Ok whats the deal with critical hits?

DKnight

Ive been trying to figure out what/how you get more critical hits in game. other than putting points into single weapon prof and two-handed prof. It might make sense that I've figured it out unless someone can prove me wrong. So far I have based this on imoen with a bow while my thief is dual wielding daggers. I have 18 thaco while imoen has 14. She not only gets criticals all the time (against low lvl creatures), she also seems to have a much better shot at getting criticals than my thief.

I have come to the conclusion that just like the eschalon series, critical hits are somewhat mired/related to your to hit rate/thaco of said character. This means that you would likely not get as many crits against sarevok (unless you are a much higher lvl with much higher thaco than possible). Its more or less that you can outclass enemies with low armor class and thus get more crits.

Can anyone confirm this as I just had a really cool idea for a thief character that wields a two-handed staff/two handed prof.

Tresset

I think Imoen just got lucky. Also helmets prevent critical hits.

DKnight

I don't know, it seems like ever since she got the bracers of archery she has been a crit hit master. She is only lvl 4 but is deadly with a bow. There's got to be some formula for how she is so good at criting.

Lord_Tansheron

Hits are not tied to your hit rate, not even statistically. It's a single to-hit roll, not a roll for hit, then a roll for crit like in some other games.

The chance is a static 5% (1 in 20). It's modified only by very few abilities (the Critical Strike HLA, and 1h/2h specialization). Note that a critical hit (a "natural" 20 roll) will also *always* be a hit; this matters if, for example, the enemy is immune to critical hits - they will not take increased damage, but it will still be a guaranteed hit.

Your luck with Imoen is just a statistical fluke. Sample sizes during play are fairly low, so clusters happen every now and then.

Schneidend

You get a critical hit on an attack roll of 20. Single Weapon Style and Two-Handed Weapon Style allow you to get a crit on 19 and 20. End of line.

DKnight

So with single/two handed it goes from 5% to 10% chance per hit?

CaptRory

The more attacks you make the better your chances as well. If you're attacking three times a round you have 3 chances to roll that natural 20.

Lord_Tansheron

DKnight said:

So with single/two handed it goes from 5% to 10% chance per hit?

Yes. It will give you a crit on either a 19 or 20, so 2 in 20 = 10%.

Schneidend

@DKnight
Yes. Those fighting styles double your chance per roll. And, as CaptRory points out, more attacks means more rolls which means more chances to roll those crits. I'm not sure exactly how much having more attacks increases that 10% chance overall, though. Each attack would still individually only have a 10% chance, but taken as a whole it is more likely that at least one crit will be rolled in that set of attacks.

FinneousPJ

Wait, there's no critical hit confirmation roll in BG?

Lord_Tansheron

Only one roll, on a d20. If it comes up 1, it will *always* miss, if it comes up 20 (or 19 with 1h/2h spec) it will *always* hit, and be a critical strike if the target isn't immune. The in-betweens depend on your THAC0 and the enemy AC, but the two extremes do not.

Awong124

FinneousPJ said:

Wait, there's no critical hit confirmation roll in BG?

I think that only started in 3E.

FinneousPJ

Oh, OK. Never noticed the lack of it before.

ajwz

Critical confirmation was a sensible mechanic to introduce. It was also less fun

CaptRory

Anything that makes combat more dangerous ultimately hurts the players. So adding the confirmation roll for the damage multiplier was probably a good thing.

ajwz

I agree, although actually I have had a lot of bad experiences with DMs not introducing some sensible form of applied critical confirmation to circumstances out side of combat.

Such as forcing all 6 members of a party to make two seperate balance rolls to scale along a cliff edge, and making the assumption that a 1 is an auto fall.

CaptRory

The 1 being an autofail thing is actually a house rule that just stuck around for so long everyone assumes it is a real rule.

atcDave

CaptRory said:

The 1 being an autofail thing is actually a house rule that just stuck around for so long everyone assumes it is a real rule.

Actually not quite true. P. 161 of the 2E Players Handbook under "Impossible To Hit Roles" says that a 20 always hits, and a 1 always misses. The "critical hit" role of a natural 20 is an optional rule in 2E.

In 1E, I don't believe there were officially any automatic hits or misses or criticals. But every DM I ever played with used some sort of house rule, usually for all three of those things.

So you're right in saying it started as an unofficial but common "house rule" in 1E. But when 2E came around it was made official.

Lord_Tansheron

I think he was referring to 1 = autofail being used for rolls other than to-hit rolls.

atcDave

Lord_Tansheron said:

I think he was referring to 1 = autofail being used for rolls other than to-hit rolls.

That also is a part of 2E rules, but I believe was not official in 1E.

franco

ajwz said:

I agree, although actually I have had a lot of bad experiences with DMs not introducing some sensible form of applied critical confirmation to circumstances out side of combat.

Such as forcing all 6 members of a party to make two seperate balance rolls to scale along a cliff edge, and making the assumption that a 1 is an auto fall.

When I DM'd, a critical miss along a cliff edge meant you were starting to fall over. Then I would have them make another normal roll (probably on DEX) to determine if they were able to grab on to something or get a foothold. This second roll was pretty much my confirmation roll, and it saved an awful lot of adventurers, yet the feeling of danger was clearly there.

Lindeblom

My first ever AD&D character, Schnawitze the dwarf fighter was killed by a critical hit by a mighty KOBOLD some 29 years ago........how i loathed that rule back then

franco

Schneidend said:

@DKnight
Yes. Those fighting styles double your chance per roll. And, as CaptRory points out, more attacks means more rolls which means more chances to roll those crits. I'm not sure exactly how much having more attacks increases that 10% chance overall, though. Each attack would still individually only have a 10% chance, but taken as a whole it is more likely that at least one crit will be rolled in that set of attacks.

Technically with the single weapon/two handed fighting styles the critical hit chance goes up to 10% on each attack. Mathematically, then, if there are three attacks per round the crit chance rises to 30%. That's powerful stuff.

FinneousPJ

Actually it doesn't rise to 30 %. This is intuitively clear, because with 20 attacks you are not guaranteed a crit. The chance of critting at least once with SWS over three attacks is 27.1 %.

franco

I don't believe that's the way it's calculated. Of course there's no guarantee you will get a crit in 20 attacks. This is mathematical statistics. The probability of critting on one swing is 1/10. In the same round you get three swings, so the probability per round is 1/10 + 1/10 + 1/10 = 3/10 (30%). Unless, if you hit on the first swing you don't get the other two.

Also, as a matter of interest, the chance of not critting on 20 consecutive attacks is: (9/10)^20 =0.12577
or about 12.6%.

FinneousPJ

Exactly, and similarly, the chance of not critting in 3 consecutive attacks is (9/10)^3 = 0.729. Hence, the chance of critting (at least once) is 1 - 0.729 = 0.271;)

Cardinal_Thor

franco said:

The probability of critting on one swing is 1/10. In the same round you get three swings, so the probability per round is 1/10 + 1/10 + 1/10 = 3/10 (30%). Unless, if you hit on the first swing you don't get the other two.

If you follow that logic, then getting 10 swings in one round would give you a 100% chance of getting a critical hit. This is obviously false.

Consider a simplified example of a coin flip. If heads is a critical and tails is not, then on a given "roll" you have a 50% chance of getting a critical. If you "roll" twice, you don't add 50% and 50% to come up with a 100% chance.

Your chance of rolling any given number on a die, or getting heads on a coin flip, is the same whether you do it a million times or once. Your odds do not increase based on the number of die rolls.

-CT

FinneousPJ

Yes, this is the point I'm trying to make.

franco

@FinneousPJ. Thanks for the critical parry, and @Cardinal_Thor thank you your holiness for the great explanation. Lassen mich denken ein bisschen.

Cardinal_Thor

I'm tracking, FinneousPJ. Math.

CaptRory

You have three different 10% chances of rolling a 19 or 20. You could roll three crits or none or anything in between. In practice, the more times you try to get a given result, the better the odds of getting that result. If you roll a d20 1,000 times each one has a 10% chance of getting that critical hit. Technically you could go your entire life and never roll a 20 on a d20. Its also possible you could roll nothing but 20's on a d20. The odds of that happening are roughly the same of turning into vapor and reappearing on the other side of a brick wall.

Anyway, the more attacks you throw, the better your chance of getting your results. This is because, if the system is unbiased, with a large enough sample, everything will come out even. If you roll a d20 20,000,000 times you're probably going to see an almost equal distribution of every number.

In smaller sample sizes you're more likely to see exaggerations. Like, the guy playing a paladin that rolled three 1's in a row and managed to kill himself and the Big Bad Evil Guy he was fighting by himself.

All this stuff is extremely hard to wrap your head around and frequently comes across as counter intuitive. No wonder everyone believes in luck and superstitions. It makes more sense!