Guild Wars 2

I’ve been gambling with rares (weapons) in the Mystic Forge for some time now, and after the last patch, it seems like the odds of getting an exotic have been reduced in half. Anyone else experiencing something similar?

Prior to the patch, the overall odds seemed 1:5/1:6. Sure, there would be “bad luck” here/there — sometimes 1:10, if not worse — but it’d eventually “even out” (2-3 exotics in a row, for example, and a few days ago I did get Dusk).

After the patch, however, I attempted 65~ forges, and it resulted in maybe 4-5 exotics. My wife later attempted 20~ forges (also weapons), and it resulted in 2 exotics — she used to have 1:5/1:6 odds as well.

Nothing has really changed with the items we’ve used (they’re crafted), and while a recent experiment (for precursor) on here suggested a 10% odds anyways (for exotics), my experiences, and the experiences I’ve read from other sources, suggests a 15/20% instead (eventually.. randomization considered).

At first I thought it was because I removed my MF gear (paranoia), so I bought an entire replacement set (and ate nomnom food) for my last attempt — it didn’t change the odds. I also thought it was due to alternating crafted-weapons (started this to avoid a possible diminishing return), but it didn’t change the odds either.

I’m not sure if it’s just a very large failure-rate (bad luck), or I was just “lucky” for weeks, or if some numbers were tweaked in the patch. All I know now is instead of eventually breaking-even or making profit (long-term), I consistently lost half of my investment (the exotics I got seriously equaled half of my costs!).

If anyone has any pre/post patch experience to share, then that’d be great. I doubt ANet would like to reveal much (less it be exploited), but perhaps a “Yes” or “No” to whether it has changed would be nice. ;P

It could just be a string of bad luck, but I did experience this after the patch. I went through about 160 rare daggers (40 combines) and got two exotics. I was in full exotic magic find gear with a magic find booster. Ordinarily, that would have resulted in 6-10 exotics.

Now like I said, there’s obviously RNG involved, but that’s a pretty big deviation from the norm.

I just asked Linsey about this very issue here. In the first post of that thread, Linsey said this:

We did fix a few bugged recipes that were rewarding Legendary precursors more often than they should have been. These recipes were mostly using Rares in the level 70-75 range and mostly affected the precursors for the greatswords and the dagger. This has had some effect on the drop rate when using lower level rares to attempt to get a precursor but it was never intended for that to be the best way of acquiring these weapons. The recipes for throwing in level 80 exotics have not changed at all.

Unfortunately, this statement doesn’t say anything about the situation with level 80 rares, and it seems that my question may have now been drowned in further discussion.

So far, we know that the drop rate for lower-level rares is much reduced, and the rate for exotics is unchanged. This narrows it down to 3 notable possibilities, I think:
1) The chance of an exotic never meant to be that high from putting level 80 rares into the forge, and has now been fixed.
2) The chance of an exotic from level 80 rares is not much different from before, and the concerns over this are mostly a case of mass hysteria – perhaps the chances beforehand weren’t as high for 80s as everyone thought, or maybe they’re still that high and some people have been unlucky.
3) The chance of an exotic from level 80 rares is currently bugged, and may end up being fixed.

Andykay.5198:

It could just be a string of bad luck, but I did experience this after the patch. I went through about 160 rare daggers (40 combines) and got two exotics. I was in full exotic magic find gear with a magic find booster. Ordinarily, that would have resulted in 6-10 exotics.

Now like I said, there’s obviously RNG involved, but that’s a pretty big deviation from the norm.

Ouch — yes, there’s RNG, but it seems as if it’s been made worse. I used to get 2/3 in a row on occasion — mainly 1:4 <-> 1:8, however — but after so many forges, it was “restricted” not to be lower than 1:10 odds!

My condolences to your gold.

lackofcheese.5617:

I just asked Linsey about this very issue

This isn’t in relation to the recent patch, however, is it? The rates I experienced pre-patch were fine.

Hmm, yeah, those changes were made a while back. Still, Linsey’s post isn’t really patch-specific, I’d say, though it’s not entirely clear with respect to some things, like level 80 rares.

A lot of people were saying the rate was already down to ~12% for level 80 rares ever since they fixed the problem with lower-level rares (particularly the Godskull rares). Maybe that ~12% figure was just wrong, but it seems unlikely that it would have gone up and then back down again.

One thing that would definitely help is if we had more data to work with.

Perhaps we’ll get more input from other gamblers. I’d gamble more myself if I didn’t think I’d consistently lose half my gold in the end. Fortunately, materials were somewhat cheap for my last attempts — I think it cost about 30~ gold to forge 60+ times, but the exotics only covered half of those costs. Prior to that, the exotics eventually multiplied investments by at least 3-4x (excluding pre-cursors).

In that same thread Linsey said that the chance of getting the legendary base depends on level and rarity of the items you put in. So highest chance of getting the base is 4 lvl80 exotics. I’m guessing the bug they fixed was regarding rares with different levels not taking into account the lowest levels, and only the highest (pure speculation). She didn’t mention anything about the chance of getting exotics from rares, though.

yesterday, after the patch, i put 4 random lvl 70-80 rares and I got a named exotic sword, the resonator.
So maybe I got very lucky.

Heh, I’d never even heard of that one before.

Yup, i put 1000’s of dyes into the forge and normally get a good amount of yellow from blue and a nice amount of green. This time i got a nice amount of green and zero yellows. So i brought a couple for thousand to test again and still none.

They have nerfed something.

It is really, really hard for people to understand just how randomness really works. People want to look for patterns – it’s how we all live every day. But just because something happens to you 100 times in a row doesn’t mean ANYTHING. If 100 people come to this forum and post that something happened to them 100 times, that STILL doesn’t mean anything. Think about the scale here: things are happening to millions of people millions of times per day. Your personal experience with random number generation doesn’t amount to more than a grain of sand on a beach.

The devs haven’t changed anything. Not only that, but then they specifically said they didn’t change anything. Sorry.

arabeth.2361

It is really, really hard for people to understand just how randomness really works. People want to look for patterns – it’s how we all live every day. But just because something happens to you 100 times in a row doesn’t mean ANYTHING. If 100 people come to this forum and post that something happened to them 100 times, that STILL doesn’t mean anything.

you are saying if 100 people did 100 attempts and all 100 didnt get a 1/10 ratio at all at all that doesnt mean anything? maybe if the odds where huge like 1/100000 however if it is infact 1/10 then you would be wrong and that would infact mean something.

I did my first few rolls of about 56 (reqlevel 80) rare staves… not a SINGLE exotic…

The only reason I even considered it was because I had heard the odds were about 1/6 to rank up to the next rarity. CLEARLY this is not the case… I wish there was some transparency about these “undocumented changes.” I could have saved myself a lot of gold and grief.

Everything means something. The problem is that people often overestimate the significance of what they do observe, and are subject to myriad biases.

mythos.1254:

I did my first few rolls of about 56 (reqlevel 80) rare staves… not a SINGLE exotic…

The only reason I even considered it was because I had heard the odds were about 1/6 to rank up to the next rarity. CLEARLY this is not the case… I wish there was some transparency about these “undocumented changes.” I could have saved myself a lot of gold and grief.

56 rares is only enough for 18 tries.
If it was a 1/6 chance, then there would be a (5/6)^18 ~= 3.76% chance of you not getting any exotics from that 18 conversions.

You might think that’s low, but at the same time you have to consider that with so many people playing this game, 4% chance RNG rolls / sequences of rolls are probably happening as many as millions of times a day. This kind of thing happening to a single person is hardly enough evidence to discount the probability being at or above 1/6.

If someone did 100 conversions in a row without getting a single rare, that would be enough to justify talking about it, although when you consider the number of people playing a game there’s still a decent chance it has happened to someone.

If someone did 200 conversions in a row without an exotic, I’d be willing to agree the chance is probably lower than 10%.

LinseyMurdock.4165:

We have made no changes to this stuff in the 10/7 build. The only changes that went in related to the Mystic Forge are in the patch notes.

Thanks for the response — I’ll have to try gambling the odds again next time I play. :P

I probably should’ve been less precise in my question, however, since any updates to the RNG algorithm probably wouldn’t require a client update (since it’s server-side).

Well something has deffinatly changed, i getting exoits 60% of the time now its under 10%

Regarding probabilities:

If I log on now and do 100 conversions, there is about a 1 in 40,000 chance of never getting an exotic, assuming the probability is 10% each time. So if 40,000 people do this, we’d expect at least one of them to get no exotics. And since it’s really kittening annoying to have that happen, we’d expect people who got 0 or 1 or 2 exotics to post here more often than the equally uncommon people that got 18 or 19 or 20. Because even though those numbers are about as unlikely as 0-2, it probably doesn’t feel as lucky as getting 0-2 feels unlucky.

But yeah, if you did 200 in a row with nothing, that’s suddenly the much more striking probability of 1 in 1.4 billion, and it would indeed suggest that something strange was going on.

That’s a good way of putting it, although your numbers are off when you say “equally uncommon”.

Getting 0 is about as uncommon as getting 25; getting 1 is similar to 21 or 22, and getting 2 is close to 19-20.

I know the normal approximation for a binomial distribution isn’t accurate at the ends of the relevant interval, but I didn’t feel like doing the extra math.

standard deviation.

http://en.wikipedia.org/wiki/Standard_deviation

basically you need huge samples. 1000 samples really isn’t enough, to make an accurate assumption. Even if its random, there has to be a probability.

personally, i got destroyed crafting my first bunch of clovers. everyone else, is getting 1/3-1/4, mine was 1/8-1/10 in my first 75 attempts. But I am taking everyone’s word for it that it evens out.

A thousand samples could definitely be enough to give strong reason to reject the claim that the underlying probability hasn’t changed. If my 1000 trials before the latest update gave me 100 exotics, and my 1000 trials afterwards gave me only 50, that’s a decent reason to suspect that something changed. Because if it’s actually the same before and after, at least one of those results is more than 2.63 standard deviations from average. Which is typically considered statistically significant.

Of course, if I didn’t do those trials myself, and instead only notice that someone did and posted their results here, I should be more skeptical, because I have no information about how many hundreds or thousands of other people ran similar tests and got more “normal” results.

Yeah, selection bias is something to be very wary of.

Food for thought…

If, in the long-run, one could profit from gambling with the Mystic Forge (weapons specifically), then what’s to prevent one from continuously trying? There would be no risk involved, and also no risk in trying for pre-cursors — one would simply increase their wealth while trying!

Unfortunately, this would (further) saturate the market with exotics (pre-cursors included), but it’s not very visible at this point (with demand still high). How would the market look, however, when more and more players “gamble” without loss, and less and less players demand their exotics? I haven’t thought much on the outcome, but it’d probably generate further economic issues (quick thought: deflation w/ no simple means — aside from new content — to implement disinflation).

So yeah… if they didn’t “tweak” this, then they probably will, and simply respond, if at all, with: “changed it to how it was supposed to be”. What it’s supposed to be is probably along the lines of: dump junk (you don’t want) in there, and maybe you’ll get lucky. It’s just unfortunate we might never know of these changes, or even what it was changed to/from.

Perhaps my reasoning went MIA somewhere up there, so pardon me if I’ve made any outrageous assumptions.

Doubt it. Yesterday (oct 9th) I was lucky enough to get the shield precursor when throwing in some random rares I’d gotten while working on my explorer achievement.

neoxide.7320:

Doubt it. Yesterday (oct 9th) I was lucky enough to get the shield precursor when throwing in some random rares I’d gotten while working on my explorer achievement.

Unfortunately, being “lucky enough” doesn’t quite help determine the current “odds”. Many have probably gotten an exotic (or exotic pre-cursor) from their first or second forges of the day, but that doesn’t mean the odds aren’t 1:1000.

Anyways, I’m not (or wasn’t) questioning the success-rate of obtaining pre-cursors. With their low rate, several thousand attempts alone would probably only yield a “general idea”. What I was questioning, however, were the success-rates of obtaining exotics in general, which, with an assumed 10-20% rate, is more reasonable to obtain. ;P