Quote from: The Traveller;636269I'm not talking about single rolls though, rather about the aggregate. If we're on about victory conditions where an exceptionally high roll is required, it gets messy. The number of rolls is messy, the probabilities per roll are messy, the number of times an exploding dice are used is messy; it's a mess.

Are you thinking of how many rounds it takes to kill a big boss end guy?

Quote from: The Traveller;636261You miss what I'm saying though, Joe might only make a climbing check once per session out of scores of other rolls, and each climbing check would almost certainly have different probabilities. The number of rolls, the percentage chance in each game are all highly variable, far too much so to make any sweeping statements about randomness evening out. Options for exploding rolls complicate even that picture tremendously, especially if you're taking success into account rather than randomness for its own sake.

But the odds are never random... well random in as much as the designer or GM have a modle of difficulty that is consistent across their game.
The odds of Joe climbing a tmple wall are (Base odds + instance modifier) compared to his climb skill.

So If the game sets a standard or 0 modifier for most surfaces then Joe can usually suceed 80% of the time. If the temple is wall is rate 'easy' he might get +10%.
None of these odds are set randomly they are all drawn from the games physics engine as static values.
I do not know of a game where the odds of success are randomly determined before each dice roll.
The odds are extrapolated from the game engine.
The PC then rolls against those extrapolated odds.
This generates a result

Yes any single climb roll is 'random' but the set of (all climb rolls) confirm to the games probability model.

Quote from: Bloody Stupid Johnson;636266Perhaps you could see 'Always Fighting Orcs' as the outcome of trying to keep the amount of bias on the roll constant mechanically? Dunno.

Perhaps this is the case.

Some video game combat mechanics attempt to do this too but in real time, such as Dragon Age.

http://dragonage.wikia.com/wiki/Combat_mechanics

Quote from: jibbajibba;636282Yes any single climb roll is 'random' but the set of (all climb rolls) confirm to the games probability model.

Only if the results of any individual roll have a deminimus effect.  If the results of a single botched roll could have catastrophic effect (character falls and dies), then you may never get to the point where there are enough rolls to trend toward the average.

If we look at 3e and 4e D&D, there is more randomness in 3e.  That is because 4e very specifically reduced the impact of what one failed or successful roll could potentially mean in a combat situation.

Quote from: jibbajibba;636243Randomness is intertesting but I think random systems need to be looked at. I play a fair bit of poker. Some people tell me it's all luck. You have to explain to them that there is very little luck. Over a course of 100 hands all the luck will even out.

It is a bit of both.  In poker, when the players are of disparate skill level, luck will fade into the background and skill will typically win out.  However, when the players are of similar skill level, luck will be the determinate factor.

Quote from: Old One Eye;636325It is a bit of both.  In poker, when the players are of disparate skill level, luck will fade into the background and skill will typically win out.  However, when the players are of similar skill level, luck will be the determinate factor.

See not sure I agree with that as mathematically all luck will even out after a number of hands. So what matters is how you play.
Then you game into game theory and how best to maximise returns etc.

Luck some up in corner cases where 2 players both have good hands and probability predicts each of them has a high chance of winning. For instance there is J, J, 10, 7, 8 showing and I have 8, 8 in my hand so have a full house but my opponent holds J, J.... now that is luck but luck based on the chance that the chance my opponents holds 2 x J , or 2 x 10 which are c 2% and the only hands that can beat me I should still play the odds.
Over a set number of those events the luck will even out again.

Same with randomness in games.
Joe may fail a climb check but the chance of 100 climb checks failing are consistent and not random.
Now that might be poor consolation for Joe as he falls into a lava pit....

So its not a question of randomness as opposed to a question of stakes and hedging.

Quote from: jibbajibba;636282But the odds are never random... well random in as much as the designer or GM have a modle of difficulty that is consistent across their game.
The odds of Joe climbing a tmple wall are (Base odds + instance modifier) compared to his climb skill.

So If the game sets a standard or 0 modifier for most surfaces then Joe can usually suceed 80% of the time. If the temple is wall is rate 'easy' he might get +10%.
None of these odds are set randomly they are all drawn from the games physics engine as static values.
I do not know of a game where the odds of success are randomly determined before each dice roll.
The odds are extrapolated from the game engine.
The PC then rolls against those extrapolated odds.
This generates a result

Yes any single climb roll is 'random' but the set of (all climb rolls) confirm to the games probability model.

Did you even read what I wrote? One good example of what I'm talking about is combat, opposed rolls. The odds themselves are determined by the opposing dice roll and a myriad of other factors. If ever an RPG is released where the players do nothing but climb the same wall all day you'll have a point, and who knows maybe the shared narrative crowd might find that entertaining, but until then you're way off base.

Quote from: jibbajibba;636328Luck some up in corner cases where 2 players both have good hands and probability predicts each of them has a high chance of winning. For instance there is J, J, 10, 7, 8 showing and I have 8, 8 in my hand so have a full house but my opponent holds J, J.... now that is luck but luck based on the chance that the chance my opponents holds 2 x J , or 2 x 10 which are c 2% and the only hands that can beat me I should still play the odds.
Over a set number of those events the luck will even out again.

For the gambler with deep enough pockets, yes.  For the gambler that looses his $500 ten hands into the night and has no more money to continue gambling, no he cannot continue playing until the odds even out.

Quote from: The Traveller;636369Did you even read what I wrote? One good example of what I'm talking about is combat, opposed rolls. The odds themselves are determined by the opposing dice roll and a myriad of other factors. If ever an RPG is released where the players do nothing but climb the same wall all day you'll have a point, and who knows maybe the shared narrative crowd might find that entertaining, but until then you're way off base.

I read it . How many combat systems used opposed rolls?
D20 - no that is roll v target, WoD - no, Traveller - No, SW - No... etc

Yes I can see that in a game where Player B rolls their defense and then Player A must roll higher than a target number generated by player A to hit them that the odds might get complex. But I don;t knwo of any games that do it. I know of a few where its roll to hit v target number then roll dodge v target number and a handful where a parry mechanic means you need to roll higher than the number your opoent rolled over their target number but they are few and far between.

Whilst Cliffhanger the RPG might not be to everyone's tastes .... since all checks in a prticular game probably follow the same pattern chances are that the group 'climbing checks' is a synonomous subset of 'all skill checks' and that we can consider all skill check in the game to follow the same pattern.
No a game that involves everyone sitting round making skill checks seems a bit less prone to hyperbole.....

Anyway my actual point is more clearly demonstrated by One Eye's post re poker. Its about setting stakes and mitigating risk , and of course about roleplaying. In mean this in character not in a storygames way.  If you risk $500 on 10 hands of poker and loose then that is fine the probabilites would have predicted that was a liekly outcome. One assumes you are playing poker for fun and get enjoyment from it hopefully $500's worth. A 'serious' player with a $500 stake would be playing 2/4 games where over time (and 200 hands) they can eliminate the luck but playing enough hands to play the probabilities.

So if Joe the climber consistently climbs over huge abysses with no safety rope then he will die, eventually. We could compute quite easily the odds of him succeding 2, 4, 8, 16, 32 times on the bounce. Now that is fine. It might well be that Joe the Climber is a risk taker and the palyer has conciously decided to play him like that. But then when Joe fails the climb check and dies its not a display a luck and randomness its a concious roleplay decision by the player.

Quote from: jibbajibba;636549I read it . How many combat systems used opposed rolls?
D20 - no that is roll v target, WoD - no, Traveller - No, SW - No... etc

Even roll versus target is quite random, as the target is most often set by the GM. Firing neurons or rolling dice, in fact I'd say GM decisions are if anything more random.

Quote from: jibbajibba;636549they can eliminate the luck but playing enough hands to play the probabilities.

Okay this seems to be the crux of what you're saying, and characters can reduce risk by making wise decisions of course. But I've never seen a game where a player eliminated luck through good playing, the nature of the game goes against it, being about adventure.

It's way more complex than poker odds. It would be like trying to aggregate the odds from playing poker, snap, 52 card pickup and bridge simultaneously in environments ranging from a dining room table to hanging off the wing of a plane.

Quote from: jibbajibba;636549So if Joe the climber consistently climbs over huge abysses with no safety rope then he will die, eventually. We could compute quite easily the odds of him succeding 2, 4, 8, 16, 32 times on the bounce. Now that is fine. It might well be that Joe the Climber is a risk taker and the palyer has conciously decided to play him like that. But then when Joe fails the climb check and dies its not a display a luck and randomness its a concious roleplay decision by the player.

Per roll it's simple to work out the odds and the effect of randomness. Over an entire session or campaign, no. You can make a rough estimation based on the per roll setup, if you have stat (1-5) plus skill (1-5) plus d100 versus average difficulties of 80, randomness will always have a much bigger effect in that game than say (1-5)+(1-5)+1d6 versus 10.

Quote from: Bloody Stupid Johnson;635959Never played Uno though, sorry.

If that's the case, you've been seriously missing something in your life.

RPGPundit