Healing surges and other such fail.

[JonSetanta]

What chance should the heroes have of actually finishing the campaign?

75%

[Absentminded_Wizard]

What chance should the heroes have of actually finishing the campaign?

75%

Apparently, it's greater than 0 and less than 1. Which really narrows it down.

[Draco_Argentum]

You being insulting doesn't make your arguement more believable or well founded, you know.

You being an ignorant, rude jerk doesn't make him wrong. You are a fucking pest. How about you go learn high school maths then come talk with us. You should also sue your high school maths teacher, assuming you've graduated.

[violence in the media]

You being insulting doesn't make your arguement more believable or well founded, you know.

I'm sorry, I just had the mental image of you stamping your foot while making that last statement.

If you seriously cannot grasp what people are talking about, that's ok. Own up to the fact that you don't get it and move on. Quit trying to argue that your inferior grasp of probabilities is somehow the correct one and that math is wrong. I know we all play in magical worlds here, but that does not mean gaming lives in one where math doesn't apply.

[norms29]

You being insulting doesn't make your arguement more believable or well founded, you know.

You being an ignorant, rude jerk doesn't make him wrong. You are a fucking pest. How about you go learn high school maths then come talk with us. You should also sue your high school maths teacher, assuming you've graduated.

High school nothing, this is stuff I could've followed in the 5th grade.

[Parthenon]

It does contradict. Either I have a 50% chance of winning in both of the encounters or I don't.

I registered just to help argue this point. You should be proud of that.

(there were lots of other reasons, I just never got around to it)


Lets say we're playing a completely stupid RPG. In it, all the players and the GM describe whats happening in turn, then the GM flips a coin. If its heads, then the players win the encounter, and if its tails, then they lose. With me so far?

Okay, so, the chances of winning an encounter are 50%. The chances of winning two encounters is 25% (50% * 50%).

But lets say you win one encounter. Whats the chances of winning the next encounter? According to Elennsar maths, 25%, because thats the chance of winning two encounters. So, at this point, you have a 25% chance of getting a heads and a 75% chance of getting a tails.

Anything seem wrong with that? Or is the cognitive dissonance giving you a headache so you'll whine that I'm being insulting so my argument is unfounded?


The thing is, for any single flip there is always 50% chance of getting a heads. However, for multiple flips, conditional probability affects it so that the chances of you getting heads every time is less. So if you want to do something that requires getting all heads, then the more times you flip, the less likely this is to happen even though the chances for you getting a head on each individual flip is always the same.

The thing is, for any single encounter there is always an average chance of surviving. However, for multiple enounters, conditional probability affects it so that the chances of you surviving every time is less. So if you want to live to the end of the campaign, then the more encounters you have, the less likely this is to happen even though the chances for you surviving on each individual encounter is always the same.

[Tshern]

Parthenon, your heroic act made my day.

[RandomCasualty2]

I understand how probability works. What I don't understand is why -after winning one fight-, winning another is any more or less challenging, or why we assume the probability of rolling above a 12 (or under, for that matter) will be reflected reliably in a given set of dice rolls.

Winning a second fight is more unlikely because to get to the second fight, you must have survived the first one. A series of corpses automatically lose fight #2. So indeed to survive fight #2, you must have survived fight #1.

So if you die half the time (50%) in each fight. You'll die 50% of the time in fight #1, and then from the other 50% that survive, only half the time will they survive fight #2. And half of 50 is 25, so we're left with a 25% chance of survival.

And yes, 25% chance means that it is technically possible. So your argument that "probability may not be reflected" is nothing that anyone is denying, and this is why I constantly say you don't understand probability. You keep claiming that we're saying that because there's a 25% chance that it can't happen. If you understood probability you'd understand that we weren't saying that.

But the point is that the continuation of your game is entirely based on your PCs have amazing luck. It's not that it can't happen, but it's very likely that any campaign you run will go for like 3-4 battles and then end abruptly when the PCs are slaughtered.

[Roy]

[Elennsar]

You being an ignorant, rude jerk doesn't make him wrong. You are a fucking pest. How about you go learn high school maths then come talk with us. You should also sue your high school maths teacher, assuming you've graduated.

You should actually show something resembling civility. And Marxism should work. The likelyhood of the latter is still higher than the former.

I'm sorry, I just had the mental image of you stamping your foot while making that last statement.

No, I am just tired of you thinking that being insulting will somehow surmount any and all of my objections to your comments and I will bow before Math, which will always explain everything, and no campaign will ever have improbable consequences.

But lets say you win one encounter. Whats the chances of winning the next encounter? According to Elennsar maths, 25%, because thats the chance of winning two encounters. So, at this point, you have a 25% chance of getting a heads and a 75% chance of getting a tails.

No, according to the people who are treating "the odds of winning both" as "the odds of winning the second". Which is not including me.

Anything seem wrong with that? Or is the cognitive dissonance giving you a headache so you'll whine that I'm being insulting so my argument is unfounded?

Being insulting =/= having a bad arguement. It just is being offensive for no reason other than to be insulting, and it does nothing to make your arguement -better-.

So if you want to live to the end of the campaign, then the more encounters you have...

Is there some reason people assume that 50-50 odds of winning = 50-50 odds of (not) dying?

No, seriously. Is there a requirement that gamers never retreat, never surrender?

Now, as to the main point:

So if you want to live to the end of the campaign, then the more encounters you have, the less likely this is to happen even though the chances for you surviving on each individual encounter is always the same.

And if this was stated from the begining, instead of making it as if the odds of rolling badly shot up for encounter #2, #3, etc, etc. we wouldn't have this arguement.

And yes, 25% chance means that it is technically possible. So your argument that "probability may not be reflected" is nothing that anyone is denying, and this is why I constantly say you don't understand probability. You keep claiming that we're saying that because there's a 25% chance that it can't happen. If you understood probability you'd understand that we weren't saying that.

If you weren't arguing that 25, 12.5, 6.25, 3.125, etc. ensured that it would go this way, and that no one COULD win a string of such encounters (and therefore assuming it is possible shows no understanding of probability), then I wouldn't be refering to it not necessarily being reflected.

People do get lucky. Not often. But it happens. Saying "yeah, but its really, really unlikely" isn't arguing that at all...saying "expecting it to be possible shows you don't get probability." is.

It's not that it can't happen, but it's very likely that any campaign you run will go for like 3-4 battles and then end abruptly when the PCs are slaughtered.

Again with the idea that losing and dying are linked. How morbid.

[Roy]

Last I checked, surrender was worse than death. If an equipment dependent game, you lost all that equipment which means at the least you are no longer level appropriate and therefore might as well retire and reroll now because you're gimped forever and more likely have also just lost more than dying and coming back, which would allow you to keep playing. Regardless of whether it is an equipment dependent game or not... you rot in jail. Or you become a sacrifice. Or whatever. Such that you are dead or effectively dead anyways, but don't at least get a good death, or a chance to live intact as you would if you just fought.

I dunno why I'm actually taking you seriously here, but eh.

[Elennsar]

And of course, you are unable to recover your equipment (or get new equipment), and you are unable to get out of jail or whatever.

And meanwhile, retreating will do Xd6 damage to all your items, where X is equal to your bust size.

Or something.

[norms29]

So if you want to live to the end of the campaign, then the more encounters you have, the less likely this is to happen even though the chances for you surviving on each individual encounter is always the same.

And if this was stated from the begining, instead of making it as if the odds of rolling badly shot up for encounter #2, #3, etc, etc. we wouldn't have this arguement.

this was stated from the begining, anyone who had even passing familiarity with Probability and Statistics would have seen no ambiguity in the statements being made here

And yes, 25% chance means that it is technically possible. So your argument that "probability may not be reflected" is nothing that anyone is denying, and this is why I constantly say you don't understand probability. You keep claiming that we're saying that because there's a 25% chance that it can't happen. If you understood probability you'd understand that we weren't saying that.

If you weren't arguing that 25, 12.5, 6.25, 3.125, etc. ensured that it would go this way, and that no one COULD win a string of such encounters (and therefore assuming it is possible shows no understanding of probability), then I wouldn't be refering to it not necessarily being reflected.

People do get lucky. Not often. But it happens. Saying "yeah, but its really, really unlikely" isn't arguing that at all...saying "expecting it to be possible shows you don't get probability." is.

The thing is no one here considers the highly improbable relevent to designing a game that appreciable numbers of people will play. If only one in 2^260 ends in a way other then a TPK then the game offically sucks. no one here wants to make or play I Wanna Be The Guy: The movie: The Game: The RPG

[Elennsar]

Silence may be a speech and other such metaphors, but online, it is extremely unclear.

So if you have a point regarding me thinking people treat retreating or surrendering (in general - if it was with D&D's specific assumptions it would be a bit more reasonable to treat surrendering as worse) in ludicrious ways, I'd like to hear it.

"I want to avoid dying, but I don't want to do anything cautious to avoid it, I want to be able to win by sheer badassery and audacity, even when that wouldn't work."

[norms29]

that's genre dependant, in heroic fantasy (which I recall was the orignal focus of the thread) winning by sheer badassery and audacity are par for the course.

[Roy]

And of course, you are unable to recover your equipment (or get new equipment), and you are unable to get out of jail or whatever.

And meanwhile, retreating will do Xd6 damage to all your items, where X is equal to your bust size.

Or something.

Anything that has taken your equipment has done so because they owned you even with it. Since most of your effectiveness is gear based, and you clearly could not win with your gear, what in the blue fuck makes you think you can counter StarTropics death by beating them without it, to get it back? Exactly. So you're fucked. Also, to get more gear, you have to kill things first. See above, not to mention the fact you're permanently behind by an increasingly large margin, such that the next few months of play are made of Suck and Fail, instead of just rerolling and joining in again this session or the next.

Retreating doesn't work because enemies are faster (pre teleport) and because enemies will reinforce and become even harder to overrun (pre and post teleport). Since you clearly could not beat them before, you certainly cannot now. Also, enemies will hunt you down and finish the job (post teleport) so the fact you ran away, at most just means picking another battlefield. Small condolence seeing as they're running their Countdowns of Awesome, and you likely are depleted, or else you would not have had to run away.

Nice straw man, but you fail. Again.

[Elennsar]

Some (<100%) of the time. Not always.

There are times heroes hold back, or wait until a better chance, or similar - not just leap out at the first possible second.

Roy: Nice example of how D&D design is utterly and irredeemably fucked up.

It actively punishes trying to overcome challenges as distinct from the easiest thing that is still worth xp and gives more gold than it costs to beat.

I'm not sure how that even works for dungeon crawl, let alone for heroic fantasy.

[Roy]

Just to clarify: In StarTropics, if you die you restart with three hearts, no matter where you are in the game or how many hearts you're actually supposed to have. Seeing as the maximum is 20 or 22 depending on which installment in the series it is, enemies do progressively more damage the further you go, and your offensive power is tied to your life, such that three hearts gives you the weakest weapons in the game even if you've found better by now the end result is that if you die, or lose, or whatever you're doomed to a cycle of losing until you get fed up and reset because you can't deal with enemies that take forever to die, but 1-2 shot you. Which is exactly how capture scenarios go. Even if it's not a gear dependent game, you're still in a prison, and still don't have any real weapons. So barring subversions like the Worst Prisons Ever that pop up in... every video game that has prisons you're fucked. End of Story.

So basically, you either ruin your campaign by making it into a living joke like those video game dungeons, or you ruin your campaign by crippling the party such that they say fuck you, reroll time bitch. How about we go with Let's Not Fucking Ruin The Whole Fucking Campaign, you Fucking Fucker eh? Yes, I am saying Fuck a lot. Fuck. Fuckity Fuck Fuck.

[Parthenon]

Notice how I specifically mentioned an average chance of surviving, not winning. This ignores the issue you have of insisting that we're ignoring the times when the PCs neither win or die.

I know that it can be difficult to differentiate between what different posters write and details like that, but even if you have say a 99% chance of not dying an encounter, the chances of not dying in a campaign is a lot less than that.

---

In fact, lets try that. Lets say we're trying to create an RPG commercially. We want to have one million people playing it. This means about 200,000 groups. They each try to play one campaign to get an idea of whether they really like it. This means that 200,000 campaigns are played. Can someone correct me on the no doubt horrendous mistakes I'm about to commit for the rest of this post?

Now this game is very lethal, and for each encounter has about a 60% chance of winning a battle, 20% chance to force the PCs to surrender, be knocked out or retreat, and 20% chance to die. These probabilities are averaged among the different encounters and chances of having the encounters.

(These numbers are completely made up and I don't know yet whether they'll help my point or not. They are made up, say, 1/3 of the encounters have a 80% chance to win, 10% chance to lose and a 10% chance to die, and 2/3 of the encounters having a 40% chance to win, 30% chance to lose but survive and 30% chance to die. However, about half the difficult encounters can be avoided or made into easy encounters by stealth, tactics, diplomacy or puzzles. They are just an example.)

So, lets say the group has to win 10 encounters to win the campaign, but if they lose more than 5 they lose the campaign.

So, looking positively, they have up to 15 encounters they can have.

The probability of winning the campaign should be

sigma( 0.6^10*0.2^x ) where x ranges from 0 to 5.

That right? I think it is.
So, its 0.0060466176 + 0.00120932352 + 0.000241864704 + 4.83729408 × 10^-5 + 9.67458816 × 10^-6 + 1.93491763 × 10^-6.

That adds up to 0.00755778827. As in a 0.7% chance of winning the campaign. That number suddenly looks very, very small. I'm really worried I'm wrong. Let me know if right please. Pretty please?

Okay, so assuming I'm right, (if not I can swap around numbers), out of the 200,000 campaigns played, Google calculates the prediction that about 1,500 campaigns will win. So, out of the 200,000 groups, probability predicts that only about 3 in 4,000 will successfully see the end of the campaign.

Lets look at another prediction: how many don't manage to get to halfway through the campaign.
This is 1-(0.6^5 + 0.6^5*0.2 + 0.6^5*0.2^2 * ... 0.6^5*0.2^5). As in 1 minus the probability of getting halfway.

Google calculates this as about 181,000 people. As in 9 out of 10 people are unlikely to get halfway through the campaign.


Now, this is where it gets important. Out of the 800,000 players that could continue to play, if 720,000 die less than halfway through the first campaign, they are likely to give up and not continue playing. So, if you have a lethal game with 20% chance of failure without death and 20% chance of death, then about 90% of your market will give up and ignore your game.

This is with only a very short campaign. D&D is supposed to last a lot longer than 10-15 encounters. A lot, lot longer.

---

A high lethality does not mean that you, individually are not going to survive the campaign. It means that out of the large number of people that will play it, the probability is that not enough of them will survive the campaign. If not enough of them survive, then it is not fun for them.

A high lethality means that the game is not fun for a large number of people. This means that it is not a fun RPG in general. We (I'm assuming this point for everyone else on the board) want a generally fun RPG. (If you want a fun RPG just for you, then, well, fuck you).

[Absentminded_Wizard]

I'm sorry, I just had the mental image of you stamping your foot while making that last statement.

No, I am just tired of you thinking that being insulting will somehow surmount any and all of my objections to your comments and I will bow before Math, which will always explain everything, and no campaign will ever have improbable consequences.

Well, repeated explanations of your errors and the holes in your reasoning don't work, so some people feel that being insulting is the only alternative.

But lets say you win one encounter. Whats the chances of winning the next encounter? According to Elennsar maths, 25%, because thats the chance of winning two encounters. So, at this point, you have a 25% chance of getting a heads and a 75% chance of getting a tails.

No, according to the people who are treating "the odds of winning both" as "the odds of winning the second". Which is not including me.

Except that nobody ever actually said that but you.

So if you want to live to the end of the campaign, then the more encounters you have...

Is there some reason people assume that 50-50 odds of winning = 50-50 odds of (not) dying?

No, seriously. Is there a requirement that gamers never retreat, never surrender?

Well, a lot of posters here are hardwired to assume standard D&D assumptions. But even if you try to design a game where surrender isn't an extremely unpalatable option, good luck trying to sell it to the average gamer. My suspicion is that you'll have an easier time convincing most gamers to avoid some combats than getting them to surrender.

Now, as to the main point:

So if you want to live to the end of the campaign, then the more encounters you have, the less likely this is to happen even though the chances for you surviving on each individual encounter is always the same.

And if this was stated from the begining, instead of making it as if the odds of rolling badly shot up for encounter #2, #3, etc, etc. we wouldn't have this arguement.

Again, you're the only one who ever said that. It's a total mischaracterization of what anybody has ever said here.

And yes, 25% chance means that it is technically possible. So your argument that "probability may not be reflected" is nothing that anyone is denying, and this is why I constantly say you don't understand probability. You keep claiming that we're saying that because there's a 25% chance that it can't happen. If you understood probability you'd understand that we weren't saying that.

If you weren't arguing that 25, 12.5, 6.25, 3.125, etc. ensured that it would go this way, and that no one COULD win a string of such encounters (and therefore assuming it is possible shows no understanding of probability), then I wouldn't be refering to it not necessarily being reflected.

People do get lucky. Not often. But it happens. Saying "yeah, but its really, really unlikely" isn't arguing that at all...saying "expecting it to be possible shows you don't get probability." is.

The problem is that probability calculations give you the average number of battles before "party loss." The fact that it's an average means that you are just as likely to get worse results or better results. You keep making the argument that "it's possible" for the PCs to do better. True, but it's also equally possible for them to do worse. If the probability of winning one combat is 50%, it's pretty easy for the PCs to lose a bunch of encounters in a row. Then it becomes a question of how many combat losses makes the campaign a loss.

It's not that it can't happen, but it's very likely that any campaign you run will go for like 3-4 battles and then end abruptly when the PCs are slaughtered.

Again with the idea that losing and dying are linked. How morbid.

Again, this is going to come down to how many consecutive combat losses make the party lose the campaign. The problem with a 50-50 average win chance is that it's the same as a coin flip. Flip a coin a hundred times and count the longest run of consecutive tails. It's probably quite a few in a row. So unless it takes a ridiculously high number of consecutive losses to lose the campaign, the players are quite likely suffer such a string of consecutive defeats at some point.

[MartinHarper]

Now, as to the main point:

So if you want to live to the end of the campaign, then the more encounters you have, the less likely this is to happen even though the chances for you surviving on each individual encounter is always the same.

And if this was stated from the begining, instead of making it as if the odds of rolling badly shot up for encounter #2, #3, etc, etc. we wouldn't have this arguement.

I bet you can't find one post, from anyone, on this thread, that has that implication. Feel free to prove me wrong. You have nearly thirty pages of posts you can draw on.

[Elennsar]

But even if you try to design a game where surrender isn't an extremely unpalatable option, good luck trying to sell it to the average gamer. My suspicion is that you'll have an easier time convincing most gamers to avoid some combats than getting them to surrender.

If the average gamer is going to chant "Victory or Death!", then the average gamer is going to be faced with the fact their choices are ensuring that those are their options.

Again, you're the only one who ever said that. It's a total mischaracterization of what anybody has ever said here.

No, it isn't. It may be a result of misreading or mistyping rather than anyone intending it, but it is implied that if you face one 50-50 encounter, winning a second is somehow less likely.

You keep making the argument that "it's possible" for the PCs to do better. True, but it's also equally possible for them to do worse.

No one is denying that they can do worse.

If the probability of winning one combat is 50%, it's pretty easy for the PCs to lose a bunch of encounters in a row. Then it becomes a question of how many combat losses makes the campaign a loss.

Depends on the combats in question. Losing something major (as in, major consequences to ultimate victory or defeat based on who wins) is much more devastating than losing something minor, whatever the relative odds of victory are.

...the players are quite likely suffer such a string of consecutive defeats at some point.

Or victories. If it is equally likely that you will roll higher than the average than lower, the players are quite likely to enjoy a string of consecutive wins.

I bet you can't find one post, from anyone, on this thread, that has that implication. Feel free to prove me wrong. You have nearly thirty pages of posts you can draw on.

I bet that you can't tell the difference between something that is implied whether someone intends to imply that or not and something that is deliberately implied.

[MartinHarper]

It may be a result of misreading or mistyping rather than anyone intending it, but it is implied that if you face one 50-50 encounter, winning a second is somehow less likely.

Where is that implied?

[Absentminded_Wizard]

But even if you try to design a game where surrender isn't an extremely unpalatable option, good luck trying to sell it to the average gamer. My suspicion is that you'll have an easier time convincing most gamers to avoid some combats than getting them to surrender.

If the average gamer is going to chant "Victory or Death!", then the average gamer is going to be faced with the fact their choices are ensuring that those are their options.

The thing is that RPGs are, at least partially, escapist entertainment. People don't do escapist entertainment to vicariously lose. That's why the heroes win in action movies and why the odds are stacked in favor of the PCs in RPGs.

Again, you're the only one who ever said that. It's a total mischaracterization of what anybody has ever said here.

No, it isn't. It may be a result of misreading or mistyping rather than anyone intending it, but it is implied that if you face one 50-50 encounter, winning a second is somehow less likely.

Or, more likely, just the result of other posters having a more sophisticated understanding of probability than you. People seriously didn't think they had to explain a lot of this stuff to anybody who was into game design enough to post on this board.

You keep making the argument that "it's possible" for the PCs to do better. True, but it's also equally possible for them to do worse.

No one is denying that they can do worse.

If the probability of winning one combat is 50%, it's pretty easy for the PCs to lose a bunch of encounters in a row. Then it becomes a question of how many combat losses makes the campaign a loss.

Depends on the combats in question. Losing something major (as in, major consequences to ultimate victory or defeat based on who wins) is much more devastating than losing something minor, whatever the relative odds of victory are.

No, but you seem not to understand the consequences of that fact. You act as if the small chance that the characters might have an improbable run of good luck and win, say, 10 encounters in a row completely overshadows the fact that it's easier for them to lose three in a row when those three losses lose the campaign for them.

...the players are quite likely suffer such a string of consecutive defeats at some point.

Or victories. If it is equally likely that you will roll higher than the average than lower, the players are quite likely to enjoy a string of consecutive wins.

The problem is that the consequences of victory and defeat are likely to be asymmetrical. Presumably, you have to make it to the final "boss" encounter to win the campaign. The journey to that encounter is probably going to involve a lot of combat encounters. Unless the number of consecutive losses that loses the campaign is pretty darn close to the number of total wins it takes for campaign victory, the PCs are much more likely to suffer a campaign-losing string of defeats than to make it all the way to the end without such a string.

[Elennsar]

Martin:

And you would get the same number of average survivals if you had everyone try roll a 9+ on a d20 to survive as if you had each of them make those 10 95% chances one at a time. Seriously. That's how risk works.

In two 50/50 encounters, 75% of the people will die. Not everyone, just three quarters of the people. As we've said, if you're fine with running a game where half the players have to make new characters every session, that is a perfectly acceptable mechanic.

Primarily the first.

The thing is that RPGs are, at least partially, escapist entertainment. People don't do escapist entertainment to vicariously lose. That's why the heroes win in action movies and why the odds are stacked in favor of the PCs in RPGs.

No, but it murders the drama in The Empire Strikes Back to whisper to someone (when Han is frozen in carbonite) that "he'll be perfectly okay, this is just to show that Vader is really evil."

Or, more likely, just the result of other posters having a more sophisticated understanding of probability than you. People seriously didn't think they had to explain a lot of this stuff to anybody who was into game design enough to post on this board.

And you would get the same number of average survivals if you had everyone try roll a 9+ on a d20 to survive as if you had each of them make those 10 95% chances one at a time. Seriously. That's how risk works.

Or people are wording it as if losing in encounter #10 is somehow more likely than in #1.

No, but you seem not to understand the consequences of that fact. You act as if the small chance that the characters might have an improbable run of good luck and win, say, 10 encounters in a row completely overshadows the fact that it's easier for them to lose three in a row when those three losses lose the campaign for them.

Not really. It does mean that if you want to be battling the odds, you need something like that (or less extreme) instead of eliminating the odds and just claiming it was tough.

Unless the number of consecutive losses that loses the campaign is pretty darn close to the number of total wins it takes for campaign victory, the PCs are much more likely to suffer a campaign-losing string of defeats than to make it all the way to the end without such a string.

Personally, I assume that it is more important whether or not you are able to get to Yavin with the information to destroy the Death Star (hopefully from more reliable sources than the Bothans?) than whether or not you win ten battles or lose ten battles between evacuating Hoth (which was not planned to happen...by the Rebellion, that is) and Yavin.

Having to win consecutively without losing to win overall is a bad set up - being able to lose three battles, come back and win big, and go on to use the fact you shoved this into 80-20 or whatever to win is much more dramatic and cool than any insistance on winning five in a row.

Besides, there are only so many incest jokes we can make about what Luke and Leia did on Hoth, so it was probably a good thing the Rebels left.

But more importantly, see the italic stuff.