------
Edit: to explain, "Rolling a die" is basically throwing a dice. Dice is the plural term, but you get the idea. If you don't know what a dice is, it's (usually) a cube with a number of dots on each side corresponding to a number. Opposite sides of a 6 sided die add up to 7. Still not with me? Well here's a nice little picture, courtesy of a google search
http://www.thinkport.org/microsites/...mages/Dice.jpg
------
I've seen, both in game and on forums, people complaining about getting 3-5, 5-10, or 10+ bad runs in a row. This applies mostly to UW farming, and I have experienced this myself too.
Just because you have had a string of X bad runs in a row does not mean that Anet must have nerfed the farm. I'll try and explain, using the drop rates of Globs of Ectoplasm, since they are the biggest source of complaint.
Let us just assume here. I don't know the exact % chance of drop rate for an Ecto, so I'll just make one up. It drops fairly rarely, so let's use 2% as our chance of getting an ecto. I'm also going to totally ignore loot scaling because, as you probably know, it doesn't affect Rare Materials.
Before we continue, I'd like to clear up a common misconception:
A 2% chance to drop an ecto does NOT mean that, for every 100 Smite Crawlers you kill, you will get 2 Ectos.
Now lets set the scene. You're in the underworld running a W/Rt build. It's your first run, so you've only pulled one group of smites. In this group are 5 Smite Crawlers, and you're just that good that you pull it off without a problem, killing them all. To make it simple, we're going to call them A, B, C, D and Bob.
A has a 2% chance to drop an ecto. Do you get one? Well let's be fair...let's pretend we're rolling a die with 50 faces, only one of which, has a picture of an Ecto on it. The other 50 are money, Holy Rods, etc. If you get an ecto landing face-up on your die, then congratulations, A has dropped you an ecto. The probability of you getting that Ecto-side facing upwards, out of a whole 50, is slim (0.02 to be exact). And ouch. Tough luck kiddo, you just got a blue Holy Rod with an "Improved Sale Value" mod.
B is up next. Again, you roll your 50 sided die (because you know, the Smite run is unimportant, and you've now decided to stop for a second and roll dice around with the Reaper). This die roll is totally independent of the first. This means that, no matter what you just rolled, Ecto, gold Stormbow, whatever, this die roll has no reason to be the same or different. It may be easier to think of it by saying that you took 5 50-sided die with you into the UW, and you're rolling a seperate die for each Smite Crawler you kill. Again, sorry kiddo, no Ecto. But you did just get a r12 Gold Tactics shield to ID on your Mesmer later.
Now it's C's turn. Out comes your third die. You might know what's coming up here. You roll it, and wow! It's your lucky day! You got an ecto. Against the odds, you managed to roll the 50 sided die with the side you wanted landing face up. Again, as you can probably see, this is totally independent of the previous rolls. Pocket your Ecto, and let's roll again.
D didn't look very nice, and your die lands on Phantom Residue. You can scrape that off your nice shield later; surely someone out there will want Crawler Remains for something.
Finally, it's time to roll last your die and see what Bob has dropped you. Bob is a very unique Crawler, as you might have guessed. And low and behold, your fifth die lands on the ecto. I'd say you're having typical beginners luck, because I've never had the first group drop me 2 ectos. But let's look at how this is possible. You've rolled a die, independent of the other 4 (sorry for repeating myself guys) and got an ecto. Each Smite has a totally seperate "roll" for dropping an ecto, and this can lead in 3 directions; occasional ecto-drops/average run, lots of ecto-drops/good run, or no ecto-drops/bad run. Now I bet you wouldn't run over to guru, tab across to alliance chat, or scream in all capitals in ToA (apologies for stereotyping) "SOMEONE HAS NERFED THE SMITE RUN!" Why? Because it favoured you. Infact, I'd bet you'd whisper the poor guy who hasn't got any ectos in 17 runs and say "Hey, I just got 2 ectos on the first group lol! They didn't nerf the run"
Now you can apply this up for every group of smite crawlers down there. I don't know how loot-scaling will affect the idea of rolling a die, so you can just pretend that some of the 50-faces of the die are blank. I know, that's not how it works, but it might help you to think of it that way because you'll still have 1/50 sides being the one you want.
So yes, this means that it is possible to get every Smite Crawler dropping an ecto. But, and this is assuming there are 40 Smite's down there, the probability of this happening is 1.09951163 × 10^-68.
To those with no mathematical knowledge, that is 0.000000000000000000000000000000000000000000000000 00000000000000000019951163 (out of 1). So yeah, congratulations; if you make that happen, and pull off a 40-ecto run, you are probably the luckiest man/woman/other alive today. You can settle down, start your average family, have your 2.4 children, and live the rest of your life knowing that there is almost nobody else alive luckier than you . Or as they say here on guru "Screens or it never happened."
So you could go through one run after another, getting the rolls you want, and walking away with 6 or 7 ecto after an hour of farming. Or you could have a couple of runs where you don't get that lucky side of the die, and you get nothing. And there is nothing in the game that is going to feel pity for you after you've had 16 of these bad runs and give you an ecto. Sadly.
I hope this post has cleared up some of the misconceptions of bad runs coming back to back. Constructive feedback is welcome. Flamers can go find some other thread to build their bonfires of idiocy in. And remember guys; try and trust Anet at least a little. They would most probably leave an update-message saying something along the lines of "Lowered drop rate of Glob of Ectoplasm from Smite Crawler, Mindblade Spectre, Bladed Aatxe". I mean sure, they might not, but like I said. Have a little faith.
----------
Sethellington and Masao raised an interesting point. Skinny Corpse and his brother did some experimentation into the theory that drops are assigned at the beginning of a zone, and they went on to prove this. It was a very interesting test, which you can read about here (credit to them on their work). But as Masao said, this doesn't affect the theory I have put forward here.
...all of the drops are set upon entry of the explorable area. By entering at the exact moment in time that another person does, you'll get the same drop table and thus each mob will drop the same as the corresponding mob in the other person's map. Think of it as rolling all the dice at the beginning of the map, rather than when you kill the mob. Marty Silverblade
People say 'Ill get 2 ectos for every 100 smites' because it's what should happen according to chance. Granted, it may or may not work out that way, but if you took every smite that has ever been killed, it'd be close to 2%.
MrSlayer
Quote:
Originally Posted by Marty Silverblade
People say 'Ill get 2 ectos for every 100 smites' because it's what should happen according to chance. Granted, it may or may not work out that way, but if you took every smite that has ever been killed, it'd be close to 2%.
That's exactly it. A 2% chance doesn't mean that every 100 smites = 2 ectos must happen. It means that's the average. It variates around that 2/100. Like some people could get 19 or 20/100 smites dropping an ecto, whilst those people who consistently get 0/100 bring it down again.
People can't just say "my run has been nerfed" when they don't get that 2/100 ectos. That's my point here. gremlin
I think you cleared that up quite well, not that it will convince the farmers out there.
Luck has no memory even if you toss a coint 10 times and it comes up heads all 10 times the chances of it coming up heads or tails remains the same for the next time. You would probably have to toss the coin many thousands of times before you saw a 50/50 split heads to tails. And you would have to add together all the runs by all the players in the game before you would be able to work out the true odds for an ecto drop. upier
(Found this as I was searching for the term "to roll a die". Non-native English speaker here and I had never heard it before.)
http://espse.educ.psu.edu/edpsych/fa...7/dieroll.html http://espse.educ.psu.edu/edpsych/fa.../coinflip.html For the kids among us that dislike long words but prefer games! Btw - just managed to get 5 ONCE in my first 29 tries! I guess the dice caught my GW luck! MrSlayer
Quote:
Originally Posted by gremlin
You would probably have to toss the coin many thousands of times before you saw a 50/50 split heads to tails. |
Quote:
Originally Posted by upier
...as I was searching for the term "to roll a die". Non-native English speaker here and I had never heard it before...
Apologies. I've edited (or if you're reading this already, I am currently editing) the first post to explain it a little better.
My record of ecto drops in 1 run remains at 3. All off one group of ~18 smites too. Sethellington
i've had 6 ectos in one run and 0 for the next 3, it's all chance, but it might have something to do with the chance other people are getting aswell.... remember that thread about loot on SV vs. 55? They both got the same items word for word when they entered at the same time.
Masao
@Sethellington: The reason for that is that all of the drops are set upon entry of the explorable area.
By entering at the exact moment in time that another person does, you'll get the same drop table and thus each mob will drop the same as the corresponding mob in the other person's map. Think of it as rolling all the dice at the beginning of the map, rather than when you kill the mob. MrSlayer
Quote:
Originally Posted by Masao
@Sethellington: The reason for that is that all of the drops are set upon entry of the explorable area.
By entering at the exact moment in time that another person does, you'll get the same drop table and thus each mob will drop the same as the corresponding mob in the other person's map. Think of it as rolling all the dice at the beginning of the map, rather than when you kill the mob. I'm going to add this to my first post, if that's ok. It kind of clarifies the findings of that thread. I'd forgotten about it. Sethellington
Yeah, that's what I was trying to get at... Not to worry, reading it back, I wasn't very clear ^^
Masao
@MrSlayer: Sure, if it helps.
@Sethellington: Yeah, I thought what you were trying to get at was that drops are weighed against others who enter, which isn't necessarily true, but also that they might be the same as the drops of those other people. One thing that might be an issue as to the thread as a whole. I'm not sure, but I'm was fairly certain that 'special' drops such as collectibles, golds, greens and materials have a separate loot table from that of 'normal' drops like weapons and armors. It doesn't really affect anything, all it does is make special drops and normal drops mutually exclusive and multiply the dice rolls per mob, possibly resulting in double drops of of a normal and special item, but also resulting in more empty boxes. Anyone have info on this from the .dat or are the loot tables stored serverside so as to avoid modification? Sethellington
Quote:
Originally Posted by MrSlayer
Masao raised an interesting point
I'm just wondering... how did he raise the point? I was referring to the thread you've linked... NOt really trying to gain "glory" for this or anything, just kinda confused...
MrSlayer
Quote:
Originally Posted by Sethellington
I'm just wondering... how did he raise the point? I was referring to the thread you've linked... NOt really trying to gain "glory" for this or anything, just kinda confused...
He suggested the dice rolls occur at the beginning of the zone, but are still a valid way of looking at the drop rates. I'll go add your name in too though, just because I'm nice
jonnieboi05
this is all very interesting. i've read about this somewhere before on how drops will be the same for 2 different people if they enetered the same zone at the exact same time. very facinating.
~LeNa~ It's A Me
Quote:
Originally Posted by MrSlayer
And ouch. Tough luck kiddo, you just got a blue Holy Rod with an "Improved Sale Value" mod.
I would be happy with that, itll help me make my 1k back on those "unlucky" runs lol. My Lipgloss is Cool
tl;dr Bad runs are bad luck, stop complaining.
^Did I get it right? =) MLiC Skinny Corpse
Nicely explained there.. hopefully it'll straighten
things out for a few people. Good job! Talarian
Nice, clear explanation. Although from the experience of trying to explain the same sort of thing to people about lockpicks, some people just don't seem to believe in chance...
capblye
so to compound your theory a bit more ...
Assuming what your saying is true, then when a player belongs to a group, every person, hero and hench in the group rolls a die for the same creature killed as to what drop they MIGHT receive IF ... A larger die is rolled with a side for each character in the group to see which player will recieve that drop. MrSlayer
Quote:
Originally Posted by capblye
so to compound your theory a bit more ...
Assuming what your saying is true, then when a player belongs to a group, every person, hero and hench in the group rolls a die for the same creature killed as to what drop they MIGHT receive IF ... A larger die is rolled with a side for each character in the group to see which player will recieve that drop. Personally, and this is not an answer based on anything other than an educated guess, I think it happens something like so. -> Monster Dies -> Big Die (as you put it ) is rolled to decide who gets loot -> Little Die is rolled by that person to decide what it will be As the findings from the other thread mentioned show, however, this would all be done at the very beginning of an area. Would be interesting to see if you get the same loot in the UW/FoW as you do when sync-entering a zone at the same time. xxx wraith xxx
verry clear guide, even though i had figured a part of this myself too I still think its kewler to think Anet nerfed stuff just because i need some1 to blame. But really good explanation and funny too. Great Guide i'd like to see some more guides bye you about loot and such Cheers
~Wraith~ Washi
Nice explanation although not really exact. Close enough though.
The thing is you can't really compare rolling a die to computer generated "random" draw. This is because computers are unable to generate random numbers. Rolling a die is perfectly random, but computer generating a random number uses an algorithm usually based on strigns. I can't really explain it in english cause it's not my native tongue, but the point is that computer generated random events are NOT independent. That is the biggest flaw. It's "random enough" for game purposes but still not exactly random. Masao
Quote:
Originally Posted by Washi
The thing is you can't really compare rolling a die to computer generated "random" draw. This is because computers are unable to generate random numbers. Rolling a die is perfectly random, but computer generating a random number uses an algorithm usually based on strigns. I can't really explain it in english cause it's not my native tongue, but the point is that computer generated random events are NOT independent. That is the biggest flaw. It's "random enough" for game purposes but still not exactly random. Of course not, that's why the hypothetical die roll is simply an analogy to an RNG that picks out a box from a loot table. The die roll is simply easier for people to understand, but I believe you'd be hard pressed to find someone thinks a computer has the ability to mimic the flaw of random chance. Spiritz
Good theories.
Its hard to state why people get bad runs but i look at it this way thru what ive noticed in other places. Ive been on cantha and into underground city and wiped the afflicted say 50 and yet only had a few drops and gold , i see it theres also something that affects it - you could kill 10 enemy before u get a single drop and a few times a dop on first kill. We say theres a "formula" for dropping but can there be one that isnt drop rates but "drop chance" also - eg "smite has 2% chance of an ecto drop" you kill 20 and no ecto but u got say 4 items and 300g - first the server uses drop chance .. then it applies drop rate ... so each smite has a chance to drop item then it works chance for rare drop. Is it most of the ppl that say about drop rates decreasing etc do just the smite runs , ive noticed u can do same run time and time again say 4 in hr but compared to a 4 man team who kill aatxes ( anything that drops ectos ) will always have more ectos drop ( not refering to loot scaling ). We kill more in given time than solo but possibly have same amount of items. We cant do random numbers ourselves mentally - our brain cant do it , dice throwing has many factors , speed/angle/how it hit and so on , a program has less factors to use , just a basic randomize 1 - 10000 and act upon the number. Div
Too bad people in Riverside will still QQ when they get bad runs.
Jay0
very nice thread, informative, funny, and easy to read, my fav things in any thread
A Simple Farmer
Nice thread, I can see this being linked to all the 'major loot nerfed' threads.
Next step, can you explain how to use the search feature? MrSlayer
Quote:
Originally Posted by holymasamune
|