Tree of Savior Forum

[DropPerKill] Drop rate system- DPK- Explained [Updated]2

No they don’t. For an item that drops roughly 1 in every 200 kills, it’s statistically likely for you to see about 100 of them in 20,000 kills. However, it’s nearly a statistical impossibility to encounter 100 of them dropping every 190-210 kills unfailingly and never dropping within 150 kills of eachother.

People who’ve farmed for things like the white boater at 100,000 kill counts, or the combined reports of numerous people farming for monster gems, who all report the same patterns, are a large enough sample size when dealing with an event that is so astronomically unlikely in a normal RNG system.

Just a reminder that the drops listed with % chance values in tosbase.com were data mined when the information was still made available on the client side (this has been change for obvious reasons). This is also why the information found in tosbase.com is limited up to Lv200 only (with a few hits and misses above that level). Following this, the drops they listed with a ? are the ones we’re assuming to be DPK, and to a certain extent, have been proven to be such.

Orsha was originally going to be a high-level town (that’s why it’s surrounded by mostly high-level maps as opposed to Klaipeda) and the information found for Orsha based mobs/drops are a bit iffy due to these changes.

I understand a lot of people are dubious of the findings presented in this thread and that’s what we need. We need curiosity and patience. People curious will try it out. People that are doubtful will try to disprove it. Either way, both will serve to further the information provided here.

Before anyone comments further that they got a rare drop back to back, you have to understand what is defined as a) RNG-rare vs B) DPK-specific. RNG permits you to have back-to-back (hell to infinity if you may) drops of rare items. Whereas DPK will only allow an item to drop specifically at the moment its threshold is reached.

The distinction has to be made first before anyone else tries to prove/disprove the information in this thread.

1 Like

I farmed at Sicarius 1F near 2F warp so maybe it’s different. I find it strange that even at .04% you can get 4 cottons in under an hour considering my farming session from yesterday. Not complaining though!!

@PWICE This is (assumed to be) DPK [unless listed incorrectly]. So if possible, please conduct a test. :slight_smile: Also let us know if you were ever able to get two to drop right after the other.

The Shaman Doll Cotton listed with a % chance to drop is RNG (Mage Tower). Hence, as I mentioned, you are inherently more likely to experience it occurring numerous times (even back to back) in a specified timeframe.

Whereas the flip-side of this drop is the ? chance which is DPK. Thus, a specified number of Shaman Dolls have to die first before it drops.

You were farming at the DPK ("?" for Sicarius version of Shaman Doll according to tosbase.com) side of this table. The consistency you felt, given that you were quite likely killing the mob at a generously fixed rate and with no one else interfering (since this place is rarely populated), then you would have reached the DPK threshold for the Shaman Doll Cotton 4x in that less than an hour time-span.

If you could, next time, provide numbers for how many you had to kill before the first one dropped (baseline). How many you had to kill before you got the 2nd one (number difference) and so on and so forth. This is where the pattern will start revealing itself.

[JUST AN EXAMPLE] If it consistently dropped, say for example, after every 50 (+/- for variation) mobs killed in a solo channel; then the DPK for Shaman Doll Cotton would be every 50-ish in the Sicarius version of Shaman Doll [JUST AN EXAMPLE].

Additionally, the thing I dislike about DPK ("?" chance) is that it will NEVER allow you get two of the same item in quick succession [back to back if you may] (in a solo environment with an item that needs a high DPK threshold to be met).

For reals? Then thats a relief. So i don’t have to worry about some random coming on to the channel im hunting and snipe my drops.

The evidence is entirely inconclusive though.

For example I can objectively say that the Hanaming evidence through this thread really doesn’t align with my experience, and I’m willing to bet I’ve got more kills. I have ~50k Hanaming kills all farmed during the dead of night with 250 people on the server on an empty channel around 2 weeks ago, now I’m number 3 Hanaming killer on Fedimian behind the 2 bots in number 1 and 2 position, so anyone can easily verify the claim. I’ve still got 4 in storage.

The experiences they’re bringing up don’t really align with my own at all. You’ll get two gems within a couple hundred kills, or you might go 20k kills with nothing. I’ve had 6 gems and believe the drop rate to be a 1in10k chance, it’s 100% not DPK based.

So either the argument people are making needs to be scaled back to not claiming that it applies to all gems, or some other theory needs adjusting.

This is why I’m so adament that these are all superstitions. I’ve logged 1500 hours now, I’ve played with the system a plentiful amount, I see unusual things happening now and then but nothing that I can’t actually chalk up to being purely random oddities that could occur at any time.

RNGesus smiles on us sometimes, and he hates us other times.

I’m entirely willing to change my opinion when someone provides something that doesn’t sound shakey. I’m really only seeing shakey anecdotal things based on 1 drop though, or other things that others are often refuting. I’m not denying the function as a matter of just arguing, I’m denying it as a matter of it not being very logical from a developer pov and not really having much evidence to suggest it exists other than anecdotal shakey pattern stuff and what it “feeeels” like.

EDIT: I know you already know this Goat, but for the sake of others - I should also add that I know the population counts on channels I’ve farmed on with 100% certainty. Was never interrupted in my times.

We use an addon to collect this information in my guild. It’s primarily for worldbossing but it comes in handy for looking into drop mechanics that could be manipulatable.

I’ve discussed it on the forums elsewhere here:

4 Likes

@Awoooo It’s been suggested here that every spawn… entry(?) has its own DPK value instead of every monster type.
This would explain why countless people have reported finding Hanaming gems back to back but never an Infrorocktor gem or other monsters with a single entry.

Two entries for Hanaming:

@PWICE .04%? I’m assuming you meant 0.4%.
I dropped 11 Leather Bangle (0.25% chance to drop) until I dropped one single Shredded Piece of Cloth.
4 Shaman Doll Cotton in one hour at 0.4% isn’t that unlikely considering the decent spawn rate of Shaman Dolls.

@DiMeowgio I think it’s just listed as “?%” because it’s a newer map but still the same chance as the one in Mage Tower.
I didn’t notice a pattern when I farmed there but that was around two months ago so I could be wrong.

In that case the people building theories need to also know exactly which Hanaming were part of my farming:

I haven’t checked, is this spawn area a combination of two Hanaming spawns for the zone or just one? Farming this section can occur on the map without any need to channel hop or travel the rest of the map so my assumption is that it is one complete spawning area.

I would guess this because normally if you only farm one small section of the map then the spawning “entry” will stack up somewhere else on the map until you see greatly reduced spawns for the mob you’re farming because there are too many spawned elsewhere. That doesn’t occur when farming this circuit back and forth as drawn. The black scribbled out bit oddly has no spawns in it at all.

I feel a little sad knowing such a small area of the map so intimately. But others can test for themselves to verify.

Now, on another note. If I were to draw up an experiment for something like this then I would suggest taking a 3-5 person team, logging the total kills of a mob and all the drops, then hitting a channel together and logging all the drops they receive. With data like that THEN we can start drawing conclusions on the nature of the drop system in relation to drop per kill or drop per spawn. One person can do this too, but the data would be relatively quick to gather with a full team.

If I was going to guess at a mob that might be suitable to do this kind of testing on I’d possibly suggest farming for shredded cloth on the later maps in the game where their spawn rate is excellent. I can’t recall the map name off the top of my head right now but I’m sure people know what I’m referring to.

Over 12k venucelos kills now and still no sage wall.

I don’t doubt your experiences with the Hanaming grind; all I can say is that grinding 90,000 (102,000 total) infrorocktors within a two and a half day period is all the evidence I need. Whether anyone believes me or not doesn’t really matter. I’ve proven it for myself and that is good enough for me.

Regarding what SlyGoat said, it exactly matched my experience with the club and rod drops (although, they require slightly less kills), but you can predict when one will drop within a 15 kill margin of error, and that is the 5-10% variance that is often experienced with DPK.

As DiMeowgio said, both drop systems are in the game and it all depends on the items being farmed (and the mobs’ spawn locations). I’m not sure why IMC would put both systems in the game, but perhaps it was their way of trying to mask DPK.

Anyway, I understand your need for evidence, and here is some work done by @Ingel at the end of June.

dont sweat it, i have 18k with no drop

But but but… The point I’m making is that one person’s anecdotal experience isn’t enough when so many experiences are all colliding with each other in opposing ways.

People don’t get to disregard the data of everybody else just because their data when viewed in a vacuum confirms the theories they want to believe. That’s not good science.

I just farmed some Melinas Staff for Audra, Acolyte shoes and gloves drops alternately. it dropped alternately, the same part never dropped twice at the same time/in a row. Yes using LKChannel to confirm i’m alone on the map.

So far I’ve seen no testing done from the opposing side, just “yeah but I got two monster gems once so you’re wrong”.

Technically if someone got 2 monster gems in between a small amount of kills in the same channel, the DPK theory is wrong. Nothing more is needed.

I totally agree with you from a developer’s perspective. Unless they specifically wanted this system, there is no reason to spend extra time to implement DPK along with extra tracking of each item’s current drop situation. That’s also more memory needed on the server.

That’s all, not saying there’s no DPK but I feel there’s no real way to do heavy testing on this since no one has the time to continuously farm something for a super long period of time.

But Shredded Piece of Cloth have a drop chance of 0.1%, so no DPK value.
Wouldn’t you want to test it on an item that has been confirmed as DPK item?

For example:

From my experience, these gloves have a DPK value of 100.
Of course, it’d take a while to farm enough of those to have convincing statistics…

If you don’t want to try it on Infrorocktors, how about Pokubu in West Siauliai Woods?
They only spawn in a small area, respawn fast and drop Crude Short Bow (?%) and have a few other drops that are RNG.
Archers killing them will increase the DPK count by 2 instead of 1, so one could set up a team consisting of two archers and two mages which each member on their own channel.
The archers should drop it twice as fast.

1 Like

I completely understand what you’re saying.

These conflicting viewpoints between RNG and DPK come down to the fact that both drop conventions are in place. I hope that nobody truly thinks that DPK is the only drop system that exists in this game, and I really doubt anyone would make such a claim. This thread’s original purpose was to show the rules for DPK item drops.

Now, I don’t have a problem with people denying DPK’s existence, but I’m mainly just dissatisfied that this thread has become an arguing ground for the past couple of months (although, it is calm currently). As you’ve said, anecdotal experience isn’t enough, but I believe Ingel’s results are pretty clear from a DPK standpoint.

As lordshredder said, to see if DPK really exists–or to attempt to disprove it–it would be best to try to do this on items that have already been confirmed as being DPK. I can understand wanting to try this with Hanaming gems, but there isn’t enough gem data across multiple mobs to ensure that all gems are DPK (I can’t rule this possibility out, at least). I only know that infrorocktor gems are DPK with certainty.

With these low drop rates and extremely low spawn. Nobody on the server is selling the item you are after. You spent so much time trying to hunt it like a no life during your days off. Doesn’t that make alot of people feel like quitting?

This system doesn’t promote casual farming. In this specific situation, Mausoleum F4 and F5 are maps that usually get people doing quests or looking for Cerberus.

The first time i farmed Venucelos, at 1.5k± i lost the Sage wall recipe to someone else, decided to farm it no stop, got it like 8 hrs later, and almost got ks’d again! Same experience when farming Medakia Metal decoration.
My only advice if you can’t farm longer than 6 hours straight, is try out your luck on weekends, and farm on a specific spot, if you farm the whole map, the recipe/item will most likely drop anywhere, therefore, more chances to get ks’d. Farm at night and channels you know people won’t show up.
And the most important: Don’t give up.

1 Like