I've forged 23 Precursors. Here is my data.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Samuirai.4561

Samuirai.4561

I have already shared this on reddit, but I guess some people are interested in my research here as well.

http://smrrd.de/static/gw2/precursor.html

kind regards,
samu

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Lucky.9421

Lucky.9421

You are a hero among heroes…

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Obtena.7952

Obtena.7952

Thank YOU so much. Finally a statistical approach to figuring this out. Unfortunately, if people get a clue, information like this is the kind of thing that could potentially affect an increase in price.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Getefix.9150

Getefix.9150

increase in price? it doesnt change the probibility of gaining a precursor, theyre already stupidly expensive

“Nothing is true, Everything is permitted”

Kiel Replacement Movement

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Astralporing.1957

Astralporing.1957

Slight correction to your info:
(on different percentages per weapon type)

Of course this variance has nothing to do with the different weapon type. This just shows how random numbers work

There is nothing “of course” about this.
There are several ways RNG may be set up to select precursors. There might be a flat % chance – in which case your statement would be true. There might be however just a chance of getting a random named exotic. In this case, the chance of getting precursors would depend on how many named weapons of that weapon type exist.
Hint: the number is different for each type.

Actions, not words.
Remember, remember, 15th of November

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Samuirai.4561

Samuirai.4561

I doubt that. It wouldn’t make sense to make certain precursor probabilities higher or lower than others from a game economic design point of view.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: cranked.3812

cranked.3812

Wow. That is impressive work there.

/Tips hat to you……

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Svetli.4276

Svetli.4276

RNG was on your side BUT yes this is pretty awesome

i like to ask if token exotics have the same chance as drop/crafted ones ?
i currently poses over 25 000 dungan tokens and one day will attempt at mf with them

“What you wish for may not be what she wishes for.” – Skull Knight

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Malediktus.9250

Malediktus.9250

You could have saved money by not selling to the highest offer.

1st person worldwide to reach 35,000 achievement points.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Redenaz.8631

Redenaz.8631

Bravo. I knew I didn’t have nearly enough gold for the Mystic Forge to ever be worth my while, but it’s nice to see that, expensive as precursors are, gambling is simply a worse idea than saving up and buying one outright.

Thanks!

~The Storyteller – Elementalist – Jade Quarry~

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: MrKahwk.9524

MrKahwk.9524

I love you. I hope people will realise now that one should BUY the precrusor from the trading post.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Malediktus.9250

Malediktus.9250

If noone gambles precursor prices would be even higher because supply decreases.

1st person worldwide to reach 35,000 achievement points.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Cynz.9437

Cynz.9437

the problem is he would need to make at least 10 000 precursors for data be somehow reliable

All is Vain~
[Teef] guild :>

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: LadyRhonwyn.2501

LadyRhonwyn.2501

Nice research (and I really liked that video you made! Wonder how much you cursed at yourself at that point :P)

I’ll just randomly throw in stuff during MF dailies and hope for some nice items

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: fadeaway.2807

fadeaway.2807

Incredible research, thank you.
One question from a mf newcomer though. Why exotics and not rares?

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Brother Grimm.5176

Brother Grimm.5176

the problem is he would need to make at least 10 000 precursors for data be somehow reliable

Not really as statistical accuracy does NOT depend on obtaining an outcome X times, but making Y attempts. He likely needs to make 10,000 FORGES to make the data statistically sound, but what he has here is no doubt within 2.5% of the number you would get from another 7,000 flushes. This is certainly better than anything else currently available on the subject.

……..
There are several ways RNG may be set up to select precursors. There might be a flat % chance – in which case your statement would be true. There might be however just a chance of getting a random named exotic. In this case, the chance of getting precursors would depend on how many named weapons of that weapon type exist.
Hint: the number is different for each type.

If one assumes it is simply based on named Exotics, then trying for a precursor from the smaller number pools would be better odds (to get a precursor). Somewhere on this forum, I did a compilation of the number of named Exotics per weapon type….. I recall that Focuses had over twice as many name Exotics as the norm (nearly 30) and I think maces had the least (6 or 7).

We go out in the world and take our chances
Fate is just the weight of circumstances
That’s the way that lady luck dances

(edited by Brother Grimm.5176)

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Wanze.8410

Wanze.8410

Also a +1 from me because of the nice presentation of the data.

Even though I personally, or the Tin Foil [HATS] never speculated or invested into pre cursors, your research seems to be what i expected.
I know that many people do stuff like this and have way better info on drop rates from the forge than the average person.

If you mass forge pre cursors, you will break even or make a little profit after taxes, if you sell to the highest bidder, If you want to make some good profit, you need to undercut the lowest seller.

I would estimate that you wont make more than 50g profit on average for each pre cursor you get from the forge.

And I think i know faster ways to make 50g profit on the TP.

Tin Foil [HATS]-Hardcore BLTC-PvP Guild
Bloin – Running around, tagging Keeps, getting whack on Scoobie Snacks.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Brother Grimm.5176

Brother Grimm.5176

……Why exotics and not rares?

If we assume the 20% next tier up is valid here, why cut your chances by a 5th? I don’t think Exotics are 5 times more expensive (but I could be wrong…at least for certain weapons).

Not saying the data might not be interesting as it would answer the burning question if using only RAREs your chances are 20% lower…..(what one would assume) or totally different in some other respect.

We go out in the world and take our chances
Fate is just the weight of circumstances
That’s the way that lady luck dances

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: ZudetGambeous.9573

ZudetGambeous.9573

Great presentation and nice data. I’d be interested to see if there is any corralation between the number of possible named exotics and the frequency of precursors. Theoretically you have an equal chance of getting any named exotic, which means some weapons like foci would be significantly harder to get precursors than something like maces.

Also it would be interesting to see if the chance to get a GS precursor is really 2x that of other weapons.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Astralporing.1957

Astralporing.1957

Also it would be interesting to see if the chance to get a GS precursor is really 2x that of other weapons.

Can be. there are two GS precursors, after all. That saying, number of tries in GS category was significantly lower than with other ones, which may also effect the results.

Actions, not words.
Remember, remember, 15th of November

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Tachii.3506

Tachii.3506

This is a great read.

And I believe 2000 attempts is good enough in terms of statistical relevance, for anyone who took stats. Should be stickied!

SBI – Thief and the occasional Guardian & Warrior.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Zanshin.5379

Zanshin.5379

the problem is he would need to make at least 10 000 precursors for data be somehow reliable

News channels do that all the time: ask a hundred people in the street a question and then say “x% of the country thinks that…”.

I think 2901 tries is good enough. In the end, it’s still just an estimate.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Oghier.7419

Oghier.7419

Thank you for doing this research. You’ve saved me from the temptation to use the forge to fish for precursors

Snit Dirtnap (Thief)
Ratbag Dogsticker (Guardian)
…Yak’s Bend

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Samuirai.4561

Samuirai.4561

Of course more data would be more accurate. But if it is 1%, 0.5% or my 0.8%, it shows, that the probability is not 0.0001% and not 10%. And that is worth something, I guess

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Brother Grimm.5176

Brother Grimm.5176

Of course more data would be more accurate. But if it is 1%, 0.5% or my 0.8%, it shows, that the probability is not 0.0001% and not 10%. And that is worth something, I guess

Very much so. Again, thank you for the data.

We go out in the world and take our chances
Fate is just the weight of circumstances
That’s the way that lady luck dances

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Wolf Fivousix.4319

Wolf Fivousix.4319

@ Samuirai, the data you gathered there is very impressive, indeed! I was planning on doing something like it once I managed gold enough BUT my sells on TP have been slow, so I’m not getting that money.

You presented the data terrifically, congratulations for that too! May I ask your current profession, if you have one?

I would like to ask you how you managed to amass that huge amount of gold, if you be so kind to give some tips.

Also I would like to ask your in game nick so I could send you a nice present for your effort and contribution to the community. If you do not want to post it here, feel free to message me in game, my nick is in my signature (which happens to be the same of my forum name).

Once again, I salute you for your achievement and hope you best luck on your next projects. Keep that way and you’re going far, my friend!

// Dragonbrand
Wolf Fivousix – Elementalist
Black Wolf Trading Tool

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: SageGaspar.5012

SageGaspar.5012

Of course more data would be more accurate. But if it is 1%, 0.5% or my 0.8%, it shows, that the probability is not 0.0001% and not 10%. And that is worth something, I guess

I was just teaching someone binomial probability today! Neat that I can actually use it for something.

Assuming Anet just has a set probability for each combine regardless of what you’re combining, there’s 95% confidence that it’s within .32% of .79%, so your sample is pretty accurate!

Doing 10000 samples would decrease the margin of error to around .173%, and if we combine your data with that other person who did 8000ish attempts, we get pretty much exactly that.

So unless Anet is screwing around with the probabilities based on (specific weapons/how many drops you’ve gotten thus far/phases of the moon/first letter of your character’s name/what direction you’re facing/if you’re doing the YMCA at your computer) then we have the drop rate pretty well nailed down.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Ensign.2189

Ensign.2189

Do you happen to have data on all the outputs you received from the forge? I’m using the ratio of named to unnamed exotics to supplement the precursor data and tighten confidence intervals.

Based on the OP’s data the 95% credible interval on the chance to get a precursor it 0.53% – 1.19%.

I can throw my own data in there at face value and tighten it to 0.47% – 0.84% (the OP got some really lucky rolls, I’ve gotten some really terrible rolls).

Most likely value seems to be somewhere in the 1:150 range.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Samuirai.4561

Samuirai.4561

@Wolf Fivousix.4319 message sent

@SageGaspar.5012 wahaaaa so awesome \o/ that is what I really love to see
See kids, math is awesome. So go back to learning!

@Ensign.2189 sorry, I have not kept track on named exotics.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: CryxTryx.9208

CryxTryx.9208

Wow this is great info. I’m not a math guy at all and this was explained in such away that I understood what I needed to know.

Thanks again.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Spiky.8403

Spiky.8403

I just spent another 600g in mystic toilet. Well, I think am finally done with this game. Can’t stand this stupid RNG in gw2, just can’t.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: OreoWolf.9564

OreoWolf.9564

I just spent another 600g in mystic toilet. Well, I think am finally done with this game. Can’t stand this stupid RNG in gw2, just can’t.

Wait, so instead of buying the precursor (almost all precursors are around the 600g mark) you went and gambled that gold away? Seems legitimate to me.

As for the OP, a very well organized and explained collection of data, no matter how accurate it is. Many thanks from me and many others on this forum!

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Spiky.8403

Spiky.8403

I just spent another 600g in mystic toilet. Well, I think am finally done with this game. Can’t stand this stupid RNG in gw2, just can’t.

Wait, so instead of buying the precursor (almost all precursors are around the 600g mark) you went and gambled that gold away? Seems legitimate to me.

I have already wasted too much gold to just buy precursor from TP. When I see guildies getting 3 precursors out of few tries (first two came after ~30 exotics) and another one from 4 rare weapons, I am not OK with buying out of TP. This is probably the only game that is ruled by pure random factor, which is not good. Doesn’t matter what data was made here, you can do same number of attemps and end up with empty hands.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Silver.3284

Silver.3284

Well done, for the patience and more for the presentation.
I also wonder what you do irl

Your sample is large – @ 95% confidence lvl the highest error margin is 1.88% – that is for 50% distribution, but given we are measuring a 1-2% probability it is more like 0.5%. So yes, it is not 0.0000001% and it is not 10%.
Even looking at only the sword sample the error is low – 1.35%.

2 bias I can see
First different kinds of precursors – while they should not have a different chance they may actually do so.
I assumed you stopped once you got a precursor. For obvious reasons, but the more straightforward way would be to do a set number of forges on each precursor – so we can take mean of successful tries rather than the mean of forges required. .

Also, I would had liked to see all the results – the idea of a chance to get a named exotic is tempting.

But once again, this data is just lovely. Thank you!

@ Spiky – unfortunately that is the nature of chance. It appears random in small samples and only becomes deterministic in large numbers. But this is also why this data is useful : it shows how rare it ACTUALLY is, rather than “I heard someone got one in 2 tries so I should do the same”

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Spiky.8403

Spiky.8403

I was just trying to say that random is killing progression for some people (like me), which is not fair. You may say “you spent 900g on mystic forge, why didn’t you rather buy precursor from TP?”. Well, because I didn’t spent all my money at once and also haven’t had them all together.

So yes, when I see someone with legendary I dont say myself “this one must be playing a lot, he deserves it”, but rather “this guy is kittening lucker”

Now someone who got precursor from MF should come here and say me am wrong, stupid etc.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Cyninja.2954

Cyninja.2954

I was just trying to say that random is killing progression for some people (like me), which is not fair. You may say “you spent 900g on mystic forge, why didn’t you rather buy precursor from TP?”. Well, because I didn’t spent all my money at once and also haven’t had them all together.

So yes, when I see someone with legendary I dont say myself “this one must be playing a lot, he deserves it”, but rather “this guy is kittening lucker”

Now someone who got precursor from MF should come here and say me am wrong, stupid etc.

You do know that a lot of people just buy the precursor from the TP. Nothing wrong with going that route. No luck involved in that.

Don’t get me wrong, I feel for you. Spent 300g myself. Just a small fyi:

I have already wasted too much gold to just buy precursor from TP.

Makes literally no sense. You have 0 progress from failed tries on the mf. Your chance of getting a pre after spending 600g is just as high as getting one after spending 10g as far as in game rng is concerned. I know the feeling though. Buying a precursor after gambling away 600g makes you feel as though you spent 1,200g (assuming the pre costs 600g). Just know this, going on gambling might net you a precursor after a few more tries (yay barely breaking even) or might give you nothing (which leaves you stuck with having spent another 600g with nothing).

The argument “I’ve already spent so much on the mf.” Is not the healthiest to stay interested in the game.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Keiel.7489

Keiel.7489

Don’t use the MF if you are bad at gambling. I don’t mean “bad” as in you have no control over the roll of the dice, “bad” as in you can’t handle taking losses. Mystic Forge is for those who enjoys to play the risk, sure some make huge profits, some get a margin of it, and most probably just lose everything they have. It’s all for the risk of getting that “win”. For me I like gambling, I’ve had pretty good luck in the past and I would say I am on the high end of the streak. Last night I decided to dump some gold into the toilet and see what I could get. First I spent 300g on GS, got dawn. Then another 300g on Staves, Got the Legend. Got everything on the TP and for both I think I got the precursor somewhere at the mid way point, could have sold the rest of the exotics and made some gold back but I kept trying, just for fun. I then also bought Warhorns 300g. Same deal half way through I got Howl, kept dumping the rest into the Mystic Toilet.

Probably will just sell The Legend and Dawn, since I have no need for either, it’s about a 400g profit, which I’ll probably lose to the MF later on.

[DONE]

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: KOPPER.1458

KOPPER.1458

I gamble a lot in the forge, what the best way to keep track of this info so I can add to this.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Samuirai.4561

Samuirai.4561

This is how I kept track:

https://docs.google.com/spreadsheet/ccc?key=0AqjFp3hDPMLUdHp0SnllYmhid2J3ZDBtN0RUNU5paEE#gid=0

You could create a copy of it, clear it, and sent me the link when you filled it with data, then I can add it to my data that would be really cool

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: insaneshadow.1654

insaneshadow.1654

So this is just data on exotics? I’m using rares because I’m trying to not blow a lot of money at once. I plan on doing it over an extended period of time.

Gust Root | Letum Folium | Lo Bridge | Snow Spot | Roland East | Nascharr | Bjorn Microbrew
Yak’s Bend
Lincoln Force [BOMB]

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: KOPPER.1458

KOPPER.1458

Ok I always do random Exotics, but I’ll play around with it and see what kind I can send you.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Spiky.8403

Spiky.8403

So I came back to game after MF precursor fiasco in december, sold my leftovers from bank. When I realised what patch brought us (another 2 hours waiting for some greens, yay!) I said myself I really have nothing to lose and bought 64 exotic axes. 15th forge was my first successful try in MF (after more like 1000g spent over the last ~18 months). It is still one big failture for me, but at least I got something to play with.

Attachments:

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Cassocaster.4576

Cassocaster.4576

Thank you very much for this, it’s an excellent examination of the RNG in regards to the forge.

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: FuriousPop.2789

FuriousPop.2789

how about less bragging and more giving out free precursors eh

I've forged 23 Precursors. Here is my data.

in Crafting

Posted by: Inimicus.7162

Inimicus.7162

Thank you for doing this research. You’ve saved me from the temptation to use the forge to fish for precursors

Here, here…but why fish in the mystic toilet? Need a picture of what you will find the most of? :P

This is the reason the thing earned its nickname, and the fact you can “/wiki mystic toilet” in game and get taken to its page…

While I like the idea that I don’t have to buy some recipe for crafting every time for some neat skin… I don’t gamble the thing because the odds are stacked so epic in the house’s favor, and I’m not feeling lucky :P