vs Mark(d1100):714 = -500 (-222%)
You notice ((DEV Rias)) glance your way, causing you to quickly withdraw your hand from his wool
drawstring pouch (open).
Roundtime: 5 seconds.

Post Reply