The Gaming Den

First, you moved the goalposts from your last post; you were talking about how some intervening rolls had to be failed then, and now you're talking about tactics.

Both apply.

If a NPC aims for your arm, your chances of dying due to that hit are essentially nil (There are situations that's not true - but in a duel, they're not going to come up. And there aren't many cliff situations...indeed, there may not be any.)

However, if he aims for your head, you have to fail the defense roll and the soak roll and the vs. death roll.

So blocking an attack is a good idea - if it does hit, it may well (I don't have a clear enough idea on how I want to write the rules for blocking to know) hurt your shield arm instead, and your head is safe (for now).

There's a reason I prefer using the phrase "a chance" to giving an exact number - depending on what happens with a given move, you might have a 0% chance of dying (though you're not likely, to say the least, to know that in advance) or a high chance.

So how does it balance out? Depends on what you do!

Personally, I'm for having a fairly simple system in this regard - having an elaborate set up doesn't add to anything, though I'd like as much detail as possible without sacrificing working out as desired.

And if things do work out as desired, you always have a chance in a given combat that if things go wrong (which you have some control over, regardless of the dice), you can die.

That's the thing. If things DO go wrong (which includes your opponent getting things going very well), you should risk dying. If things -don't- go wrong, you won't.

Having every individual attack or action have a chance of killing you doesn't accomplish that - it just means that you will sooner or latter be dead.

If, on the other hand, dropping your guard because you don't think someone is a threat might give them enough of an advantage (CAN style) that they -are- a threat - you have to be sure to keep your guard up.

As it should be.

Elennsar wrote: There's a reason I prefer using the phrase "a chance" to giving an exact number - depending on what happens with a given move, you might have a 0% chance of dying (though you're not likely, to say the least, to know that in advance) or a high chance.

I just wanted to address this statement. Why would a player not know? If you cannot roll anything that turns sword to arm into I'm dead then, as a player, I'm eventually going to figure that out. It may be through game play and getting swords in my arms a lot. It might be from reading the rulebook to understand how to play the game. Once a player does figure out that sword to arm cannot lead to death due to the mechanics of your game, what then?

What you can't know with absolute certainty is whether or not someone is going for your arm. Not that arm wounds are survivable.

If you say "Oh, he's only going for my arm, I can deal with that." and ignore it - there's a chance (I'm not sure on the mechanics for this) that he'll take advantage of you ignoring his attack and what you thought was going for your arm turns out to be going for your head.

-That- is what makes you need to treat each attack seriously. Sure, the majority of attacks will not be that serious, but you can't afford to treat an attack from a competent (not necessarily "Equally skilled") opponent as being not serious - if it is, you can be in trouble very swiftly.

So if you drop your guard, and pardon me for being punny here, you CAN give your opponent sufficient CAN to screw you.

Maybe you'll get away with it, but the base TN to hit someone, ignoring their defense roll, is fairly easy.

IGTN wrote:Seriously, do what Frank asked. Explain what you do know about probability, so that we can build up from there.

From what Elennsar's posted, it looks like he understands how to calculate probability for single die rolls, even if that "single roll" involves multiple dice (e.g. 3d6). However, he doesn't appear to understand how to calculate probabilities for trials that involve multiple die rolls. He also doesn't seem to understand how to calculate probability over multiple trials (iterative probability) or conditional probability.

Good lord, man. Read a prob-stat book or take a class.

I'll get more specific. I'd recommend Probability for Dummies. And I'm not saying that as an insult, Elennsar. It explains a lot of the basic concepts you're having a problem with in a way that's understandable to the average person with plenty of everyday examples. And if you can't understand these basics and keep spamming the Gambler's fallacy, you're highly unlikely to design good game material, at least for anything involving an RNG.

It is either likely or unlikely that you will get two failures in a row.

I don't believe in the gambler's fallacy, what I beleive is that if you are going to get something 50% of the time, then at some point, heads will occur. Of course, betting on a given roll turning to heads stands a good chance of leaving you broke.

The likelihood of your next two die rolls both meeting a certain condition is equal to the likelyhood of one die roll meeting that condition, squared.

The likelihood of both your last die roll and your next die roll meeting that same condition is either zero (if your last die roll does not meet that condition), or equal to the chance of a single roll meeting that condition (if your last does).

The likelihood of your last two die rolls meeting a given condition is either zero or one, depending on whether or not they do. This part isn't really necessary to state, but it's here for completeness. The point it makes is that, as you go through a trial involving multiple probabilities, the probabilities change.

For example, let's say someone hits you PC with a Phantasmal Killer in D&D. You have the ability to make Fortitude saves in place of Will saves for some reason, and succeed on a rolled 11.

Before you roll your saves, you have a 25% chance of dying.
Once you roll your will save, you have a 0% chance of dying if you make it, or a 50% chance of dying if you failed it.
Once you roll your fortitude save, you either died or your didn't.

So, before you roll, your knowledge is incomplete, and you can say it is unlikely that you will fail both saves. If you blow your will save, though, your knowledge is more complete, and so the probability changes. Probability describes incomplete knowledge.

Also, you didn't describe the gambler's fallacy right. It's exactly what you're doing.

Elennsar wrote:What you can't know with absolute certainty is whether or not someone is going for your arm. Not that arm wounds are survivable.

If you say "Oh, he's only going for my arm, I can deal with that." and ignore it - there's a chance (I'm not sure on the mechanics for this) that he'll take advantage of you ignoring his attack and what you thought was going for your arm turns out to be going for your head.

-That- is what makes you need to treat each attack seriously. Sure, the majority of attacks will not be that serious, but you can't afford to treat an attack from a competent (not necessarily "Equally skilled") opponent as being not serious - if it is, you can be in trouble very swiftly.

So if you drop your guard, and pardon me for being punny here, you CAN give your opponent sufficient CAN to screw you.

Maybe you'll get away with it, but the base TN to hit someone, ignoring their defense roll, is fairly easy.

How exactly do you propose to resolve that? I'm trying to see a way to codify such a thing fairly. Either the player knows you're going after his arm, or he doesn't. If he doesn't know that, and defending doesn't cost anything, then he just defends normally. If he does know that, and there's a cost/benefit analysis to defending and he elects not to, then switching it up just seems...unfair, as you're simply punishing him by changing the deal he agreed to. (He elected to not defend a shot to his arm, not one to his head.)

You could probably have some sort of poker-style system of raises and calls in terms of combat (I slash at his arm!/I leave myself open./I turn my cut towards his head!/Aha! That was a trap! I stab him in the gut!) but that would get bogged down and dull if you had to do that for every attack and every opponent.

I know you have this hard-on for forcing players to take every opponent seriously, but that's just not going to happen. Nor should every opponent be a credible threat. Not everyone with a sword is a swordsman, after all. The PCs, based upon your descriptions however, are. If they're the elite companions of Arthur, it's a little unbelievable if the rank-and-file soldiers of the opposition are their peers. Or did you overstate the PCs stature?

Also, you didn't describe the gambler's fallacy right. It's exactly what you're doing.

Expecting a fair coin to roll heads at some point is the gambler's fallacy?

Either the player knows you're going after his arm, or he doesn't. If he doesn't know that, and defending doesn't cost anything, then he just defends normally. If he does know that, and there's a cost/benefit analysis to defending and he elects not to, then switching it up just seems...unfair, as you're simply punishing him by changing the deal he agreed to. (He elected to not defend a shot to his arm, not one to his head.)

Simple. You know you're being attacked. You can choose to dodge, parry, or block the attack (assuming its a melee attack...you can't parry ranged attacks, and if you don't have a shield or something that can be used as one you can't block).

You might have some idea where the guy is aiming at, but its a guess - you don't know until steel bites into your flesh (if you ignore it) if that's actually where he was aiming.

I know you have this hard-on for forcing players to take every opponent seriously, but that's just not going to happen. Nor should every opponent be a credible threat. Not everyone with a sword is a swordsman, after all. The PCs, based upon your descriptions however, are. If they're the elite companions of Arthur, it's a little unbelievable if the rank-and-file soldiers of the opposition are their peers. Or did you overstate the PCs stature?

I have a "hard-on" for combat actually involving an actual risk of actual pain and/or injury even vs. inferior opponents if you don't act with some level of sense.

I fully agree that not every opponent should be a credible threat, but -ignoring- a trained opponent is not safe.

The rank-and-file soldiers of the opposition are skilled enough that they can get hits. The fact that if you do focus on one, he's going down, and you aren't, doesn't mean that you can just treat him as irrelevant. If you do, he -might- be able to get CAN from that (and make Arthur short one Companion, if he really gets lucky and you really fuck up.).

Being able to take on infinite lesser opponents is neither realistic or dramatic.

On the other hand, I would be very irritated if immitating Leonidas (the historical one) fails - "Infinite" is out, but "a great many' sure isn't.

A rank and file opponent is probably not going to kill you. Assuming he can't is capable of being fatal to your health, assuming he can and treating him accordingly will be fatal to his.

Elennsar wrote:

Either the player knows you're going after his arm, or he doesn't. If he doesn't know that, and defending doesn't cost anything, then he just defends normally. If he does know that, and there's a cost/benefit analysis to defending and he elects not to, then switching it up just seems...unfair, as you're simply punishing him by changing the deal he agreed to. (He elected to not defend a shot to his arm, not one to his head.)

Simple. You know you're being attacked. You can choose to dodge, parry, or block the attack (assuming its a melee attack...you can't parry ranged attacks, and if you don't have a shield or something that can be used as one you can't block).

You might have some idea where the guy is aiming at, but its a guess - you don't know until steel bites into your flesh (if you ignore it) if that's actually where he was aiming.

Then why would you ever elect to not do one of those things? Were you suggesting that someone was foregoing their system-granted defense options just for the hell of it? So my first case it is: the player doesn't know where it's going and has no reason not to defend. What were we arguing about? Oh, that's right, your refusal to cite specific concrete numbers so we can get a fucking idea of how you envision any of this working.

I have a "hard-on" for combat actually involving an actual risk of actual pain and/or injury even vs. inferior opponents if you don't act with some level of sense.

I question what you determine as an appropriate level of sense. You were presenting someone as simply not defending against an attack for no reason earlier.

The rank-and-file soldiers of the opposition are skilled enough that they can get hits. The fact that if you do focus on one, he's going down, and you aren't, doesn't mean that you can just treat him as irrelevant. If you do, he -might- be able to get CAN from that (and make Arthur short one Companion, if he really gets lucky and you really fuck up.).

How lucky does he have to get? How badly do you have to fuck up? I want numbers here. And if the numbers indicate this guy can't hurt me, I'm going to ignore him and mock him.

Being able to take on infinite lesser opponents is neither realistic or dramatic.

Neither is being overly threatened by them.

Then why would you ever elect to not do one of those things? Were you suggesting that someone was foregoing their system-granted defense options just for the hell of it? So my first case it is: the player doesn't know where it's going and has no reason not to defend. What were we arguing about? Oh, that's right, your refusal to cite specific concrete numbers so we can get a fucking idea of how you envision any of this working.

Well, you don't -get- infinite defenses in a round, so you might well feel that risking that it -could- be serious vs. one opponent is okay because you're more worried about the other one.

As for concrete numbers: You want concrete numbers before I -have- concrete numbers. What do you expect? Me to pull random numbers of out of my giant invisible hat and pretend they're a representation of how I'd like to see this work?

I question what you determine as an appropriate level of sense. You were presenting someone as simply not defending against an attack for no reason earlier.

(Over)confidence is a reason. If you expect an attack to be something you can just take on your armor, you might well do that for any number of reasons, some of which are even valid.

How lucky does he have to get? How badly do you have to fuck up? I want numbers here. And if the numbers indicate this guy can't hurt me, I'm going to ignore him and mock him.

And you're not going to -have- numbers until this is actually written out to some extent.

Neither is being overly threatened by them.

What's your idea of "overly"? So far, I don't see any sign you're really okay with the idea that a rank and file guy has -any- chance of killing you without you putting your head on a block for him.

You're going about this the wrong way. Rather than decide what the percent chance of getting killed in a given fight is first, decide what sort of things will be going on in combat. Then make mechanics that allow for it. Then tweek numbers that are too extreme or too puny. Start with this:

Two knights are on foot, fighting with the intent to kill eachother, both with identical kightly "fighting on foot against other people on foot" equipment. Assume both are completely equal in skill, size, strength, tactics, luck and neither are wounded nor tired. Assume that it's just these two fighting and nobody else jumps in or shoots at one one the knights with an arrow. Answer the following three questions:

1. List what the kights should have physically on ther persons during this fight. No stats for it, just "a sword, a helmet, pair of boots, etc."

2. Name the different attacks can they use to try to hurt eachother. As many as you can. Give us a description of each type. Not attack bonuses or mechanics yet, just what the attack looks like.

3. Name the different ways they can defend against getting hurt. As many as you can. Give us a description of each type. Again, no mechanics just yet. That can be done later.

Then do the same for the situation of one of the above knights, still in full health and not tired, but this time the knight is fighting a run of the mill barbarian in single combat on foot.

From there, we'll come up with mechanics which handle both knight on knight violence and knight on barbarian violence.

I can answer some but not all of this - so any gaps should be treated as "will be filled in when possible":

1. List what the kights should have physically on ther persons during this fight. No stats for it, just "a sword, a helmet, pair of boots, etc."

A sword (standard 3' straightsword). A helmet. Pair of boots. Mail hauberk (does not cover lower legs). Shield. Possibly a spear (Eight feet? Standard size, at any rate). Long knife/dagger. Hands protected by leather gloves.

That about covers it.

2. Name the different attacks can they use to try to hurt eachother. As many as you can. Give us a description of each type. Not attack bonuses or mechanics yet, just what the attack looks like.

Offhand:

1) Standard vanilla attack.
2) Flurry of blows (offensive stance only). - defense sacrificed for speed.
3) Precise attack (offense or normal stance only) - speed sacrificed for accuracy.
4) #1 in defensive stance.
5) Shield bash. Yes, you hit someone with your shield. (normal or offensive stance)
6) Unarmed attack.
7) Grapple.
8.) Trip.

3. Name the different ways they can defend against getting hurt. As many as you can. Give us a description of each type. Again, no mechanics just yet. That can be done later.

Dodge - hardest, but if you succeed you're completely unhurt. Can be done without any weapons or shield. Hindered by encumbrance.
Parry - standard. Cannot be used against ranged attacks.
Block - uses a shield, easiest, chance of shield breaking or attack hurting shield arm or both.

Then do the same for the situation of one of the above knights, still in full health and not tired, but this time the knight is fighting a run of the mill barbarian in single combat on foot.

Essentially the same thing, though less armor and more axes among the barbarians.

One thing that I'd like to note is that there won't be much where you're fighting equally armored and identically trained people - this is focusing on the guy who beat the Saxons back for a generation as a source of inspiration, not the guy who slept with his half sister and wore shining plate armor.

Elennsar wrote: 1) Standard vanilla attack.
2) Flurry of blows (offensive stance only). - defense sacrificed for speed.
3) Precise attack (offense or normal stance only) - speed sacrificed for accuracy.
4) #1 in defensive stance.
5) Shield bash. Yes, you hit someone with your shield. (normal or offensive stance)
6) Unarmed attack.
7) Grapple.
8.) Trip.

Dodge - hardest, but if you succeed you're completely unhurt. Can be done without any weapons or shield. Hindered by encumbrance.
Parry - standard. Cannot be used against ranged attacks.
Block - uses a shield, easiest, chance of shield breaking or attack hurting shield arm or both.

Ok, how does the following sound:

Between the two fighters, there is a distance, measured in feet. Every attack and defense has an opitmal distance at which the move which is most effective. Being at a distance different from the optimal distance makes the move less effective. Different weapons and different attacks have different opitmal distances. During combat, you can move closer or further from the enemy.

Example:
There are two fighters, one with a spear and one with a short sword. The opitmal distance of the spearman's attack is 6 feet. The opitmal distance of the swordsmen's attack is 2 feet. Thus, the spearman wishes to remain at a distance of 6 feet from the swordsmen, while the swordsmen wishes to close to 2 feet. At a distance of 4 feet, both are equally disadvantaged.

Offense orrented moves tend to have shorter opitmal distances, while defensive moves tend to have longer opitmal distances. As such, we have the following attacks for a man with a sword:

1) Swing. Range 3, does damage to random limb.
2) Flurry. Range 2, does large amounts of damage.
3) Jab. Range 4, does damage to a specific limb.
4) Swing, but from your chosen defense's optimal distance.
5) Shield Bash. Range 2, hurts enemy.
6) Punch. Range 1, low damage, close range, not very good.
7) Grapple. Range 0, prevents both fighters from moving away from eachother.
8.) Trip. Range 1 Prevents enemy from moving without preventing you from moving.

And his defenses:

Dodge: Range 6. Avoids an attack and moves you one foot towards or away from enemy. Very difficult when close to enemy.
Parry. Range = weapon's range. Avoids damage.
Block. Range = shield's range. Avoids damage.

Wahla. We now have some basic sword fighting tactics, which can be picked apart and used as a basis for other fighting styles.

Elennsar wrote:

Also, you didn't describe the gambler's fallacy right. It's exactly what you're doing.

Expecting a fair coin to roll heads at some point is the gambler's fallacy?

No, but expecting having rolled a 3 on one roll to affect the odds on your next roll is. And you've stated that this is true. Any time you expect a short run of particular results in a series of independent trials to affect upcoming results in any way, it's the gambler's fallacy. If you bet on a fair coin flip to come up heads because you've been getting a lot of heads lately, that's the gambler's fallacy. If you bet on that same coin to come up tails because the run of heads means tails are "due," that's also the gambler's fallacy.

At some point, betting in the first place is a fallacy.

I expect that if I have a given outcome (a 3, say) that it is improbable that I will have two 3s in a row.

Grek: Solid enough basis, I suppose.

How can a bet be a fallacy? A bet is a wager on an outcome - IF P THEN Q...

P might be impossible, or highly unlikely - but the wager is a logical statement.

Also - if you just rolled a 3, why would rolling another 3 be improbable?

Did you read frank's post?

Asking for the next 2 flips to be heads is 25%

Asking if the flip after a head flip is a head is 50%

Expecting that I will roll two 3s in a row is improbable. Expecting I will roll a 3 at all is improbable.

As for betting being a fallacy: You're assuming that an event is going to go a certain way based on guesswork. That's not the wisest thing known to man.

the word fallacy has a definition.

It doesn't apply to unwise decisions.

spasheridan wrote:How can a bet be a fallacy?

The point being made is that the reasoning behind the bet is (based on) a fallacy, not that the bet itself is a fallacy.

And picking anything other than randomness for gambling (for say, slot machines and roulette, not so much for poker and blackjack) is inevitably going to be based on something that is a fallacy of one sort or another.

Of course, you could happen to be right long enough to win (it would still be a fallacy, but you could get away with it).

Elennsar wrote:Expecting that I will roll two 3s in a row is improbable. Expecting I will roll a 3 at all is improbable.

The probability that your next two 3d6 rolls will be 3s is 1/46,656. However, if you've already made roll #1 and it was a 3, the odds that rolls 1-2 will both be 3s is now 1/216 (the odds of making a 3 on one 3d6 roll). Rolling a 3 on 3d6 is highly unlikely, but it is no more unlikely simply because your last roll was a 3. And that's what you said earlier in this thread.

I give up. Why? Because none of this boils down to anything with any meaning that is remotely useful to actual design based on actual play.

Either you're likely to fail (roll low) twice in a row or you're not.

One or the fucking other.

Either a 50-50 chance two times in a row is a 50% chance of surviving the second encounter (if you survive the first, if you don't, you don't have a second encounter to worry about surviving) or it isn't.

For the hundredth time: Probabilities depend on what you already know.

The odds of rolling two threes in a row are low. The odds of rolling a second three, when the first one is already known, is substantially higher (the same as your odds of rolling the first).

If you have a 50-50 chance of surviving an encounter, then the odds of surviving two in a row are 25%. If you already survived the first, then it becomes a known and you don't care about it. It's a 50% chance of surviving the second encounter, conditional on your seeing the second encounter, so a total 25% chance of surviving it before the first encounter.

Elennsar wrote:I give up. Why? Because none of this boils down to anything with any meaning that is remotely useful to actual design based on actual play.

Either you're likely to fail (roll low) twice in a row or you're not.

One or the fucking other.

Either a 50-50 chance two times in a row is a 50% chance of surviving the second encounter (if you survive the first, if you don't, you don't have a second encounter to worry about surviving) or it isn't.

I think you don't like the probability of a loss occuring, in which case you should simply deny chance for those situations.
Someone says they do something and it happens. No risk. No opposition.
I am serious like you wouldn't believe; if a guarantee is what you want, why bother with chance at all?

Elennsar wrote:I give up. Why? Because none of this boils down to anything with any meaning that is remotely useful to actual design based on actual play.

Either you're likely to fail (roll low) twice in a row or you're not.

One or the fucking other.

Either a 50-50 chance two times in a row is a 50% chance of surviving the second encounter (if you survive the first, if you don't, you don't have a second encounter to worry about surviving) or it isn't.

Hm?

If you state that there is a coinflip chance of surviving an encounter, then I can tell you that there is a 1 in 2 chance of surviving any given encounter, and that in order to survive a certain number of encounters you are basically going to be taking (1 in 2) to n, where n is the number of encounters. So yes, you are likely to fail two encounters in a row, mostly because you were likely to fail one encounter and you keep having these highly hazardous events occur. You are not, however, any more likely to fail any one encounter as an isolated event just by having multiples occur over time, assuming all else being equal. It's the set which is unlikely. No single pearl is terribly unlikely but when placed upon a string with other pearls you have a rarity.

If you add in ways to manipulate the results of the dice, however, which is fairly common in gaming systems, the probabilities are not as straightforward as a direct transposition from the dice. Usually what happens is the chance of failure drops dramatically on anything that can and will have odds-improving resources spent on it, such as survival. So if I can take a coinflip I dislike and do it again, I have to fail that particular check twice in a row, for a 25% chance of actual failure, or a 75% probability of survival. Even though the probability of failure is initially high, because of the resources, it drops sharply. I can then take this and raise it to a power to discover the odds of surviving multiple encounters on this basis.

However, note, that here is the rub. If I have a system which works like this in any manner, I am encouraged, if I want to survive, to have this resource available every time I make a survival check. This is why you might as well provide the resource either virtually always or effectively never in many scenarios. If it's a cute benefit you can never guarantee like picking up the Quad Damage, then people don't care much about it, but if it's something you can reliably get, especially by, say, sleeping 8 hours, you now will be observing quite a few people patterning themselves around that so they have the resource always.