I’ve seen stupider mistakes in coding. Not usually so efficiently clustered around one design element. And especially not focused around a design element so integral to the revenue stream. But still…
“Nice” dataset is going to be a bit of a stretch, at least on my budget. Of course, if Tpenn wants to lend us the company credit card, we can crank this into high gear. 
I don’t mind contributing some data. I do however only summon on full moons so please take that into account as it could be affecting my results. It has worked pretty well for me so far. Maybe the seed also uses lunar cycles? Full moon summons between :35 and :55 could be the way to go.
My results to date on Atlantis pulls, by the way, are depressingly consistent with an unbiased RNG.
| Category | Measured | Theory |
|---|---|---|
| S1 3* | 19.25% | 21.80% |
| S1 4* | 12.21% | 11.90% |
| S1 5* | 0.47% | 1.00% |
| S2 3* | 47.42% | 49.20% |
| S2 4* | 19.25% | 14.60% |
| S2 Non Featured 5* | 0.47% | 0.20% |
| S2 Featured 5* | 0.94% | 1.30% |
| HOTM | 2.35% | 1.30% |
How much total tries? just curious 
213 so far in that data.
Awesome job @Garanwyn
Call me petty but no there is no comaprison between repeatedly waiting for a drop rate to increase and a roulette wheel.
To elaborate. You had a new level which drops a new character. You need 14 of the character to max it.
Subject a (high end) subject b (middle) subject c (low end)
Subject a. Has incredible luck with the event. He farms it in a few hours and maxes the character only a few runs where the character didn’t drop lets say 4 out of 18 runs where the character didnt drop.
Subject b. Farms for a day (over 25 runs) and winds up with 9 of the character. He farms 5 more the next day in 12 runs. (This is how it was for majority of the player base.)
Subject c
Farms all day (40+ runs) and winds up with 4 drops. Next day lets say his rate gets a bit better. 8 more in 20 runs. Last day he gets last 2 right away. (Some players had extremely bad rates last for 3+ days)
Those 3 patterns represent the player base every single event.
Each time there was a character drop level like this you had a group of players on the high end some in the middle and some on the low.
So I in turn would not farm a character if it did not drop in 3 runs. I would stop and wait till later that day or next day. Try the next day no, next day no. Eventually i would hit a day where it would drop almost every run much like the players on the high end. Thats when I would farm it and it worked seemlessly. I tried to tell the player base this but they didnt take to it. So I sat back and watched players struggle with farming while for 3 years never had trouble farming anything and usually finished hours from when I started. Aside from waiting for the right time I would only spend a few hours maxing any character drop.
The most effort I had to put in farming was if I decided to farm in the middle spectrum. This was rare out of me as its much easier to farm when the character drops almost every run. Usually I wouldnt even farm this and wait for the high end.
I had suspected all of this on my own but when I went to reddit there where 3 “veterans discussing the very same thing”. They unknowingly confirmed what I had found that a better rate could be had by waiting. However going back and finding that post once the arguement ensued proved too difficult. Players remained that it was purely random and farmed regardless of how it was dropping. I was either waiting to farm or done already. This remained consistant throughout my coarse of playing the game.
Honestly I thought certain players finishing in hours and other players not getting a drop all day would be proof enough that everyone is not farming with the same rate at the same time but they didnt think so and chalked it all up to random. If it was truly random, i wouldnt have been able to farm 14 of every character release with very little effort consistently for 3 years. Thats over 60+ characters if you limit it to 2 new characters a month which It was surely more than that.
Sometimes the rate would change while I was farming. Like I would get up to almost done (9or so) and suddenly the drop rate disappeared because I took too long to farm it (twice this happened at gift reset but also tried to proof that too but, it was also incosistent indicator of rate change. Rarely rate change would happen at reset but nowhere near everytime). Near perfect runs then 8 straight runs and no drop. Needless to say, I would wait again then finish farming when rate got better, again. Never farmed anything that wasnt dopping easily and in doing so, I still maxed everything and saved myself tons of frustration and stamina but watched other players get frustrated with drop rates.
You’re petty.
That sort of behaviour is typical of pattern-seeking humans and especially typical of gamblers. It lacks any rationale and fails to incorporate any means of measuring its efficacy. The way you describe it suggests very strongly that it didn’t work, but you seem not to be aware of that.
If you try three times and pull nothing, you give up. This happens many days in a row, until “eventually” your character drops “almost every run”. That sounds no better and no worse than the typical experiences you described the player base having:
So the way you yourself chose to describe your process makes it sound very much like your method achieved nothing except perhaps making you feel better about yourself (which I’m heartily in favour of: kudos). By definition, you would stop playing after short streaks of losses, and persist in playing when you hit a streak of wins. That’s a good way to avoid the misery of a long streak of losses, but it doesn’t mean you have any fewer - you just spread them over more days.
You have given no indication that you made any attempt to measure or test what you were experiencing. Did you even log your attempts? You haven’t mentioned it, so I’m going to uncharitably assume that you did not. And your perception of what you experienced is horrifically distorted, because you are human like the rest of us.
This is why we need to measure our experiences, record the measurements, and use a robust methodology to assess those measurements. If we don’t do that, all we are left with is our tendency towards superstition and irrationality.
You should take comfort that the misperception you were suffering from is not unique. It is extremely common. People also think rabbits’ feet are lucky, and the number 3, and the number 7, and the number 8. They even post these thoughts on reddit. Never pay attention to reddit. You know better than that.
Have you seen those sad types in casinos, standing by the roulette wheels with their little pads and pencils? They think that the wheels will get ‘hot’, and that they will be able to ride a streak, if they can just identify when it is coming. Sometimes they get lucky and hit a few winning bets in a row, but they never seem to make any money. They are sure that they can, though, because when they take their pads home at night they see that the streaks are there. If only they could get on them a bit quicker. And get off them a bit quicker. And have better personal hygiene.
Randomness is streaky. But streaks are not predictable, and walking away when you think things are going badly does not make them any more likely to go well when you return.
You deserve credit for feeling good about yourself. Well done! But - unless you are keeping it a secret - you made no attempt to measure whether what you were doing worked or not, and your claims that it did are utterly unconvincing.
Next time you think you’ve cracked something, just log the results. It’s easy and it can shed a lot of light on the situation.
i feel you mate… ive saved for the last guin around 500€ worth on gems… never did a single pull before from nothing sow i’m talking around 250+ summons here… it was for me also a huge dissapointment drawed 1 1 5* think around 220 3* and some 4*… maybe people say i’m crazy cuz i do that… but i didnt do a single buy before just the vip… then you see some people use 300 gems and pull guin or king arthur… i can easily afford it don’t need to life different for it… and i’m supporting the game i like to play… but i think there must be a sort of a bonus system that rewards their loyal paying custommers… without them theres no money to earn for sg or something… i know ive played a game before that worked with a vip system… they can do like if you used xxxx gems in a 3 month period you get some kind of bonus wheels with 5* heroes in it and just spin it so people have a look out for it… you can do atlantis heroes in it cuz they seem to be very popular and highly wanted… will bring extra life in the game and people wont feel bad if they spent some and got nothing in return… it will increase their income tbh…
I agree with Fatal1 about it’s some things they can do for pay to play players after spending so much on the game. Other games I play do this and they get more money than this game from me in the long run 
But I’m not worried about this game in the money aspect because I’m mostly free to play now because I don’t need the second builder anymore and I don’t do pulls for hero because I have no ascension items to give them. So I just run TC20 all 4 camps.
Rewarding loyal customers would be a great idea and show of good faith. They could make it like atlantis where after 10 summons you get some asc items, heck even 100 summons would be something.
Well. Agree to disagree. I see you follow the other religion.
That reply was one huge assumption with pre disposed beliefs. There is no way I should be able to reliably farm a card everytime. To put this in perspective.
I would have to have beyond lottery luck to make that happen.
Your assumption is the game is actually random and I just had beyond incredible luck for 3 years straight. Which flat out does not happen farming randomly, you will hit the low end multiple times throughout events. I however avoided the low end with almost prefect result. Again that does not happen farming randomly never? Never. You may have good rates for a event or 2 but eventually you will have to spend several days trying to farm a character. Vs a method that takes 2 hours or less every single time.
Ive run into your kind of logic before as well. We’ll just agree to disagree. Based off information given what are the actual odds of me be able to do that twice a month.
Also what are the odds that Id be able tp finish within 20 runs everytime?
The way I got the player base to at least respect my findings was finding other flaws in the rng. For example the game used a board of multi colored orbs much like this one although not exactly the same. The board was pseudo rng with different color orbs dropping.
I had an alt account so I logged on both at exact same time. Went through menues at exact same time. And on both accounts the boards were identicle even the ones that dropped until a variation was made then they would differ. That shut up a lot of the its all random followers because they realized they dont know everything even though they thought they did because they couldnt “its all random” explain it.
And his theory put into practice in this game what would it be? I am inclined to believe that you are right about what I have seen (manipulations of devs in the margin), but I can not see its practical side.
Well I havent really found a legit theory for this game I just get irritated when I saw psuedo random is random because I know its definitely not.
Not even all rng manipulation is practical. Such as the trick I mentioned previously that only worked on a certain seeds/algorythems. The fact that you had the to find the algorythem to get it to work (sometimes this took a week and change sometimes it was immediate) ment you still had to cash to find it. When the games character list grew a lot and rates became better, the trick itself became impracticle, because you were now getting top cards at same rate or better with normal randomness. So since the games rates improved over time, the trick became useless.
Before that, How I knew it was working was a que from the game. The game has 3 winning summon, somewhat super rare animations. When the seeding was right the trick would cause the same winning summon animation that starts with your screen cracking then goku saying bye guys. I would get the cracked screen bye guys animation back to back to back to back. Once again this is technically possible but doesnt really happen. It was the norm while I was doing the trick.
I just want all gamers to know rng is not true random. Even researching rng will tell you thats its not true random. Will this help us right now? No. But most rng does have tendecy flaws so keep ypur eyes open.
Not at all. My assumption is that you didn’t bother measuring your results, recording those measurements, or testing their significance. So when you make this sort of claim
then we can’t really take you seriously, can we? Because you never recorded your results or even quantified your expectations.
Did you even achieve better results than we would expect random draws to achieve? We don’t know, because you never bothered measuring anything. (Although it bears observing, again, that the way you described your results made them sound strikingly similar to the results you say were achieved by the rest of the player base - and you seem to be unaware of this.)
So what are your claims of success based on, if not measured results? What you feel? Lot’s of people feel like their weird little superstitions pay off. The way to figure out whether they’re right is easy: measure the results. (Spoiler: superstitions are usually useless.)
Congratulations on your detective work, however you certainly did not find any “flaws in the rng”. If you generated identical game boards on different accounts (for more than a few dozen tiles), then you uncovered the fact that tiles are chosen deterministically, not stochastically. No PRNG generator can generate such results - not even one from 1985.
So far you have provided no evidence whatsoever to this effect. You have, however, given a strong indication that you can’t even tell the difference between deterministic results and stochastic results. This, among many other things, does not inspire confidence.
You mean that when you got bad results, you attributed them to the algorithm (please, for the love of the gods, absorb the correct spelling of algorithm) not being in your favour, and stopped playing. Then when you got good results, you attributed them to the algorithm being in your favour, and kept playing… unless things started to go badly, in which case you would attribute them to the algorithm having changed, and would stop playing again.
This is so typical of gamblers’ behaviour when faced with randomness that you might have copied it from a text book.
We can tell that this is all nonsense by the fact that not a whit of actual evidence has been produced in support of any of this. I’m not interested in hearing stories about goku and their cracked screens, whoever and whatever that is - I just want to see the numbers that back up these nutty claims. Unsurprisingly, there don’t seem to be any.
Sounds like they generated a single seed every few seconds and pushed it to everyone on that server without generating independence between clients in the subsequent mixing. That seems…unwise. But it doesn’t at all suggest that the core PRNG itself had problems. It just suggests really bad seed management.
My question to @Skeezy: why do you think that the PNRG quirks you were able to exploit in your previous game exist in E&P?
That’s a smart theory that hadn’t occurred to me but seems the most likely explanation, using Occam’s Razor. It also seems extremely lazy - why use a seed based on a seconds counter, when you could use a milliseconds counter and avoid the possibility of replication by fat human fingers?
On the other hand, thinking about it a little more, why not use a seconds counter? There’s no downside to two devices receiving the same seeded PRNG stream, aside from the possibility that paranoid types will seize upon it as evidence of some undefined problem.
