Talk: Concealment

Odd percentages

25% with Blind Fighting? That's not how probability works, is it? If you get a second roll with the same chance of missing (50%), the total chance of hitting is still 50%. You get two 1 in 2 chances of hitting. Together, that makes a 2 in 4 chance of hitting. 50%. 84.28.178.18 12:27, 3 May 2009 (UTC)

• No you get a reroll only if you fail the first roll. Thus to succeed you have the initial 50% success and then 50% of the remaining 50% for when the reroll succeeds. 50% + 25% = 75% success and thus 25% failure. Or you can see it as to fail with blind fight both the first roll and the second roll must fail. Thus it is 50% failure * 50% failure = 25% failure. WhiZard 17:00, 3 May 2009 (UTC)

• Another way to look at it is that you do not get a 2 in 4 chance of hitting. The 4 possible outcomes are hit-hit, hit-miss, miss-hit, and miss-miss, and you only miss if both outcomes are misses. That is, there are 3 of 4 outcomes that involve hitting. --The Krit 22:28, 17 May 2009 (UTC)

Each attack, not just hits

"Concealment is a chance that your opponent will miss you when he would normally hit otherwise."

According my tests, this is not true, concealment works for every single attack, no matter if that attack would hit or miss. (I tested this with opponent with 70AC + 50% conc. and my char have ab 20, so he could hit only on 20, however I saw in log "miss due to concealment" with all possible values. I would update the page myself, but Im not sure how to formulate it. ShaDoOoW 20:46, September 20, 2010 (UTC)

• Right. This article was not written completely accurately back in 2005, and no one had gotten around to cleaning it up. (Not that it makes any effective difference to tactics — the number of hits is the same regardless of whether or not concealment is rolled for misses.) --The Krit 19:26, September 24, 2010 (UTC)

• Well im not really sure but... if it would work only for hits, the concealment would be more powerfull for those with high AC value, as opponent must first hit - lets say he must roll 19 or 20 and then there would be 50% chance to miss anyway. But concealment is rolled before, so no matter whats the character AC, it do not gives advantage to anyone. ShaDoOoW 21:05, September 24, 2010 (UTC)

• It shouldn't make any difference. In order to hit, both the attack roll and the concealment roll have to pass. If the concealment chance to hit was 30% and the attack role chance to hit was 40% then in order to hit the chance would be 0.3 * 0.4 if concealment is rolled first and 0 .4 *0 .3 if attack were rolled first. In either case the chance to hit would be 0.12. Dremble 21:11, September 24, 2010 (UTC)

• If 80% of attacks will miss because of AC and 50% because of concealment, then depending on the order in which you do things, you get:

1. AC then concealment: 80% of attacks miss because of AC, then 50% of the remaining 20% miss because of concealment, giving 10% missing because of concealment. Overall, 80% + 10% = 90% of attacks miss.
2. Concealment then AC: 50% of attacks miss because of concealment, then 80% of the remaining 50% miss because of AC, giving 40% missing because of AC. Overall, 50% + 40% = 90% of attacks miss.
3. Both at the same time (roll both and the hit is scored only if both rolls succeed, which is what the game does): 80% of attacks miss because AC and 50% miss because of concealment, but that double counts the attacks that miss because of both. Specifically, 80% × 50% = 40% miss because of both AC and concealment. Overall, 80% + 50% - 40% = 90% of attacks miss.

It works out exactly the same, no matter what order you apply these things. The only thing that changes is what you credit for causing the miss, but that has no effect on the battle's outcome. --The Krit 23:36, September 24, 2010 (UTC)

• Another way to look at it (not that there was anything wrong with Dremble's simple approach): Suppose someone had cooked up the formula

AC + (AB+20 - AC)*C%

for calculating, given C% concealment, what is basically an AC rating that is an equivalent defense with no concealment, assuming concealment is only rolled for misses. (Someone might rearrange terms and write this as "AC+(20-(AC-AB))*C% = conceal adjusted AC".) This formula counts misses due to AC (more or less — there is some normalization going on here), then successful attack rolls that miss due to concealment. The same calculation assuming that concealment is rolled first would be

(AB+20)*C% + AC*(1-C%)

which counts misses due to concealment, then failed attack rolls that passed concealment first (where both terms have the same sort of normalization as above, counting lower-than-possible attack rolls as misses so that the end result is an AC rating). These are just two forms of the same formula:

(AB+20)*C% + AC*(1-C%) =

(AB+20)*C% + AC - AC*C% =

AC + (AB+20)*C% - AC*C% =

AC + (AB+20 - AC)*C%

So while someone might think the first formula needs to be reworked because concealment is rolled for misses, the math does not support the need for reworking it. --The Krit 16:10, September 25, 2010 (UTC)

• Very interesting. So basically the concealment works better for those with high AC, correct? ShaDoOoW 14:25, September 26, 2010 (UTC)

• No. --The Krit 05:59, September 27, 2010 (UTC)

Concealment vs. AC

During a recent dialog about game mechanics, specifically concealment, an experienced long-time player insisted that concealment is not worth much unless the AC is high enough to win attack rolls. The article clearly states that there are 2 separate rolls per attack that must both succeed for a hit to land. According to shadooow's testing, natural 20's are not always Autohits, not when concealment is a factor. Seems to contradict the related statement in the Attack roll article. The player's claim seems to based on misinformation, possibly based on in-game observation rather than controlled testing. Am trying to dredge up some reason for his assumption. Any clues? I suspect many players do not realize that more than one roll can be involved in an attack. I was one of those until reading the article. :P

Has the calculation of concealment (with respect to the "pairing" with the attack roll) changed from its inception (version when NWN was first released)? (I haven't had the time to examine every patch so am hoping you folks could save me the research effort.)

"Concealment rolls are not themselves displayed on the screen, but if the roll fails, the attacker is told the attack failed because of concealment and is told the amount of concealment."

I am assuming this is referring to the attack roll rather than the concealment roll, but not positive. --Iconclast (talk) 12:43, July 22, 2014 (UTC)

• A post way back by Shia Luck detailed that the order of consideration is concealment -> parry -> AC -> epic dodge -> auto-miss (on 1). As far as I know concealment/miss chance has always been first in consideration. I am not sure where auto-hit on 20 falls; it is definitely after concealment (that is 100% concealment will block all attacks), but if it is at the end (with auto-fail on 1) then a low AB attacker may not be dodged by a high AC epic dodge on their automatic 20s. The details of the concealment roll are only shown when concealment blocks an attack; otherwise the attack roll information follows. WhiZard (talk) 15:46, July 22, 2014 (UTC)

• I've looked at and tested the order it is concealment/miss chance -> parry -> AC -> deflect arrows -> auto-miss/auto-hit -> epic dodge. Slightly different from Shia's order. WhiZard (talk) 23:04, July 22, 2014 (UTC)

• Thanks for your efforts to clarify this, WhiZard. It's surprising to me that the autohit/miss gets resolved so late in the sequence. I'll have to begin networking this info. Many players will be jolted methinks. --Iconclast (talk) 01:33, July 23, 2014 (UTC)

• Why does the order make a difference? If you miss due to concealment, then get a natural 20 attack roll, you miss. If you get a natural 20 attack roll, then miss due to concealment, you miss. Either way, you miss. The order of checks really only matters when it comes to "once per round" things like deflect arrows. (Deflect arrows and epic dodge (almost) only fire for hits, presumably to avoid wasting the feat for the round. If they are checked before auto-hit/miss, then you could still waste the feat on an auto-miss and fail to use the feat on an auto-hit.) --The Krit (talk) 15:24, August 16, 2014 (UTC)

• Regarding the player: Concealment is not worth much if the AC is high enough to win (all) attack rolls. Let's look at the extreme cases. If your opponents are going to miss 95% of the time because of AC, how much value would you attribute to being hit less often, say upping that to 97.5% (through 50% concealment)? On the other hand, if your opponents are going to miss 5% of the time because of (no) AC, how much value would you attribute to being hit less often, say upping that to 52.5%? In the former case, 2.5% of strikes are impacted, while in the latter case 47.5% are.

Regarding the quote: In this context, "the roll" is being used basically as a pronoun, meaning that it has an antecedent. Look for potential antecedents (step backwards through the text until you reach a noun). In the same sentence as "the roll", there are two candidates: "screen" and "Concealment rolls", only one of which matches "the roll". ;) --The Krit (talk) 15:24, August 16, 2014 (UTC)