I like this game a lot because of its fantastic art and that nice titan play, but that summoning… For me spending some VIP months is fine, as this game has to be developed and maintained. (Before F2P was “invented” all those computer games cost around 20-50$.) But I will not spend extra money for chances. If I remember correctly, special 5* are 2.8 per cent from Atlantis and that is one special 5* for about 10K gems. Seems OK to me, but even drawing 100 times leaves almost 6 per cent to get no special 5* hero (0.972^100) and that seems really wrong.
I would pay 15K gems for one random special hero, but nothing for nothing guaranteed.
What @Brobb asked was for you to post your theory so other people can test it. Instead you just claimed that nobody would believe you.
In the absense of any actual evidence provided, my default assumption would be that you can’t back what you’re saying. Talk is cheap, and so far you’ve just talked instead of actually providing any proof of your statements.
“A wise man … proportions his belief to the evidence”
–David Hume
So far, I’ve seen a whole lot of talk, and very little evidence. You want me to believe you’ve really discovered an exploitable trick? Show me data.
I’ve collected and published data on the randomness of boards in this game. I invite you to do the same regarding your theory.
I’m going to take a guess at your strategy: what you do is a single draw first. Then, if you get a 5*, you do another draw. If you don’t get an 5*, you conclude that the RNG is in a bad place and stop. On the times you do a second draw, If it’s also a 5*, you do another draw. And so on. Is that your technique?
Aside from the smuggness, I’ll go ahead and share even though I do risk update changes. This is not an exploit but the most recent thing I am proofing.
I have only chased one character on this game so far. Mother North. During my two weeks of chasing mother north by the time I got her I noticed majority of my 5* came within a certain timeframe. So to test this I have done majority of small groupings of single pulls (3) during this timeframe with uncanny results so far to date.
However I need to proof so I havr to pull outside the time frame as well (proofing gets very expensive). After months of observation a clear result should be possible.
That is what I am currently looking at.
Other than that my recommendation is that small groupings of spread out single pulls is most cost effective way to pull. More 5* for your buck.
I find 3 pulls is a good way to conservatively check your current rate. Im also still feeling out how often this games seeding/algorythem changes. Since it can be literally anything from time, date, to opening the summon (i notice the lineup changes everytime you re-enter) to literally anything or number of things. Just getting a feel of wether its strictly time based or not is a great help.
As stated there is a high end and a low end. I havent been playing long enough to say the summon is exactly the same as dokkan or not. However thizs is true in most rbg games. Chasing will break you, if your on the wrong algorythem.
On the highest of high end, desired is your first card. This is because you were seeded that way for whatever reason or value. I find rule of 3 to be the most cost effective way to test the waters especially with the proof Im currently doing.
To give an example proofing dokkan took between 500-900 a month because again to proof you have to pull against your theory to see if your correct.
A lot of times with manipulating rng, people/fanbase want a cheat code and consistency. Thats not how most rng tendencies work, although some do.
Long story short I find 3 to be enough to check for high end seeding.
I still have not studied the summon long enough to know if streaks maintain on this game for periods of time. Nor will I probably ever spend enough to know but we’ll see
Sorry, just to make sure I’m understanding you rightly: you’re saying you always do 3 pulls at a time, no matter what? So if the first one is a 3*, you still do two more?
Oh yes. I mean you could limit yourself to 1. But you’d always need the very best seeding to pull 5*. 3 gives wiggle room to not have the best seeding but still be rewarded if your on the high end.
People do the same thing with roulette wheels and blackjack tables, imagining that streaks are predictable then using hindsight to convince themselves they knew what was coming.
What you claim is very easily testable, and contrary to your earlier assumption, I (and many others, I wager) would be happy to try out your ideas and record the results. So I again invite you to publish your analysis and subject it to some testing. We’ll do it for you - no effort required on your part at all. Otherwise it just seems like superstition, doesn’t it?
(Posted, I note, without hostility or anger. There is a technical term for when a person perceived hostility and anger that isn’t actually there.)
I see that you do understand the premise of only pulling on high and not on low. However cashing is required even for just checking the waters.
I see that you get what Im saying. Never try to turn your luck around because your probably on the low end. A high end would be quite noticible from jump but again Im more familiar with dokkan where the algorythem literally allowed me to continue pulling good pulls for a period of time.
Im still not sure about the high end of this game but I have noticed the low end is simlar, meaning yes you can spend over 10k gems and not get anything. This actually helps determine success.
Just a comment here from the developer perspective. Usually in this kind of games generate a seed server side for each success for each player is very expensive in calls-to-server terms. Usually a seed is generated every few seconds and that seed is used to N calls, and then later is mixed in some way client side and after that you have a result. Using a seed genetator directly time based is really poor design, much worse if you don’t reevaluate it client side. Maybe in that DB game that was possible (not doubt it for a second) but here, if you read all the testing results you will see there is no pattern. Still if hipotetically there were one, you would need so much data that would be almos impossible to prove.
As far as the dbz drop manipulation… dude, I did it for 3 years and it worked everytime. Not only that but other people confirmed it, who I dont talk to, as I dont post on reddit.
If you really want to help which it seems you are genuiwine then I will elaborate what I’m proofing. I notice that majority of my 5* for two weeks came between the minutes of *:35 and *:55.
In general I would agree that this means nothing but I literally had crap luck earlier in the hour. Still doesnt mean much but noticible enough to proof.
So since mother north came out Ive been doing singles in that hourly time range (no not every hour but leisurely). 3 pulls 3 times a day on for 3 or 4 days.
In the priate summon I got two lady luck and 2 other 5*. In recent atlantis summon got thoth 1st pull another thoth later that day and inari (fox tail girl). I also did summons and multis outside of my timeframe (about 40 pulls) and got Ice dude with the hammer and another regular 5*.
This all confirms nothing but, the characters gotten with singles test from both events looks promising and raises questions. Again it will take months before I can say wether there is someting to it or not.
That is a very impressive claim: I’m sure you make it sincerely. It is not, however, a test - nor even an attempt at a test.
Okay: now we have something testable. Think we can collect a decent sample of summons between *.35 and *.55, and compare them to a decent control group summoned outside those minutes? That seems doable, I would think.
(I’m not a big summoner myself - I hope a few paying players will be happy to participate.)
I will say clear your head of expectations and failure is not full proof. Multiple wins or multiple faileres gives us a guideline.
I say that because to often players expect immediate results and again thats not how rng manipulation works. In fact I would research rng manipulation and go through a couple legit programmer sites that talk about it.
Lastly thank you for helping. My initial cry on forums was one asking for help testing but you can imagine the backlash and jumdgement without help trying to proof this kind of thing. Players are often to critical and demanding of something thats variable. Like I said, a tendency not a rule but when you research rng manipulation it will cover all of that.
I specifically like the story about rare succeful armor upgrades that more likely on a certain map in a certain spot. Which would have been hard to believe until he watched someone do it. I cant remember the site but it will definitely come up if you start researching rng manipulation.
In testing even with good results you must always cast doubt just so you can be sure. What works one day may not work the next so months of testing is almost required.
Rng manipulation is cloudy and clear. Given a proper speriod if time if there is any substance it will be obvious.
The underlying theory here would have to be some flavor of a combination of periodic reinitialization of the PRNG, choosing a crap PRNG, and hard-coded seeds.
I’ll say the same thing I said regarding the claim that the perfect time to summon is between 0900 and 1000 UTC: Go for it. It’s certainly less messy than sacrificing chickens to improve your odds
I’m happy to summon in that time window. There’s literally no downside. I’m certainly not doing single pulls, though. Paying a 15% premium on the guess that maybe the RNG exhibits some degree of clustering seems like an unwise use of my money.
Yeeees. It’s a bit of a stretch, isn’t it? Still, it does no harm to collect a few data points. If the summoning is going to happen anyway then, as you say, there’s no downside to summoning at a particular time.
I hope we can collect a nice data set. In the unlikely event that it shows any advantage to summoning between *.35 and *.55, I will unreservedly recognise that the PRNG in use is flawed and open to exploitation.
