Question for the MFO'ers: Peculiar drop patterns...

Z

ziambe

PurePremium
With the closing ceremonies of the Autumn MFO still fresh in everyone's memory and the preparation of our Christmas MF wish lists just around the corner, I wanted to take a moment to ask some of our finest MF athletes the following question:

During periods when you were running the same area several hours a night for several nights in a row, have you ever noticed that a single item will drop suspiciously often one night than in comparison to the rest of the total time you have committed to that area?

Here is the background info, which after too much coincidence, has finally pushed me to seek (medical) help. :D

So I have been dedicating a lot of time to the pits lately (would've signed up for the MFO but missed it by a couple days), around 4 hours a day over the past couple weeks and, call me crazy, but I seem to be getting what I have started to name the 'drop of the day' almost every night.

That is to say, a particular item seems to be dropping suspiciously often...way more than you would expect in a such a time frame in a place that can drop every item in the game...and then just for that night.

The first time this caught my attention it was Cryptic Axes, not just uniques, but 6 or more in one night whereas 1, or more often none, every other night. Since a Tomb Reaver is such a great find, I think this definitely influenced the attention I paid to this particular item, but in thinking about the fantastic fact that I found 2 in one night along with a bunch of other blue cryptic axes, the next day I paid a little more attention to the frequency of each item and low and behold, it was Shadow Plates - no joke, 3 Aldur's Deception's in one night (along with some blues/yellows) and no cryptic axes. Then next day Ogre Axes and the next, Myrmidon Greaves...

I started to think about this, wondering how/if I could ever prove it or at the very least try to disprove it with any of the information on item generation and thinking to myself 'is this even worth posting' .... they will think I am crazy. But then tonight after Spired Helm after Spired Helm after Spired Helm dropped I had to login for the first time in years to see what you guys think...

One fun theory I came up with is that I seem to remember that your CPU's system clock played a roll in map (and maybe item) generation in Diablo 1. Maybe Diablo 2 has been out so long now that the system clock, if somehow secretly involved in item generation in D2, could be producing such strange coincidences...

In any event, thanks for reading and I'd love to hear what you think!
 
Last edited:
I have noticed this apparent effect for years, but not enough to prove it. Your mind looks for patterns in randomness, maybe it's a case of that. I figured someone would have proven or disproven it by now. Hasn't the game code been torn apart completely?

I've always thought it would be interesting to run D2 on some kind of virtual PC where you could save the entire state of the PC and roll back to it just like an emulator save state, do it just before a boss drop, and try slightly different things to see if you can see patterns in the RNG. I think if we could outright manipulate the RNG it would be cheating and make the game useless though.
 
I have seen it. Running Pindle, I had four or five consecutive jewel drops, one unique, the rest rare/magic. He will also "echo drop" where an unusually rare item (i.e. Tals Armor) will drop again in just a few runs.
 
I had a peculiar situation with stormlashes. Not one dropped in over 10 years and then 3 in a matter of days.. apart from not fitting into "item of the day" theory, I was running AT like crazy at that time and stuff was dropping left and right so.. might just be good old RNG with our built in pattern recognition :)

I'd bet on it, if it wasn't for 1.07 racks, 1.00 cow king patterns, 1.10+ (?) council spawn patterns..
 
Does finding a Mang Song's Lesson on back-to-back runs qualify? I tend to think it's just a result of some innate MF that my IRL person has that the game picks up on.
 
Often? Sure

Suspiciously often? Nah.

LC, back-to-back is nice. Finding two Mang's in the last three hours was what got me me the win in the 2013 Winter MFO, but back-to-back is even crazier!

In that same MFO, I had three Griffon's drop in the span of an hour or two. Easy to find patterns when you look for them, especially after the fact.
 
LiquidClear said:
Does finding a Mang Song's Lesson on back-to-back runs qualify? I tend to think it's just a result of some innate MF that my IRL person has that the game picks up on.
Click to expand...

Now there's a thought I never had before. Maybe I should buy some topaz jewelry IRL...
 
I know that we are psychologically programmed to see patterns where there are none. But also computers are not very random, I would not expect a game of this age to have a great RNG, so maybe there is something in it. But given the relatively small sample size we run and the complete lack of real data, I think it is mostly psychological.

On a related note, are some map seeds more likely to drop good items than others?
 
Funny that this thread is here because ... yes, in fact I have had the same impression during this MFO. Sure, the human mind tries to find patterns in everything, but in terms of blue, green, yellow and golden items with a base like Unearthed Wands, Archon Staves, ... it's definitely more than that.

So yes, I definitely have the impression that the drops aren't 100% random. As most of the random stuff on computers isn't actually 100% random (check random.org for detailed information on this). When I set my Spotify playlist on random, one day I also get a whole lot of band X (or even a specific album of this band) and the next day lots of band Y. In the same vein, I had lots of Scourges on one day of the MFO, Hydra bows on the first run or runs. I didn't get a single unique ring during the MFO, and very low amount of rings overall, except for the last day, when it seemed to rain rings, resulting in 3 uniques in the 2.5 hours I played then.

So yes, I'm really sure that a certain item drops a lot one day, while another drops a lot the day after. I don't think there's necessarily a certain pattern involved though, like there is in LK. I'm inclined to think it's just a matter or random not being entirely random.

O yes, and several Griffon's and Mang Song's in one day? I have never found several of those to begin with, let alone in one MFO or even better in one day. Some guys have all the luck :).
 
While there very well could be something weird going on I doubt it.

It seems that we often don't grasp the nature of randomness, really. All kind of suspicious pattern are exactly part of it. If everything was dropping without very random streaks of crazy incidents, THAT would be suspicious!

There's a story in a book called "Random walk down the Wall Street" how a professor would start this one course with a task that every student imagines fifty coin throws and writes the pattern down. They give their patterns to a professor, but one was asked to truly throw a coin and write the results down as well. Guess what, professor can always pick the pattern that was made with real throws. People are so poor to mimic true randomness, probably making it too even etc. That applies also when we look at randomness; we see patterns and think that can't be random.
 
Somewhere, far far away in a different multiverse, my BlizzSorc is running Pindle, and all he drops is Deaths Fathom, and all she want's is The Oculus for that extra bit of mf.

I don't disregard this drop pattern hypothesis, as I feel I am often (I even restart my PC from time to time...) a victim of it, but i think it's more a psychological thing. Some uber things keep dropping, while we are looking for something else, which NEVER drops. And I would say of course we keep an eye out for the Unearthed Wands and Archon Staffs, because they are so rare, but we don't notice the Heavy Gloves or Sallet's that could just as well drop in this "pattern". And If they drop like this, when does it just become random, and when is there a pattern at work?
 
I don't completely agree. As it's not a matter of Archon Staffs vs Heavy Gloves. In that case, it's indeed because Heavy Gloves aren't what we are looking for.

In the example I have given above, with Hydra Bows dropping left and right one day and Scourges another day, do you really think that both are dropping in about equal amounts on those days, and I only notice one but not the other, and the other day it's the other way around? I don't think so.
 
T,

No, what he (presumably) thinks is that Hydra Bows and Scourges drop equally often*, just that one happened to drop more often on one particular day, and the other happened to drop more often on the other particular day. Which is bound to happen, because randomness. Think series of coin tosses, and the results being HHTHHHTTHHHHHTHTTTHTTTTHTTT, or whatever.

* Not actually equally often as I'm sure they have different drop probabilities, but you get the point.

Edit: Also, it's kinda weird to have the exact same conversation going on in two threads simultaneously, started by the same poster no less. I wrote a pretty lengthy post about this in the other thread which might be relevant here too.
 
Fabian said:
T,

No, what he (presumably) thinks is that Hydra Bows and Scourges drop equally often*, just that one happened to drop more often on one particular day, and the other happened to drop more often on the other particular day. Which is bound to happen, because randomness. Think series of coin tosses, and the results being HHTHHHTTHHHHHTHTTTHTTTTHTTT, or whatever.

* Not actually equally often as I'm sure they have different drop probabilities, but you get the point.

Edit: Also, it's kinda weird to have the exact same conversation going on in two threads simultaneously, started by the same poster no less. I wrote a pretty lengthy post about this in the other thread which might be relevant here too.
Click to expand...

i noticed the exact same thing :) 2 posts i call for a merger!! (or am in in aqcuisition modus to much?)
 
Fabian said:
bound to happen, because randomness. Think series of coin tosses, and the results being HHTHHHTTHHHHHTHTTTHTTTTHTTT, or whatever.
Click to expand...

Except in this case it's not a 2-sided coin with only H and T. It's a whole lot of base items, then exceptional/elite applied to them. HHH then TTT should obviously be expected to happen a lot. Mithril Coil Mithril Coil Mithril Coil then Lacquered Plate Lacquered Plate Lacquered Plate should not. Only a real in-depth analysis can dismiss it.
 
Hi all, thanks for your insight and comments!

Since the discussion seems to be brewing, let me add a couple details, (hopefully without getting too wordy! :)):

The variable which I am paying attention to here is the day, not the item. In other words, I am not comparing the drop rate of (in my examples) Cryptic Axes to Myrmidon Greaves or any other item. I am simply observing the number of cryptic axes which showed up on a particular day, with the number of cryptic axes which showed up on any other day. If this particular item and this particular day was the only time I ever noticed it, I would have sided with my reluctance to report it and would not have posted anything. But what I find interesting is that I have been able to make the same observation multiple times using different items with respect to different days.

Think of it as conducting an experiment several times (of course these observations are far from an actual experiment) instead of the conducting one experiment with several results.

The reason I made my other observation a separate thread is because it has nothing to do with the process of item generation and everything to do with randomization. While randomization is a large component of item generation, item generation ultimately follows a much larger list of non-random rules (and these rules are well-documented in the wiki page Item Generation Tutorial). Just as I would have sided with my idea not to post if I had only been able to make the observation once, I would also not have posted this idea if I felt that there was any information in the wiki which could explain this (or my other thread's) phenomena.

The focus of the other thread is simply an observation that: an item which has only 11 possibilities of a stat (enhanced defense) and which drops enough to provide lots of data (100s of chain mails), should produce a perfect version of that stat much more often that it seems to. Again, if Chain Mail was the only item which I felt was doing this, I would have dismissed the idea. But I noticed it on other items as well: War Axes, Scythes, Pikes etc. Also, I want to be clear that I am not talking about the combination of a perfect roll on numerous stats, only a single stat. And that this observation (rolls on an item) relies on very common (normal) items, whereas the items I am talking about in this thread should be the rarest of the rare (ie. TCs 81 84 and 87).

My goal here was to invite the community to share any thoughts, ideas or even actual data if they had ever had a similar notion but, due to their knowledge of probability or the programming of D2 did not act on it, not to look for reasons about why it is improbable.

Think of engineers trying to design the first airplane. If you had the opportunity to meet all of the world's best engineers at once, wouldn't you want to share ideas about how an airplane could work instead of how it couldn't?

Indeed, and as was mentioned by BKC, no one was looking for the LK pattern until data appeared (screenshots of the same superchest items), yet until the screenshots were available it was easy to disprove that idea; the challenge is to look for the data that could prove it (or perhaps add new information to the item generation wiki).
 
Last edited:
The game knows what you are looking for and then drops something different.

It is also intelligent enough to know when you try to bluff it by saying “I’m going to run the Pit and I really hope I don’t find Tyrael’s Might.”
 
Log in to Reply / Create Site Account
Remove Ads - Go PurePremium

Latest posts

Back
Top