Extraordinary bad luck or intentional game design?

[Apogee]

Been playing this on my iPad for a while. Completed all Story campaigns on all difficulties no problems. Quest mode is highly frustrating. In my 20's with my Quest mode characters and ever since the last update, I've noticed a significant leap in failure, I usually did Heroic with 4 chars and barely ever ran out of time, now it's 1 in 3 that end in failing a quest from blessings running out.

 

Why am I failing? Either I'm rolling a lot of 1s or 2s regardless of die, or I'm failing checks by 1, or henchmen/villains are consistently at/near the bottom of their decks. With my die rolls, I'm now always being over cautious, treating each die as half its value rounded down, then adding 1 or 2 extra die through blessings r other means. But still about half the time, I'm getting the aforementioned failure through rolling 1s or failing by 1.

 

Is this by design to slow progression? It's not fun anymore. I spend 15+ mins on a quest digging through decks as fast as I can only to fail and I get nothing. A stupid waste of time. I understand no gold for failing but why no XP?

[Hannibal_PJV]

It is not design. I have played quest mode to level 40. Sometimes it is real pain and sometimes it goes smootly. I have had very lucky die rolls and very unlucky die rolls among those so I think that it balance out in some point.

Some wildcard combinations Also Are hard and some of them Are quite easy. So there is a lot randomnes how hard the scenario will be.

Edited November 1, 2016 by Hannibal_PJV

[Ethics Gradient]

Why am I failing? Either I'm rolling a lot of 1s or 2s regardless of die, or I'm failing checks by 1, or henchmen/villains are consistently at/near the bottom of their decks.

I can assure you that the random number generator does not play favorites or hold a grudge.  Sometimes the dice and scenario modifiers just aren't in your favor.  Without getting too far into the details, the game uses a pretty standard method to draw random numbers.  On the "randomness" scale, it should have a distribution somewhat better than the cheap plastic RPG dice that would have been found in the tabletop version of Pathfinder Adventures.

 

Is this by design to slow progression? It's not fun anymore. I spend 15+ mins on a quest digging through decks as fast as I can only to fail and I get nothing. A stupid waste of time. I understand no gold for failing but why no XP?

It's just how it is.  Not to slow progression, but to slowly crank up the difficulty as you progress in level.  As your characters gain in level, the game generates harder scenarios based on a series of quest templates.  Sometimes it takes a a bunch of failures and narrow victories to help "gear up" your characters whenever they get bumped to the next set of templates.  A 33% loss rate is pretty typical when it's time to go hunting for better cards.

 

Also, I'm not a fan of the all-or-nothing approach to XP either.  It would be nice to leave every scenario with at least the XP from defeated banes.  I'm sure there's some reason for the decision, but there should be some sort of incentive to stick out a losing scenario.

 

Edit: Need to pay better attention to my losses.  It's been pointed out that you do, in fact, get to keep your bane XP when you run out the clock.

Edited November 3, 2016 by Ethics Gradient

[Apogee]

Thanks for the responses. I do accept that part of my lack of success in Quest mode is due to not having access to better cards (I assume I will if I hit level 30? Is that when you unlock the next tier of adventure deck cards?) and more power feats.

 

However, the jump from breezing through Heroic quests pre-1.4.4.3, and often with risky dice rolls, to being over-cautious with dice rolls and failing much more considerably post-1.4.4.3 is just something I can't reconcile, particularly with the petty punishment of not rewarding XP for failing a quest, so I can't potentially move faster to a higher levels where I can improve my characters and get access to better cards.

[Thyraxus]

Also, I'm not a fan of the all-or-nothing approach to XP either.  It would be nice to leave every scenario with at least the XP from defeated banes.  I'm sure there's some reason for the decision, but there should be some sort of incentive to stick out a losing scenario.

 

Actually, unless your characters die, you do get whatever XP you earned for defeating banes at the end of a scenario - if you lose by running out of time. But you get nothing if you use the "forfeit" option from the menu.

 

The latter distinction is rather important. If you run into a situation where you know (or think) you can't win anymore, but have already defeated a bunch of banes for some juicy XP, you have to forfeit "manually", by actually playing all the remaining turns, just doing nothing. Kind of annoying, really.

Edited November 2, 2016 by Thyraxus

[Ethics Gradient]

Actually, unless your characters die, you do get whatever XP you earned for defeating banes at the end of a scenario - if you lose by running out of time. But you get nothing if you use the "forfeit" option from the menu.

Ah.  You are correct!  I'm not sure why I never really noticed the distinction.

 

If I'm playing with permadeath on, I guess I have a tendency to forfeit a losing game before doing anything that may risk killing off a character.

[Loswaith]

 

Why am I failing? Either I'm rolling a lot of 1s or 2s regardless of die, or I'm failing checks by 1, or henchmen/villains are consistently at/near the bottom of their decks.

I can assure you that the random number generator does not play favorites or hold a grudge.  Sometimes the dice and scenario modifiers just aren't in your favor.  Without getting too far into the details, the game uses a pretty standard method to draw random numbers.  On the "randomness" scale, it should have a distribution somewhat better than the cheap plastic RPG dice that would have been found in the tabletop version of Pathfinder Adventures.

 

...

 

You would think so however testing has proven other wise for me.  From 1000 rolls (for the first calculation) testing the half point (below or above the half way number on the dice) ~63.4% have resulted in below the half point within the app (from a mix of updates, scenarios and story/quest mode) on rolls for the player, while 1000 rolls of the same kinds of dice (d4 to d12, with the general spread of character stats) gave me ~52.7% below the half point.  

[Ethics Gradient]

You would think so however testing has proven other wise for me.  From 1000 rolls (for the first calculation) testing the half point (below or above the half way number on the dice) ~63.4% have resulted in below the half point within the app (from a mix of updates, scenarios and story/quest mode) on rolls for the player, while 1000 rolls of the same kinds of dice (d4 to d12, with the general spread of character stats) gave me ~52.7% below the half point.  

Yikes.  That's pretty rough.  I admit that I don't have 2000+ samples to work with, but all signs point to the app using Unity Random.Range() to handle dice rolls.  It's a pretty standard function unless you're concerned with cryptographic randomness or plan on starting an online casino.  Back-end is an xorshift128 PRNG, which isn't perfect, but is fast and fairer than a lifetime of tabletop gameplay.

 

Found a dev post I remembered reading forever ago.  No reported bias in a million simulations.

 

If there is something else causing the dice to go crooked, I haven't been able to find evidence of what it might be.

 

[xkcd]

 

Edit: Added in the link to the ancient dev post.

Edited November 17, 2016 by Ethics Gradient

[KhorkynSG]

Thanks for the responses. I do accept that part of my lack of success in Quest mode is due to not having access to better cards (I assume I will if I hit level 30? Is that when you unlock the next tier of adventure deck cards?) and more power feats.

 

However, the jump from breezing through Heroic quests pre-1.4.4.3, and often with risky dice rolls, to being over-cautious with dice rolls and failing much more considerably post-1.4.4.3 is just something I can't reconcile, particularly with the petty punishment of not rewarding XP for failing a quest, so I can't potentially move faster to a higher levels where I can improve my characters and get access to better cards.

If you wanna to acces better cards you can group tres character of same tier and one of better tier (better cards and more XP, also more powerful enemies)

[Loswaith]

 

You would think so however testing has proven other wise for me.  From 1000 rolls (for the first calculation) testing the half point (below or above the half way number on the dice) ~63.4% have resulted in below the half point within the app (from a mix of updates, scenarios and story/quest mode) on rolls for the player, while 1000 rolls of the same kinds of dice (d4 to d12, with the general spread of character stats) gave me ~52.7% below the half point.  

Yikes.  That's pretty rough.  I admit that I don't have 2000+ samples to work with, but all signs point to the app using Unity Random.Range() to handle dice rolls.  It's a pretty standard function unless you're concerned with cryptographic randomness or plan on starting an online casino.  Back-end is an xorshift128 PRNG, which isn't perfect, but is fast and fairer than a lifetime of tabletop gameplay.

 

Found a dev post I remembered reading forever ago.  No reported bias in a million simulations.

 

If there is something else causing the dice to go crooked, I haven't been able to find evidence of what it might be.

 

[xkcd]

 

Edit: Added in the link to the ancient dev post.

 

I was noticing an inordinate amount of low rolls, hence the prompt to test them (even if it just quashed the bias that our brains can recall).   The potential factor though are other things influencing the values out side of the generator itself, further data should tell me whether I hit an anomaly or not.

[justaleaf]

Sorry for the bump. I've noticed a low average in my time playing (since day 1), and it did seem to get worse after a few updates. I began over-rolling everything, and still losing many gimmie rolls. (amazing how frequently I can lose an 8 check with 2d12 and 1d4... an average of 15ish)

While I'm totally open to simply being ill-favored by the RNG gods, my regular tabletop PF ACG group has had a similar experience. We've taken to called our worst rolls at the table "apping it," in honor of how low the averages feel on the Obsidian game.

 

I've played multiple runs through Runelords, Skull and Shackles, and we're almost through Wrath right now... and in that time, I can maybe only remember one session where our averages were so bad that it was even remotely like the experience I have on the app, in my majority of games there. So I'm not saying it's rigged... I'm just saying it's rigged.

And no one is about to convince me that monster rolls aren't budged . "On a roll of 1 [blank] happens"... you're going to see 1 rolled more than any other number on that d6, lol.

[busterdog]

I agree with the not-so random dice rolls.

I've had waaaay too many low rolls on average.

I've taken to rolling actual dice while the computer sims the dice rolls in the game - what a difference!  I suggest to anyone to try that and then try to argue that the RNG in game is fair.

D4 dice rolls seems to the worst offenders - I consistently roll 1's when I am using healing potions, greater heal, rolling for weapons etc, AND I've noticed that the henchmen consistently roll 3's and 4's to increase the difficulty to defeat them.

I see far too many 1's when rolling any dice D6, D8 and particularily D12's.

To be fair - I have not played any computer game that relies on RNG without noticing some irregularities.

Just my thoughts and experience.

[Ethics Gradient]

So, are these some of the possible bugged RNG theories?

 

1. Player dice roll lower than expected average, including a high proportion of 1s.

2. d4s for character benefit roll lower than d4s that work against the character.

3. The "On a roll of 1/whatever" penalties happen more often that it seems they should.

-----

Initial thoughts...

 

(1) I had a pretty good trace on this, and it seemed to behave as expected.  The game lines up your dice, runs them through the appropriate RNG, and passes back the results.  I don't think there's anything "loading" the dice in the game's favor.  It is 100% Unity API random number functions.  If there is some sort of fault, it would have to be downstream of the RNG.  If UnityEngine.Random.Range(1,5) (max exclusive) isn't providing the App uniform random integers from 1 to 4, I'm not sure what Obsidian can do to address that.

 

(2) All d4 rolls (or all rolls, technically) come from the same stream.  It might be inexplicably annoying, but if "good" d4s and "bad" d4s are balancing out, that's a sign that things are probably ok.  Rather than a tabletop "everyone has their own dice" approach to randomness, the game essentially has one set that everybody has to share.  

 

(3) There are many different powers/effects, but they all seem to lead back towards the same core RNG.  That being said, since there are so many different powers/effects it could be possible that one specific implementation can produce funny results.  Investigating each one is probably a step further than my little userland debug experiment was meant to go.

-----

If we get back to the OPs question forever ago.  I'm relatively convinced that there is nothing intentionally penalizing the player.  Sure, there could be a bug somewhere, but there's no apparent "house advantage" when it comes to the RNG.

 

Everyone seems to have a different perception of randomness, but my experience has felt pretty fair.  I swear I've had a worse rolls playing tabletop games.  Though, Chessex and GW dice are essentially made out of distilled bad luck and plastics that must have been stolen from some cursed burial ground.

[Flounder]

We have a person on the team who consistently rolls junk.  We use his name to describe particularly crappy rolls when needed most but "Apping it"?  Ouch, sir!

 

It is random though, pinky-promise.

[Ethics Gradient]

We have a person on the team who consistently rolls junk.  We use his name to describe particularly crappy rolls when needed most but "Apping it"?  Ouch, sir!

 

It is random though, pinky-promise.

Hmmm...  official comments are usually the first step in the coverup for any good conspiracy theory.  Maybe I should rethink my analysis.