I feel like regardless of DECA proof, overdamage still exists to a lesser degree, relative to other people’s DPS. Of course, this is all down to my experiences palying realm, of which most of you have likely surpassed years ago.
'Overdamage is a myth'
BaadGuy
#22
I’m not sure what you even think “overdamage” is.
If your client tells the game you did impossible damage after the monster is dead, it will just ignore it or log an error - with you being in a group or not.
If you’re telling us more damage over the SB thresholds equals more loot, that hasn’t ever existed.
The only way you can think of an “overdamage” scenario nowadays is someone doing so much damage to a monster that other people don’t have time to qualify for SB, thus stealing their potential divided loot. But SB values have been balanced so this practically doesn’t happen
Nameness
#23
You literally just got told by a guy who has seen and read the code that it doesn’t exist in the slightest and you, who hasn’t seen it in the slightest, just gonna say he’s wrong? Like what?
Zquidx
#25
i did the same tests as OP and got the exact opposite results so i guess that balances it 
BaadGuy
#28
You obviously lying since you didn’t write random numbers and percentages to make your text bigger
Get goodnerd
Xaklor
#29
that is called a superstition. any game that features a major RNG-based game mechanic inevitably develops superstitions, like overdamage and kendo stick/fire sword. overdamage is objectively false on every level (this thread proving nothing as the sample size is pitiful but that’s been hammered in quite a bit already) but people still find reason to believe in it anyway through confirmation bias and the overwhelming urge to find ways of beating an infuriating system even where they don’t exist.
Twitchystr
#30
Definitely.
At the beginning yes, but it just feels like too small an increase to bother doing less damage than before.
xRetz
#31
Compared to the other people who didn’t test whatsoever and just went off of what they think to be the truth, my tests could be considered in-depth.
xRetz
#32
You did 100 cdepths in a row, with the same 6 people, in a controlled environment where the only factor that changed was the amount of damage you did to the boss each run? I find that doubtful.
xRetz
#33
How does ‘it’s just rng’ explain how I consistently got low-tier soulbound loot when overdamaging, and consistently got higher-tier soulbound loot when not overdamaging?
xRetz
#34
This is confirmed a conspiracy, trying to cover up the truth? Smdh.
You can’t fool me, I know the truth.
Orsome
#35
Your “in depth” testing could still be a result of random chance, regardless of the results, so they don’t prove anything, just imply that there’s a possibility your theory is true. You can’t just deny actual proof within the code because you don’t like that it goes against your idea. This isn’t like real life, code does exactly what you tell it to do and if Kidd says the code doesn’t do anything in terms of overdamage, then it doesn’t. It’s as simple as that.
Seelpit
#36
I can second this - there is no parameter in the code that’s like, “ifDamageExceedsThenLowerDropRate”, so unless it was somehow hardcoded in - which is something that I can’t check of course, but I presume Kidd can - there just…isn’t any proof XML-wise.
Unicorn’s post is just the only real type of “overdamage” in the game.
Mynamerr
#37
…except when you do the math, it actually is completly sufficently. The chance of getting a discrepancy this big at 50 runs is <0,0000011, which I think is very safe to conclude that something is going on or the data is flawed.
I however don’t think Deca secretly using overdamage to prevent people from bringing alts makes sense when they deny it and no one thinks its a thing. Possible explanations would be
- OP miscounted
- The data is somehow flawed, e.g. in a significant portion of the low damage runs some of his friends nexused
- Deca changed the mana drop rate at some point and the runs were affected by that
Or maybe there is a actually a bug with the drops going on.
1On the calculation: You might be wondering how the chance can be estimated if we don’t know the drop rate. However what we know is that the drop rate will always be rather far of the 43/50 or the 10/50. We assume that the drop rate of mana is p≤0.53 and consider the 43 in 50 mana drops (if the drop rate is lower, we get the same result by considering the 10 in 50 drops instead). Now you might know from school/college that chance of getting k drops in n runs is given by the Binomial distribution. We notice that the chance of getting 43 in the binomial distribution is the highest when p is maximal i.e. =0,53.
Now we can calculate the chance of getting 43 or more drops in 50 attempts at a drop rate of 0.53 using Wolfram alpha:
(You might have realized that we’re throwing away half of our data here and our bound should be far from optimal. I suspect that we should be able to go down the square of that number, but I’m to lazy to work that out unless anyone seriously thinks 0,000001 isn’t unlikely enough.)
PhantomMod
#38
Because…that’s literally what rng is. If you want your results to be an impression of what the majority, or even the entire playerbase is experiencing, you’ll need a sample size much greater to compensate. As I mentioned before, experiments around RNG can only truly be accurate if you test it a sufficient amount of times. I mean, I’ve probably defeated Xolotl about 700 times by now and I have never gotten the penetrating blast spell, do I conclude that this is the rarest item in the game, disregarding the results from the other thicket hunters?
The more you roll, the higher the chance that your final cumulative result is closer to the actual percentage of rolling it, and sometimes, the amount of times you have to test will be extraordinarily high, this is a fact you have to accept. You cannot test 2x50 cdepths and say it is an accurate representation of the entire population of players on the game, when in one day, I am sure the loot of Cdepths has rolled for 1000’s of other players.
At the end of the day though, a dev has already confirmed that this is a myth. But as Xaklor said, you are free to believe in this superstition if it helps you to enjoy the game better.
Gidey
#40
Although Kidd already closed the matter, i wanted to point out two things:
- @starwartwo i dont think you would use a z- test for this, as z-tests are for continuous data, while loot drop is bernoulli (there is a drop, or there isnt) I cant remember what the test you would use for this is called, nor do I remember how to do it.
- Contrary to popular opinion, although 50 is too small a sample size to estimate the drop rate accurately, it is large enough to do these types of statistical tests.
Oh im mostly considering the pots for what you would test with, by the way. (that is, 43/50 in one set, and 10/50 in the other).