No. A programmed algorithm will always be more open to decode that will benefit the player as opposed to something that is RNG. Thats why after a few weeks it was cracked and players took advantage. The mistake was to publicise your findings in a forum that devs normally ignore and highlight their incompetence in neon with sirens and air horns.
EDIT
in a related thread i have commended the analytical skills of those who did this. I am far from taking the devs side but if you highlight your profits and expertise (an expertise and care they lack) then you cant expect them not to freak out at the balance sheet implications
There was no reason to believe this was a bug. The devs claim that plenty of things are working as designed, and the code necessary to implement their scheme was much more complicated than the pure RNG solution. The way that it was implemented definitely reduced the chance of winning if you pick 1-2-3, so it definitely feels like it was done on purpose to drive up Dragonite purchases. Unless they tell us how it was expected to work I can’t come up with another conclusion than malice, because the lazy/simpler implementation would have been correct.
6 Likes
I dont dispute this at all. But naively belieiving an algorithm is less predictable than RNG is preposterous. And the intelligence and data skills of the community have proved that
EDIT
algorithms are based on primary or ‘seed’ values which are determinable by maths and probability.
1 Like
Exactly. i was adding an edit to demonstrate this when you responded. Believe me i am absolutely on the side of the community here.
But they didnt necessarily programme the pattern and that is what I am struggling to convey. Patterns depend on seed values which are to some degree random based on the parameters specified by the programmer. Many here refuse this explanation and thats fine by me.
1 Like
Bingo. finally someone gets it.
But this issue exhibits with clarity the way humanity are (in the non math or engineering community) completely unable to comprehend the concepts of infinity, random and imaginary numbers. I remember those days. How many understand sguare root -1?
Lol yeah but square root -1 is the basis of imaginary numbers that define modern day electronics.
Edit
and i loved those head shaking wtf days. engineering rocks.
1 Like
I also am a software engineer and: pretty much nobody should be rolling their own RNG these days. They’re furiously easy to get wrong, the consequences for doing so can be very bad, and there are high-quality, proven, and tested PRNG and CPRNG implementations available in most OSes and/or in third-party libraries like openssl.
I would be very surprised if IP2 rolled their own RNG. I would be surprised if this were a seeding issue as the “item X not between door Y” pattern was so absolute; this was not just a slight statistical edge, this was a clear-cut never-occurs thing. I would not be surprised at all by any of these 3 explanations for the pattern:
- a deliberate “hey, let’s put a bit of a fun twist on this” choice
- a buggy implementation of forming a 6-item permutation from an integer random distribution
- they’re doing some post-filtering on the RNG to reduce runs – which can and do happen in random distributions but which can appear disturbingly non-random to casual users – and this had some weird effects on their permutation implementation
but hey; we’ll never know the why; we only know the effect, which was that a purportedly random door distribution turned out to be somewhat non-random.
12 Likes
It’s bullshit like this that makes me sad this game wasn’t more popular, but not because i want it to succeed. Hell no, far from it.
I’d love to see the likes of Angry Joe, who despise the concept of loot boxes and gambling in video games, to show everyone just how shameful and disgusting these devs/publisher have become in recent years.
2 Likes
There are two different things at play here:
1 - The three distinct exclusion rules discovered are rules, as in: no RNG, percentages or guessing involved, they always apply; without exception. Not ‘sometimes’, not 97.3%. 100%. They’re static. They were manually designed or taken from a textbook cipher or stackoverflow.
They line up to individually form not just ‘kind of’ neat patterns, they form perfect patterns. Every time, without exception. All the way down to how trivially perfectly the card numbering scheme lines up with the door numbers.
Maybe to add a fun brain teaser (fair point, won’t discount the possibility), maybe to manipulate odds. Deliberate and intended in any case.
2 - The RNG part is what can be found behind the doors after applying these exclusion rules. Actual outcomes line up neatly with expected outcomes here too (+/- 1-3%). This uses the ‘good event’ percentage as used in the tool. Here’s @CaptainAwesome’s data set for example (215 runs at the time):
Door 1 | Actual: 61.4% | Expected: 60%
Door 2 | Actual: 58.6% | Expected: 60%
Door 3 | Actual: 62.79% | Expected: 60%
Door 4 | Actual: 80.47% | Expected: 80%
Door 5 | Actual: 78.6% | Expected: 80%
Door 6 | Actual: 58.14% | Expected: 60%
Because this is across completed runs, I went over a smaller set of 100 reports and checked first picks. First picks also seemed to line up with expected outcomes (larger variance, <=6%).
We were going to check second, third and fourth picks next, but there was no time.
…and if you needed more than 4 picks to win, then your game’s broken anyway!
In other words: Exclusion rules always worked as expected and, as far as I can tell, RNG wasn’t off either.
E: This was across all players of course. There is still a possibility that game RNG still punishes certain players more than others and things just average out in the end.
12 Likes
Notice how no one has liked this post? All you could do from our feedback is “We have some ideas, and if we do any of them, expect them in no less than four months?” LOL
2 Likes
Thanks IP2 for making something that was fine enough, shit and then making it shitter again. I mean is this all you do, first GAP’s now this go fly a bloody kite. How much more do you expect people to tolerate? There arent that many people playing GoW and I can imagine you just lost a few more. Whats next? Come on give us a teaser so we can be pissed off now.
1 Like
I just crafted my 3rd brown dragon!!! how exciting!!!.. 
If you get handed the long path by RNG (by picking up the special along the way), then out of 100 people only twenty make it to the finish line (w/ tool). You still lose 80% of the time on the long path, even with the tool.
Even on the short path (boss-boss-boss) and with tool, nearly two out of three players will not get a perfect run.
Of course, there is still a possibility that game RNG punishes certain players more than others (for a while anyway) and things just average out in the end.
2 Likes
I totally agree with this sentiment. I actually started to feel like playing again - more than just grind. I was having fun trying to get through dungeons.
BUT NO… this company has become the WORSE when it comes to player base support. I have never seen in the 30+ years of playing video games of any kind, this kind of crap.
But hey, we’re a minority anyway. I remember when there were hundreds of new messages daily on the forums. Now…it’s almost crickets.
Be proud Dev - you really are anti-player base - anti-game
3 Likes
I suspect at a very basic level it was programmed to thwart people who took the “most obvious” or “most likely” paths. So if you picked 4-5-6 every day, or you picked 1-2-3 every day, the design made you less likely to “win” than if you picked almost any other permutation.
It seems quasi-random, but biased against certain combinations.
For 3 bosses in a row, these were the only possible starting patterns with 456:
456231
456312
465213
465321
546132
546321
564123
564231
645123
645312
654132
654213
=12 in total
out of 120 “tool” combinations
For 3 bosses + Special in your first 4 picks using 456 + x, you had:
246315
256134
264135
265314
346125
356214
364215
365124
416325
436215
461235
462315
516234
536124
561324
562134
614235
615324
634125
635214
641325
642135
651234
652314
= 24 in total
out of 120 possible “tool” combinations
On the other side going with 123 and get 3 bosses had just:
= 0 in total
4 Picks containing 123 + x:
234561
235641
241563
251643
314562
315642
342561
352641
412563
431562
512643
531642
= 12 in total
out of 120 possible “tool” combinations
1 Like
It feels so weird to write “had” instead of “has”…
Can’t believe it’s already gone
2 Likes
the bottom line is whether or not you put up with whatever you think happened. its your choice and i have made mine
1 Like