Pretty sure you can use binomial distributions, (as @mynamerr did) which can still be estimated with a normal distribution. IDK if that holds up or not but at this point I don’t care enough to check.
This is more what I was referring to, I wasn’t saying the conditions wouldn’t be fulfilled but yeah.
To put a more fine point on it statistically, we can calculate an estimate of the probability density of the p-value of mana dropping in both cases and see how likely it is that the same probability applies to both.
The standard error in the first case is sqrt(pq/n) = sqrt(10x40/(50^3)) = 5.7 percent, roughly.
The standard error in the second case (and this is not really estimate-able by a normal distribution because the number of non-successes is less than 10) is sqrt(7x43/(50^3)) = 4.9 percent, roughly.
The modal p-value for a mana drop in both cases would be the experimental p-value - 20% with overdamage and 86% without overdamage.
This graph shows the findings. The horizontal axis is a possible value of the real likelihood, in percentage points, to get mana. The vertical axis is something called the “probability density” which is technically not, but basically serves as, an indicator of how likely it is that a particular guess for the real drop rate of mana is correct. (E.g. if you shade in the area of the blue curve from 10 % to 20 % on the horizontal axis, that area is the likelihood that the real mana drop rate is in that range of 10 to 20 %)
The significant thing is, this isn’t a guess at the fact that 50 runs is “not enough to show anything”. This analysis (although not perfect) shows the data as it is, and there’s pretty much 0 (so close to 0, in fact, that the online calculator displayed it as a pure 0) chance that the probability for mana is the same, because, say, in the range 40 to 50 percent, both graphs are basically flat, and there is no chance that the real drop rate lies there and so it is almost definitely wrong to say that this data shows nothing.
At the very least, it shows that some factor changed between each set of 50 runs, which significantly impacted mana’s drop rate. To confirm @Mynamerr 's intuition, even though he used another method which I wouldn’t have thought of but seems mathematically valid enough to me, to come to the same conclusion, the chance that this change in drop rates is just an anomaly caused by RNG is practically 0.
but you could do 1 million cdepths and get 0 whites. It’s all rng it doesnt matter how many you do because you could do 100 and get 1 white and then do another 100 and get 7 whites. it all random
If you dont believe what a dev says. well good thing i can use YOUR proof to disprove your proofed checked conclusion with my proofere conclusion
Soo with the statistics your given us. in the first 50 Dungeons 20% of loot was blue bags, and 80% of loot was Cyan/pbags
while in the second 50 dungeons, a crazy 86% was blue bags, 2% was white bags, and 12% was cyan/pbags
now this might seem like overdamage, but i think your misunderstanding what the TRUE implications this is
this clearly proves Character loot seeds (or account loot seed) is a thing, which is a unique seed that every character has which detemines what loot said character can get and they chances. now this seed you proved is randomised at some point, which may be every 50 dungeons, or 50 certain lootdrops, or even a settime. that we dont know, but infact with your proof. we can say.
overdamage isnt real, but Character loot seed is, and Even kiddF confirmed that overdamage wasnt real. but he NEVER mentioned loot seed? you know why. because Deca is scared of people finding out and abusing the system
Checkmate Flat earther,
@mynamerr @chinadash thanks for the analysis both, though i’m inclined to agree with mynam’s reasonings that the data is heavily flawed or there are some other factors at play.
follow up question is: to my understanding both methods assume a constant p-value in each scenario. how would the discrepancy measure change if we know that for example, 1 mana each is guaranteed to 2 players who reach the sb threshold, and remaining players have a chance to get mana (with a possible cap to the number of players who can receive mana)?
You could probably just make one hypothesis that 2 random people (out of let’s say, 6) receive mana no matter the overdamage and the alternative hypothesis is that the probability is lowered due to overdamage (lower than 2/6). In this case, I’m guessing you would just disregard the results of OP’s testing of regular damage and only see if the overdamage was significantly lower than 1/3. It still is problematic though because white bags can still not have mana although hypothetically you’re still qualifying for for loot and you can still receive mana even if there’s guranteed drops that went to other people, which means even if there is a cap, the true probability is likely higher (like 2 mana guranteed, and everyone else with sb has like a 10% chance or something).
TL;DR: Theoretically you could, but the main issue here is that without a population proportion/true probability of drops, it’s extremely difficult to see how a certain factor impacts the values.
Sorry for the rambling, also, did OP ever state that the number of players remained the same through all 100 runs?
i guess it was implied
and thx for ur response but im just gonna wait for big brain people to compute an actual value 
My point was though, that we don’t know the amount of guranteed mana, we don’t know the probability of receiving additional mana, we don’t know how cyans and whites are factored into drops, and any exploration into those values would require probably hundreds of runs to accurately guess those initial values to even begin with, much less repeating those trials for “overdamage.” Didn’t make it too clear but yeah.
There is no need to “repeat trials for overdamage” or anything of the sort. You guys literally had a Deca employee reply to this thread stating with certainty that there is no overdamage mechanic. If people want to disagree and say that there is such a mechanic, no amount of talking about statistics will convince them otherwise. I don’t understand why this conversation is still continuing.
I think we can agree that OP’s data collection is flawed in some fashion. A video showing the runs would probably be necessary to figure out exactly what about the collection process is wrong.
Compiling statistics are fun and all (I love data!), but I agree with Furyborn and the couple others that have mentioned anything along these lines. I foresee this getting a little out of hand. With KiddForce offering his tidbit, it may be wise to have a similar conversation about drop rates, or even have a small group of you collaborate together on your tests and findings as it’s own thing, including if some of you still believe there might be something our there that’s affecting drop rates! But I think this specific debate no longer has much merit.
Same thing with doing abysses/sprites with for example 8/8 knight. Some 0/8s join and u get purple back every single time
I’m saying if you were to try and prove that overdamage exists, that’s the procedure and factors that you would have to take into consideration. I stated in my first post that I thought there was something probably flawed. Stop making it look like I believe this shit lmao
Honestly, I have been following this thread a bit. I agree with this analysis the most. I would probably throw out the 3rd bullet point though, that is, I think either the data is flawed or there is truly some over-damage “bug” that may not be obvious in the code to the current developers (that’s not a rare thing for software development lol, especially with code that has been managed by different groups in time).
Edit: On the point of the 3rd bullet though, actually yeah that is definitely a possibility I suppose.
Anyhow, but if the data is true and accurate, and there were no changes to mana drop rate during the tests or other drop rate influences from some event (I don’t know if anything influenced cdepths drops), I’d agree with OP that it is real lol.
I mean there could be a secret modifier which operates like a loot booster (or a loot suppressor) to increase or reduce loot chances with some mysterious factors, to encourage new/returning/weak players, or to give increased rewards at certain times such as in your first hour of logon.
If such a thing existed clearly/necessarily it would have secrecy around it because of the scandalous breach of good faith implications it would have.
What would be needed to test that, would be to repeat the test with:
a) using characters/accounts with different attributes (older/newer, lower/higher base fame)
b) noting the precise times of the loot bags compared to the login time.
Proving existence or non-existence of such a thing would be difficult. But repeated tests and more data is how you’d go about looking for signs of it. We shouldn’t flat rule it out existing, because who knows what shady practices go on in a game when money and gambling, and pushing addictive behaviour, is involved. But we should certainly view it with a large “prove it” mentality because Occam’s Razor prompts us to look for the simplest explanation (like bad data, or bad assumptions, or pure bad luck).
I’m gonna level with you. One of the very first things I did when I joined Deca and received full access to backend code was looking at the loot system. As a long time player, I had to know if any of my random skepticism was true.
I realize my word doesn’t have any more weight than Kidd’s if that apparently can’t be trusted, but the plain truth is that the loot system is… well, plain. It’s got the basic parameters defined in the XML like probability, threshold, mins and maxes (which tend to skew data like this since there’s no way to determine if the item you got was from a guaranteed roll or a post-guarantee roll, especially with potions), and that’s just about it. No secret mechanisms or modifiers.
Hell, the fundamentals of the loot system haven’t budged since pretty much the inception of non-public drops, so even the majority of pservers based on 2012 code should be using exactly what’s used today in the real game. Any sort of trickery would have long been exposed if any past owners of the game had tried to conceal some facet of it.
This is now from the mouth of two witnesses, and these were players like us once upon a time. And in my opinion, considering they took the time to even bother enlightening us shows that they at least harbor some minimal care for us, right?
Time to complete the deca designer trinity!
I’ve also checked for overdamage, and I can also say with complete confidence:
Overdamage is definitely nonsense. The entire loot computation checks the damage you have dealt to the dropper once: To see if you have met the threshold value.
Apart from that, the value is not used.
min and max constraints on loot do not break the system either: The list of qualified-by-threshold players is shuffled before any handing out is done.
So unless anyone wants to take a chance and hit up the creator of C++ and convince him the random function is broken, I think there’s not much more to be gained from this thread.
