Refining for cheap

sleepcat

Used about 500 mirages to +4 12 pieces of equipment. I used about 60 or so Tienkang Stones from the event boutique. The next day I spend 160 mirages and about 5 Tienkangs to get a piece of gear to +1. I should've recorded the exact numbers.

Mirko - Raging Tide

290 mirages, 20 tiekangs

TT99 boots +4 --> +5
TT99 belt +3 --> +5
TT99 neck +4 --> +5

11/08/2010 4:30 - 5:00 PM east coast server time
arch elder

IrishPain - Heavens Tear

Miss_Tika - Lost City wrote: »

I refined +0 --> +7 with only 9 mirages and no dragon orbs b:cute

If this was the case, your odds of success are 3.64500 * (e^(-4)) = 0.0667605037.
We are talking 6.6% chance this would occur.
Also, since you require 7 straight mirages to get to +7 on armour, assuming you used only 9 would mean that you failed either on the first try 2 times in a row, or on the second try (+2) only once. Meaning only mirages also means that if you failed you dropped back down to +0. So assuming you failed on your first 30% chance for +2, you want people to believe you beat the 30% odds 6 times in a row? And managed to break the 6.6% mark with only 9 mirages? This is not rational. I'd like proof.
Something tells me you're not being honest.

Either way, refining is all luck in the end.

wraithwolf

Honestly i really could care less about the percentages ....what i want to know since i cant seem to find an answer is what does refining do to my gear other than makin it +1 +2 and so on. In another game that i play the higher + would change the glow of the weapon as well as add stats such as patk or matk on weapon and pdef and mdef on equips...does it do the same here?

WarrenWolfy - Sanctuary

IrishPain - Heavens Tear wrote: »

So assuming you failed on your first 30% chance for +2, you want people to believe you beat the 30% odds 6 times in a row? And managed to break the 6.6% mark with only 9 mirages? This is not rational. I'd like proof.
Something tells me you're not being honest.

Either way, refining is all luck in the end.

Getting +7 with 7 Mirages is a 0.036% chance, which is small, but it's certainly comparable to many of the prizes people are constantly winning from Packs.

So there's no real reason not to believe her. People who've just had an amazing run of good or bad luck are usually pretty eager to post about it, so I'd be surprised if I didn't see at least a few amazing claims.

Personally, the best I've done was to hit +5 on the first 5 refine attempts (Mirage/Mirage/Mirage/Tienkang/Tienkang) of the day, which is a 0.9% chance.

wraithwolf wrote: »

Honestly i really could care less about the percentages ....what i want to know since i cant seem to find an answer is what does refining do to my gear other than makin it +1 +2 and so on. In another game that i play the higher + would change the glow of the weapon as well as add stats such as patk or matk on weapon and pdef and mdef on equips...does it do the same here?

Refining adds extra matk/patk to weapons, extra HPs to armour, and mdef/pdef to rings/belts/necklaces.

If you want to know by how much per refine level you can look the specific item up in http://www.pwdatabase.com/pwi/.

Noob - Raging Tide

U think the tiens u get from quest have lower % or same? >_> something about free stuff makes me suspicious

Paramedic - Dreamweaver

when you refine +7 you will fail 820 times before u succeed .

doesnt matter if you gonna have 20 fails in a row or two times +6 and then 0 - you will have to refine 820 times before you get +7 (averagely ofc)

if she got +7 on first try, then she gonna take that fail refines 'package' with her to next refinement, so she may need 1600 refines to get next piece to +7

soon or later it gonna even out. there is no good or bad luck - just math


stop QQing and accept odds or use d.orbs

Lunality - Dreamweaver

Paramedic - Dreamweaver wrote: »

when you refine +7 you will fail 820 times before u succeed .

doesnt matter if you gonna have 20 fails in a row or two times +6 and then 0 - you will have to refine 820 times before you get +7 (averagely ofc)

if she got +7 on first try, then she gonna take that fail refines 'package' with her to next refinement, so she may need 1600 refines to get next piece to +7

soon or later it gonna even out. there is no good or bad luck - just math


stop QQing and accept odds or use d.orbs

Eh... that would be gambler's fallacy. I wish it did work that way...

Bleh, tried getting from +0 to +7 on a weapon today, and 2700+ mirages and 50 something Tishas and Tienkangs later... back to +0. x3 Thus, I am switching to Dragon Orbs. b:beatup

Paramedic - Dreamweaver

Lunality - Dreamweaver wrote: »

Eh... that would be gambler's fallacy. I wish it did work that way...

Bleh, tried getting from +0 to +7 on a weapon today, and 2700+ mirages and 50 something Tishas and Tienkangs later... back to +0. x3 Thus, I am switching to Dragon Orbs. b:beatup

it does work that way (if you simplify).

warren posted averange number of refines with mirages and stones.

by averange he mean - averange, not when you have good or bad day

WarrenWolfy - Sanctuary

Lunality - Dreamweaver wrote: »

Eh... that would be gambler's fallacy. I wish it did work that way...

Bleh, tried getting from +0 to +7 on a weapon today, and 2700+ mirages and 50 something Tishas and Tienkangs later... back to +0. x3 Thus, I am switching to Dragon Orbs. b:beatup

+7 is definitely well into frustration territory. With prices being what they are I personally would refine a weapon with Mirages to +3, Tienkangs to +5, Tisha to +6, then D.Orb 7 to +7. After that it would be D.Orb Oceans.

The average savings to get to +6 are going to be about 45%, which is just too good for me not to take the risk, while the extra cost of using the D.Orb 7 is going to be about ~8 million, which for me is worth paying to avoid the frustration.

Paramedic - Dreamweaver wrote: »

it does work that way (if you simplify).

warren posted averange number of refines with mirages and stones.

by averange he mean - averange, not when you have good or bad day

The part that's the gambler's fallacy is the idea that you take your fail/success "package" forward to your next refine, and that your odds will gradually work themselves towards average. That part's not true.

Nothing gets carried forward, and so if you're ahead/behind, then on average you'll still be that much ahead/behind 1000 attempts later.

So if I +5 something on my very first 5 tries my chances of success on my next attempt are exactly the same as if I had just failed 1000 times.

Okeano - Harshlands

Got pants +6 in first try, helm went from 5 to 0 after 500+ mirages and 30+ Tishas. I don't think I can take the stress of +7.

Asterelle - Sanctuary_1381265973

I had a very good run the other week using tisha for +8 refines. I had 5 pieces sitting at +7 and was able to get them all to +8 without completely losing any refines. I think 3 hit it with 1 tisha while another dropped to +6 before hitting +8 and another went to +5 before +8. Thats like 11 mirage total. Before that run of luck I was only successful in 1 out of 5 attempts for +8 (I define a failed attempt as completely losing a +7 refine down to 0).


My grand total for +7 -> +8 refines using tisha is 6 out of 10 times. Even with that I'm not sure I'd recommend tishas unless you have a lot of pieces to refine. My first attempt at +8 I remember using over 2000 mirages trying to get back my +7, using tisha definitely involves a lot of frustration and dread. I'm glad to be beyond the help of tishas now heh... well until I redo everything for R9 gear.

Krosan - Dreamweaver

My refine session from yesterday:

2 x +5 armor piece
1 x +4 armor piece
1 x +5 weapon

121 mirages, 9 tienkang, 8 tisha

Okeano - Harshlands

Asterelle - Sanctuary wrote: »

well until I redo everything for R9 gear.

b:shocked

Knownase - Heavens Tear

Today, I was aiming for +5 on my new Forest's Wisdom Helm, and wow....guess this is my "OMG WTF they nerf the rates QQ" story.


I think in all, I blew about 500 mirages, to get a +5, and like 30 Tienkang. I was like..."Alright, **** +5, +4 is good enough for me." b:cry


On the bright side, it has 4 sockets!

Guess having it at 4 sockets killed all my luck at refining. b:surrender

D Orbs are at like 190K on HT soooo...yea....still cheaper I think! lol

Atone - Lost City

With the latest d.orb sale and pre-sale gold prices, the Zoho spreadsheet says to use a 10* orb for +8 to +10, then tisha +11? b:shocked I admire all who have the guts to try that.

WarrenWolfy - Sanctuary

Atone - Lost City wrote: »

With the latest d.orb sale and pre-sale gold prices, the Zoho spreadsheet says to use a 10* orb for +8 to +10, then tisha +11? b:shocked I admire all who have the guts to try that.

Yeah, I saw that too and did a bit of a double-take.

So I checked the numbers, and they don't add up.

Consider if you fail, then you'd need another D.Orb Ocean, which costs ~29 million. That's 16.1% of the cost of a D.Orb 11, but the Tisha only gives you a 15.5% chance of success. Clearly there's an error in the Tisha calculation.

The error appears to be that the Tisha calculation was using the wrong base cost. It was adding the cost of a +11 attempt to the cost of a +9 refine, but of course that's not right. It should add the cost of the +11 attempt to the cost of a +10 refine.

I double-checked, and the error only affected Tisha stones. With the correction made Tishas drop significantly in their overall effectiveness.

- EDIT -

The equation still doesn't seem correct for Tisha stones. I'm getting results with the simulator that don't match the projections. The other stones all seem to check out, but not Tishas. I'll have to do some more checking and figure out what's going on...

- EDIT X2 -

Huh... well, I re-did the formula from scratch using a different, more complex, approach. I then ran 100,000 simulations of a +12 refine making heavy use of Tishas and the results came out to 99.99% the predicted cost.

I then went back and checked the original formula and it turns out it gives identical results. So I was correct in the first place, and I assume this means my new, more complex formula simplifies down to the old one.

I then changed everything back to how it was before and tried to figure out why a Tisha that costs 16.1% of a D.Orb 10 and only has a 15.5% chance of success is actually cheaper. It turns out that's completely correct. What I was forgetting is that you don't have to pay the D.Orb 10 cost when you succeed.

For example, if you have a 25% chance of success, you'll need an average of 4 tries to succeed, but that means 1 success and 3 failures, and thus you only pay the failure cost (ie. a D.Orb 10) 3 times, not 4 times.

So I've gone ahead and changed everything back to the original, although I've turned the "Risk Aversion" up to 10% instead of 5%. Personally, that's much closer to what I actually do in-game.

Ubnext - Sanctuary

when I look at the Orb description it states 100% chance to refine +X or +XX depending on the orb.
ok so if 100% means 100% then what the heck is the 0.36% **** about and why is everyone using 500 Mirage celestones?
Why not just use the Dragon Orb associated with the +1-+12 refinement your want to acquire?

and 44 pages here, is there an explanation of what each +x, +xx eqauls in refinement.
I know armor is +?
and weapons is +?

and what's the best someone can hope for?
Like if I want to hang out at lvl 70 for a year, dumping exp into Genies and farming spirit to up my spells to max for lvl 70. what is the best refinement I can hope to achieve?

and the next question I saw the word simulator.
where is it?

Olbaze - Sanctuary

Ubnext - Sanctuary wrote: »

when I look at the Orb description it states 100% chance to refine +X or +XX depending on the orb.
ok so if 100% means 100% then what the heck is the 0.36% **** about and why is everyone using 500 Mirage celestones?
Why not just use the Dragon Orb associated with the +1-+12 refinement your want to acquire?

Because refining with the Mirages is a lot cheaper.

and 44 pages here, is there an explanation of what each +x, +xx eqauls in refinement.
I know armor is +?
and weapons is +?

The values are different from every item. Go here and search for the item you want, it'll bring up the stats.

and what's the best someone can hope for?
Like if I want to hang out at lvl 70 for a year, dumping exp into Genies and farming spirit to up my spells to max for lvl 70. what is the best refinement I can hope to achieve?

Well, there's two bad things in that:
1. You shouldn't hang onto a level. As you level up, you'll have less and less skills to level up but you'll get more and more spirit. There's also no cap for how much spirit you can have.
2. Refining at level 70 should be left to +1, +2 or +3 at most. Anything higher isn't necessary nor is it wise.

WarrenWolfy - Sanctuary

Ubnext - Sanctuary wrote: »

and the next question I saw the word simulator.
where is it?

It's part of the Excel version only. Unfortunately Zoho sheet is far too slow to be useful at running refining simulations.

Paramedic - Dreamweaver

WarrenWolfy - Sanctuary wrote: »

The part that's the gambler's fallacy is the idea that you take your fail/success "package" forward to your next refine, and that your odds will gradually work themselves towards average. That part's not true.

Nothing gets carried forward, and so if you're ahead/behind, then on average you'll still be that much ahead/behind 1000 attempts later.

So if I +5 something on my very first 5 tries my chances of success on my next attempt are exactly the same as if I had just failed 1000 times.

ow i see now

nonetheless, should i expect to use 8000 mirages (on averange) on 10 gear parts, if i want to refine all to +7?

or is it equally possible to use just 70 on whole process?

WarrenWolfy - Sanctuary

Paramedic - Dreamweaver wrote: »

nonetheless, should i expect to use 8000 mirages (on averange) on 10 gear parts, if i want to refine all to +7?

or is it equally possible to use just 70 on whole process?

The odds would be close to a bell curve, where the peak of the bell would be 8850 Mirages.

The more refines you do (in this case 10), the steeper the edges of the curve, so the more likely you are to end up closer to the peak. If my math skills were better I'd have the spreadsheet draw the curve.

Paramedic - Dreamweaver

it was rhetorical, so no worries

but i admit that i didnt understand how gambler's fallacy works exactly, so ty for pointing flaws out


so the point is: to not look for any 'evil powers' behind mirage refining. your luck dont depend from your connotations with devil or gms. if on averange, ref +7 takes 800-something tries then you can expect it. more refining you do - more accurate result will be. this gotta even out later or universe will implode

is that right?

Ubnext - Sanctuary

Olbaze - Sanctuary wrote: »

Because refining with the Mirages is a lot cheaper.



The values are different from every item. Go here and search for the item you want, it'll bring up the stats.



Well, there's two bad things in that:
1. You shouldn't hang onto a level. As you level up, you'll have less and less skills to level up but you'll get more and more spirit. There's also no cap for how much spirit you can have.
2. Refining at level 70 should be left to +1, +2 or +3 at most. Anything higher isn't necessary nor is it wise.

Thank you for answering most of my question. and especially for not treating me like some one you just wasted your time on in solving my lack of comprehension.

I have a wheel of fate, I am 65. I wanted to refine at least +3.
but +7 would be nice.
each time i use mirage only I go up +1 and go down -1 sometimes I fail 2 or 3 in a row,

but if I am correct...you guys are saying if I had 300 Mirages and rolled the +'s eventually I might hit 3 +'s in a row at 33% , but I would actually have to spend all 300 Mirages. no luck just numbers?

flamingahole

Ubnext - Sanctuary wrote: »

Thank you for answering most of my question. and especially for not treating me like some one you just wasted your time on in solving my lack of comprehension.

I have a wheel of fate, I am 65. I wanted to refine at least +3.
but +7 would be nice.
each time i use mirage only I go up +1 and go down -1 sometimes I fail 2 or 3 in a row,

but if I am correct...you guys are saying if I had 300 Mirages and rolled the +'s eventually I might hit 3 +'s in a row at 33% , but I would actually have to spend all 300 Mirages. no luck just numbers?

Sell the wheel of fate and get an ancient arbor instead. The -channel and vit are nice bonuses and the mag.attk isn't too far below the arbor but you could probably sell it for a decent profit and buy a prerefined/sharded arbor. It would depend on the prices on Sanctuary though so check those out beforehand.

And refining is all about luck. I got my bracers to +5 in 7 or 8 tries but when I tried to refine my chest it didn't work out too well; it's at +2 after going through around 150 mirages. Having more mirages just gives you more chances to get lucky.

WarrenWolfy - Sanctuary

Ubnext - Sanctuary wrote: »

but if I am correct...you guys are saying if I had 300 Mirages and rolled the +'s eventually I might hit 3 +'s in a row at 33% , but I would actually have to spend all 300 Mirages. no luck just numbers?

There's no way to be sure telling how many you'd actually need, other than to use Dragon Orbs.

The average number of Mirages to refine a weapon to +3 is 74, but you could get lucky and use only 6. I've gotten a +3 on my first try on at least one occasion, so it does happen. The chance of getting 3 successes in a row is 4.5%, so when you're refining without orbs you just keep trying again and again, ignoring the fact that you have to keep starting over, until you happen to hit that 4.5%.

Basically, refining without Dragon Orbs is a gamble, but it's a gamble where the odds are heavily in your favour. There's a slim chance you might spend 300 Mirages and get nothing, but the most likely outcome is that you save about half the price.

It's odd that so many people will gamble on opening Packs, even though the odds are against them, but won't gamble on refining even though the odds are 2-1 in their favour. I'm pretty sure it's because human beings have a strong emotional preference for, "I might win" instead of "I probably won't lose", even though the two are similar, financially speaking.

LenieClarke - Heavens Tear

Ubnext - Sanctuary wrote: »

I have a wheel of fate, I am 65. I wanted to refine at least +3.
but +7 would be nice.
each time i use mirage only I go up +1 and go down -1 sometimes I fail 2 or 3 in a row,

but if I am correct...you guys are saying if I had 300 Mirages and rolled the +'s eventually I might hit 3 +'s in a row at 33% , but I would actually have to spend all 300 Mirages. no luck just numbers?

when it comes to refining, WarrenWolfy and flamingahole had it right; it's all probability and chance.

when it comes to your weapon in particular, at level 6x, the ancient arbor is way better than the wheel of fate. i've +1'd mine and imbued it with the best gem it'd take; my wizard alt got the wheel of fate. i don't think i'll refine this weapon any further, since i'll just end up decomposing it for my TT70 weapon anyway, but you'd have to refine a wheel of fate unreasonably high before it'd beat an ancient arbor.

edit: if pwdatabase is telling me the truth, you'd have to at least +8 a wheel of fate before it'd match a +0 ancient arbor --- maybe +9 it. this is without considering that the grade 8 TT60 weapon will both refine better, and take better gems, than the grade 7 wheel of fate.

Teny - Heavens Tear

WarrenWolfy - Sanctuary wrote: »

There's no way to be sure telling how many you'd actually need, other than to use Dragon Orbs.

The average number of Mirages to refine a weapon to +3 is 74, but you could get lucky and use only 6. I've gotten a +3 on my first try on at least one occasion, so it does happen. The chance of getting 3 successes in a row is 4.5%, so when you're refining without orbs you just keep trying again and again, ignoring the fact that you have to keep starting over, until you happen to hit that 4.5%.

Basically, refining without Dragon Orbs is a gamble, but it's a gamble where the odds are heavily in your favour. There's a slim chance you might spend 300 Mirages and get nothing, but the most likely outcome is that you save about half the price.

It's odd that so many people will gamble on opening Packs, even though the odds are against them, but won't gamble on refining even though the odds are 2-1 in their favour. I'm pretty sure it's because human beings have a strong emotional preference for, "I might win" instead of "I probably won't lose", even though the two are similar, financially speaking.

Check ur pm.

The mentality between opening packs and refining is that on packs you're not really losing... I mean, you'll end up with several thousand tokens, but that's better than getting to +7 and resetting.

Or you could be like me, and get to +8 and reset over and over until you get +9... i did that probably 3 times b:surrender

WarrenWolfy - Sanctuary

Teny - Heavens Tear wrote: »

The mentality between opening packs and refining is that on packs you're not really losing... I mean, you'll end up with several thousand tokens, but that's better than getting to +7 and resetting.

Yeah, that's definitely why it feels tougher to risk a +4 than it does to risk opening a Pack. It's also why it feels easier to use Tishas instead of Tienkangs, even though most of the time Tienkangs are the better choice.

There's a fascinating study done with monkeys (Google "Monkey Economics") that shows we have an innate emotional bias to try to avoid risk of taking a loss, yet we'll happily accept the exact same risk if we think it's about potentially getting a win instead, even if both wagers have identical odds and payouts.

Olbaze - Sanctuary

I noticed something that threw me off for a bit.

I have the Zoho Sheet version bookmarked. When I put in 700,000 for Gold price and 33,000 for Mirages, reflecting current Sanctuary Gold price and price of 2 mirages, the chart said that the best option was to use Mirages for +1 and +2 and Dragon Orbs for the rest.

Similarly, when I put in just 33,000 for Mirages, it told me the usual: Mirages until +3, Tienkang until +5, Tisha until +7 and rest with Dragon Orbs.

Of course, at first I thought that it was just a bug in the code. However, after thinking of it for a moment, the actual information would be that the price of Dragon Orbs is actually so low that there is a point in Gold Price after which Tienkangs and Tishas just aren't worth it anymore for the odds they offer.

The same can be observed with extremely low Gold prices.

So basically, the cheap Dragon Orbs are just about making Tienkangs and Tishas obsolete for weapon refining.

The same is of course not true yet for armors, as the price per attempt is much lower.