TOPIC: Google Doc for 2017-2018 Season Treasure Pulls

kurtreznor wrote:

Matthew Hayward wrote: I added a tab at the end with some guesstimates as to the average value of tokens at various rarities, and the overall average value of a pull this year, which I put around $3.14 from the sheet.

It's not synced with the main sheet, so it won't stay up to date - but I saw someone asking about it so I put together something quick.

docs.google.com/spreadsheets/d/1g1lb5GgfnL0xW5IzZV9VUvsJCzvPU9endLbWoIfmZvI/edit#gid=1996682456

Not meaning to derail the thread, but is $3 for ring of the hallowed accurate? I can see $8 for the new greaves, but the ring is far more rare; was that supposed to be $30?

Also, i dont have exact counts on my uncommon vs rare pulls, but i know my total and the specials i drew; should i add what i know and then split the remainder 50/50 for rares/UC?

Kent if you believe they were about 50/50 that probably would be best.

jedibcg wrote:

Matthew Hayward wrote:

Steve wrote:

Mike Steele wrote:

jedibcg wrote: I have said it before and I will say it again the 50 bags are from the same mix. At least 900 maybe 950 of my pulls were from the bags both blue and clear. I spoke to Chandler who made the 50 clear bags. They were 50 tokens from the mix. He didn't look what went in them but did notice lots of colors. I pulled the first 2 clear bags (arpund 130pm) and last 3 on Saturday (around 6pm not certain if they made more). If it seems the box was hotter it is a perception. If someone pulls 3 to 19 from the box and gets something good you will hear about it. You don't hear from everyone that that doesn't an UR or better from there 3 to 19 pulls. You do hear about bad 50 bags.

My 50 bag results were far worse than any of the results I've seen on the spreadsheet so far (although I haven't looked today). Out of 700 tokens from 14 bags, I got no Relics or higher tokens, and only one UR. I plan on getting all my tokens in the future from the treasure boxes.

Between mike and myself we account for 33 of the 50 clear bags, and to get a single UR (or higher) out of 1650 pulls is ridiculously low. I'm 100% prepared to believe that they came from the same mix, but as I have been personally privy to treasure-box shenanigans from volunteers in the past (in this instance they would 'look in the box to see what is in there' and move any high value over to a corner for their kid/partner to draw from ; yes .. it was already reported .. no .. they don't volunteer anymore.) I'm more inclined to believe that it would be easier to have something like that happen ..

I think one UR , and nothing higher in 1650 pulls is low.

Let's Math!

First off, Mike - was 14 bags all your bags, or did you have 1 or 2 good bags and 14 bad ones?

Let's assume there were 33 clear bags which between them had only 1 ur or better.

Let's assume the average rate of >= URs in the mix is as implied from Jedi's sheet: 0.64%.

Odds of drawing 1650 tokens and getting exactly 0 UR or better: 0.9936^1650 ~= 1/40,000

Odds of drawing 1650 tokens and getting exactly 1 UR or better: 1650*0.0064*0.9936^1649 ~= 1/3750

Therefore the odds of getting 1 or fewer URs with a uniform random distribution are ~= 1/3,430.

This is unlikely. It's unlikely enough that if you were publishing a scientific paper on the hypothesis that "The mix was different" you'd get published .

But it's not so unlikely that it absolutely couldn't have happened.

For myself, I got 2 clear bags and found 1 UR in one.

If we add in my 2 bags for 35 total and look at the odds of getting 2 or fewer URs it becomes 1/1000.

Anyone else have bags they care to add to the mix?


By the way, I empathize, last year I had ~400 draws and got 1 UR and 1 Relic (Ice Crag Hero Earcuff). I estimated my average draw to be worth around 2.50, with the average overall draw around 4.50.

Let's only use the data that supports that the clear bags were bad or tampered with. C'mon guys you cannot only use the 'bad' clear bags as a data points. I have said multiple times that my clear bags were on par with my blue bags. True I cannot only use my pulls to extrapolate all the clear, but others have said they got stuff out of the clear bags as well.

Lastly we don't know what Jeff thought of the Greaves and Ring. Most of us are counting them under a UR because their value is less than that, however since they are more limited in ways of obtaining them Jeff may have considered them more of a premium item.

I took photos of my pulls from both clear bags and blue bags. I had three blue bags that had only monster bits, and one of those bags only had four monster bits. My clear bags were not much better, I had three clear bags that had only monster bits. Out of the 10 clear bags I pulled the 25K bar, and out of the blue bags the Legendary token. But that was with over 1K treasure pulls in bags.

Based on last year my expectations for this year:

40% Uncommon
40% Rare
15% Monster Bits
5% "Special"

What we see from the spreadsheet so far:

39.16 Uncommon
36.95 Rare
18.35 Monster Bits
5.54 "Special" (Counting the Darkwood Greaves as "special")

So, overall the big picture distribution is pretty close to what I expected. What looks different is the make up of the "special."

What was announced last year was

2.5% Enhanced/Exalted
1% Reserve Bars
1% Ultra Rare
.5% Relics

The big difference here is .37% Ultra Rares and .03% Reserve Bars.

Harlax wrote: Based on last year my expectations for this year:

40% Uncommon
40% Rare
15% Monster Bits
5% "Special"

What we see from the spreadsheet so far:

39.16 Uncommon
36.95 Rare
18.35 Monster Bits
5.54 "Special" (Counting the Darkwood Greaves as "special")

So, overall the big picture distribution is pretty close to what I expected. What looks different is the make up of the "special."

What was announced last year was

2.5% Enhanced/Exalted
1% Reserve Bars
1% Ultra Rare
.5% Relics

The big difference here is .37% Ultra Rares and .03% Reserve Bars.

Thanks for the compare. That's what I thought. I'm really on the fence regarding Darkwood Greaves as special. At minimum, they are not an equal replacement for URs or Reserve Bars. But then again, this all may be intentional to detract people from being so focused on treasure.

jedibcg wrote:

Matthew Hayward wrote:

Steve wrote:

Mike Steele wrote:

jedibcg wrote: I have said it before and I will say it again the 50 bags are from the same mix. At least 900 maybe 950 of my pulls were from the bags both blue and clear. I spoke to Chandler who made the 50 clear bags. They were 50 tokens from the mix. He didn't look what went in them but did notice lots of colors. I pulled the first 2 clear bags (arpund 130pm) and last 3 on Saturday (around 6pm not certain if they made more). If it seems the box was hotter it is a perception. If someone pulls 3 to 19 from the box and gets something good you will hear about it. You don't hear from everyone that that doesn't an UR or better from there 3 to 19 pulls. You do hear about bad 50 bags.

My 50 bag results were far worse than any of the results I've seen on the spreadsheet so far (although I haven't looked today). Out of 700 tokens from 14 bags, I got no Relics or higher tokens, and only one UR. I plan on getting all my tokens in the future from the treasure boxes.

Between mike and myself we account for 33 of the 50 clear bags, and to get a single UR (or higher) out of 1650 pulls is ridiculously low. I'm 100% prepared to believe that they came from the same mix, but as I have been personally privy to treasure-box shenanigans from volunteers in the past (in this instance they would 'look in the box to see what is in there' and move any high value over to a corner for their kid/partner to draw from ; yes .. it was already reported .. no .. they don't volunteer anymore.) I'm more inclined to believe that it would be easier to have something like that happen ..

I think one UR , and nothing higher in 1650 pulls is low.

Let's Math!

First off, Mike - was 14 bags all your bags, or did you have 1 or 2 good bags and 14 bad ones?

Let's assume there were 33 clear bags which between them had only 1 ur or better.

Let's assume the average rate of >= URs in the mix is as implied from Jedi's sheet: 0.64%.

Odds of drawing 1650 tokens and getting exactly 0 UR or better: 0.9936^1650 ~= 1/40,000

Odds of drawing 1650 tokens and getting exactly 1 UR or better: 1650*0.0064*0.9936^1649 ~= 1/3750

Therefore the odds of getting 1 or fewer URs with a uniform random distribution are ~= 1/3,430.

This is unlikely. It's unlikely enough that if you were publishing a scientific paper on the hypothesis that "The mix was different" you'd get published .

But it's not so unlikely that it absolutely couldn't have happened.

For myself, I got 2 clear bags and found 1 UR in one.

If we add in my 2 bags for 35 total and look at the odds of getting 2 or fewer URs it becomes 1/1000.

Anyone else have bags they care to add to the mix?


By the way, I empathize, last year I had ~400 draws and got 1 UR and 1 Relic (Ice Crag Hero Earcuff). I estimated my average draw to be worth around 2.50, with the average overall draw around 4.50.

Let's only use the data that supports that the clear bags were bad or tampered with. C'mon guys you cannot only use the 'bad' clear bags as a data points. I have said multiple times that my clear bags were on par with my blue bags. True I cannot only use my pulls to extrapolate all the clear, but others have said they got stuff out of the clear bags as well.

Lastly we don't know what Jeff thought of the Greaves and Ring. Most of us are counting them under a UR because their value is less than that, however since they are more limited in ways of obtaining them Jeff may have considered them more of a premium item.

You absolutely can use a subpopulation to determine this sort of thing - because we are using all data for those populations. There is nothing special about mike and steves bags.

Why don't you make an estimate of how many clear bags you pulled, and how many ur or better there were, and we can keep running the numbers.

Maybe we get to place where the odds are less than 1/20.

Dave wrote:

Harlax wrote: Based on last year my expectations for this year:

40% Uncommon
40% Rare
15% Monster Bits
5% "Special"

What we see from the spreadsheet so far:

39.16 Uncommon
36.95 Rare
18.35 Monster Bits
5.54 "Special" (Counting the Darkwood Greaves as "special")

So, overall the big picture distribution is pretty close to what I expected. What looks different is the make up of the "special."

What was announced last year was

2.5% Enhanced/Exalted
1% Reserve Bars
1% Ultra Rare
.5% Relics

The big difference here is .37% Ultra Rares and .03% Reserve Bars.

Thanks for the compare. That's what I thought. I'm really on the fence regarding Darkwood Greaves as special. At minimum, they are not an equal replacement for URs or Reserve Bars. But then again, this all may be intentional to detract people from being so focused on treasure.

If you count the Greaves as ordinary rares, that pushes the rare percentage to about 39 and drops the "Special" to around 3.5.

Matthew Hayward wrote:

jedibcg wrote:

Matthew Hayward wrote:

Steve wrote:

Mike Steele wrote:

jedibcg wrote: I have said it before and I will say it again the 50 bags are from the same mix. At least 900 maybe 950 of my pulls were from the bags both blue and clear. I spoke to Chandler who made the 50 clear bags. They were 50 tokens from the mix. He didn't look what went in them but did notice lots of colors. I pulled the first 2 clear bags (arpund 130pm) and last 3 on Saturday (around 6pm not certain if they made more). If it seems the box was hotter it is a perception. If someone pulls 3 to 19 from the box and gets something good you will hear about it. You don't hear from everyone that that doesn't an UR or better from there 3 to 19 pulls. You do hear about bad 50 bags.

My 50 bag results were far worse than any of the results I've seen on the spreadsheet so far (although I haven't looked today). Out of 700 tokens from 14 bags, I got no Relics or higher tokens, and only one UR. I plan on getting all my tokens in the future from the treasure boxes.

Between mike and myself we account for 33 of the 50 clear bags, and to get a single UR (or higher) out of 1650 pulls is ridiculously low. I'm 100% prepared to believe that they came from the same mix, but as I have been personally privy to treasure-box shenanigans from volunteers in the past (in this instance they would 'look in the box to see what is in there' and move any high value over to a corner for their kid/partner to draw from ; yes .. it was already reported .. no .. they don't volunteer anymore.) I'm more inclined to believe that it would be easier to have something like that happen ..

I think one UR , and nothing higher in 1650 pulls is low.

Let's Math!

First off, Mike - was 14 bags all your bags, or did you have 1 or 2 good bags and 14 bad ones?

Let's assume there were 33 clear bags which between them had only 1 ur or better.

Let's assume the average rate of >= URs in the mix is as implied from Jedi's sheet: 0.64%.

Odds of drawing 1650 tokens and getting exactly 0 UR or better: 0.9936^1650 ~= 1/40,000

Odds of drawing 1650 tokens and getting exactly 1 UR or better: 1650*0.0064*0.9936^1649 ~= 1/3750

Therefore the odds of getting 1 or fewer URs with a uniform random distribution are ~= 1/3,430.

This is unlikely. It's unlikely enough that if you were publishing a scientific paper on the hypothesis that "The mix was different" you'd get published .

But it's not so unlikely that it absolutely couldn't have happened.

For myself, I got 2 clear bags and found 1 UR in one.

If we add in my 2 bags for 35 total and look at the odds of getting 2 or fewer URs it becomes 1/1000.

Anyone else have bags they care to add to the mix?


By the way, I empathize, last year I had ~400 draws and got 1 UR and 1 Relic (Ice Crag Hero Earcuff). I estimated my average draw to be worth around 2.50, with the average overall draw around 4.50.

Let's only use the data that supports that the clear bags were bad or tampered with. C'mon guys you cannot only use the 'bad' clear bags as a data points. I have said multiple times that my clear bags were on par with my blue bags. True I cannot only use my pulls to extrapolate all the clear, but others have said they got stuff out of the clear bags as well.

Lastly we don't know what Jeff thought of the Greaves and Ring. Most of us are counting them under a UR because their value is less than that, however since they are more limited in ways of obtaining them Jeff may have considered them more of a premium item.

You absolutely can use a subpopulation to determine this sort of thing - because we are using all data for those populations. There is nothing special about mike and steves bags.

Why don't you make an estimate of how many clear bags you pulled, and how many ur or better there were, and we can keep running the numbers.

Maybe we get to place where the odds are less than 1/20.

I'll second the point that there was nothing special about our bags.

Harlax wrote: If you count the Greaves as ordinary rares, that pushes the rare percentage to about 39 and drops the "Special" to around 3.5.

Why would you count the Greaves or the RotH as a rare? Should we count the Volunteer Rare token as a Rare? I think they are definitely an Other....now what kind of other I don't know.

Lodestone (KH) wrote:

Matthew Hayward wrote: It's becoming increasingly clear to me that there are "hot spots" in the treasure, and that whatever physical procedure is used to mix the treasure doesn't create a uniform distribution.

And I will add there will never be a way to get a uniform distribution. It just doesn't happen in reality. Look out how galaxies are distributed - they recently found a gap so large between some galaxies it should be statistically impossible. It's just the way things work.

On another note the 50s entries I made were 1 bag of felt (came with 1 3-point enhanced item and a trade 1 item) and 1 plastic bag (had the 2 hallowed rings and 2 greaves). My overall value was $1.67 average for 50s and $1.15 average for pulls.

I need to find someone a lot luckier than me to do my pulling.

Uniform doesn't mean there are no lumpy or hot spots, it means they aren't predictable, e.g. If you pull 2 urs in a row your third token isn't any more or less likely to be a ur.

Like rolling fair dice, that would be uniform at random.

I wouldn't expect the matter in the universe to be distributed uniformly at random - as matter attracts other matter with gravitation, this causing correlations in the placement of matter.



With td treasure If you pull two urs in a row I believe you are much more likely to draw a third in the next pull then you would be otherwise.

This would be like rolling fair dice but if you get a 20 then your next roll counts as a 20 if you roll a 1, and if you roll a 1 your next roll counts as a 1 if you roll a 20. over time the same number of 1s and 20s will come up as in a uniform random distribution, but 20s and 1s will be a little clumped together.

Certainly when I pull monster bits from bags they are found in clumps of 5 or so tokens all touching.

At a guess a handful of the good stuff gets chucked in every so often and then whatever mixing occurs doesn't really move them around much.

kurtreznor wrote:

Matthew Hayward wrote: I added a tab at the end with some guesstimates as to the average value of tokens at various rarities, and the overall average value of a pull this year, which I put around $3.14 from the sheet.

It's not synced with the main sheet, so it won't stay up to date - but I saw someone asking about it so I put together something quick.

docs.google.com/spreadsheets/d/1g1lb5GgfnL0xW5IzZV9VUvsJCzvPU9endLbWoIfmZvI/edit#gid=1996682456

Not meaning to derail the thread, but is $3 for ring of the hallowed accurate? I can see $8 for the new greaves, but the ring is far more rare; was that supposed to be $30?

Also, i dont have exact counts on my uncommon vs rare pulls, but i know my total and the specials i drew; should i add what i know and then split the remainder 50/50 for rares/UC?

Whoops - I read that as ring of holly.

Sure - that should be maybe 25.

People have been trying unsuccessfully to sell these in the 40 to 50 range, before the con when they became more plentiful.

Matthew Hayward wrote:
Whoops - I read that as ring of holly.

Sure - that should be maybe 25.

People have been trying unsuccessfully to sell these in the 40 to 50 range, before the con when they became more plentiful.

I will sell mine at 20 until I run out, but I am guessing I am not going to see much movement on them. They are a collector's token I think most would agree and I don't know that there are that many collectors left without them already.

jedibcg wrote:

Harlax wrote: If you count the Greaves as ordinary rares, that pushes the rare percentage to about 39 and drops the "Special" to around 3.5.

Why would you count the Greaves or the RotH as a rare? Should we count the Volunteer Rare token as a Rare? I think they are definitely an Other....now what kind of other I don't know.

Because they are rares.

On the other hand, they are only obtainable from the treasure box. That's why I counted them as special in my analysis. As to how special, I'd put them on par with the enhanced transmutes.

I see they are in TokenDB already. tokendb.com/?fwp_slot=shins