[User Collected] Recruit token probability

user311

Looked through the spreadsheet. Thanks to those contributing. It seems like its pretty rare to get the *** and above. So I have to ask where is everyone getting their covers from? Throughout this forum I see people talking about maxed out players or 5/3/5 this or that. I have been playing non stop and cannot seem to get anyone maxed out above 2 stars. I would pay for a specific cover but I would probably be throwing my money away if it gave me another purple GSBW.

PorkBelly

user311 wrote:

Looked through the spreadsheet. Thanks to those contributing. It seems like its pretty rare to get the *** and above. So I have to ask where is everyone getting their covers from? Throughout this forum I see people talking about maxed out players or 5/3/5 this or that. I have been playing non stop and cannot seem to get anyone maxed out above 2 stars. I would pay for a specific cover but I would probably be throwing my money away if it gave me another purple GSBW.

Either luck (just got a Rag red from a standard token) or through PVP/PVE tournament rewards. I built up my Patch through the Thick as Thieves subbrackets PVE. That seems to be the best way to get a specific cover. However, lots of covers are still pretty rare.

I still don't have a Loki.

Pwuz_

I'm not convinced that there really is a way to properly collect this data. I've been wondering if the tokens we pull are from a set queue that is shared across all users on a 1st come 1st serve basis. If that is the case, without having the data from all users & times pulled, the results may not be consistent. In which case, you would increase your odds of pulling 3* tickets by stockpiling your tokens and using them in groups. That wouldn't prevent someone else from pulling one from between yours, but it would group them . Sorta like when playing scratch off lotto tickets; you have a better chance of getting a winner if you take a bunch from the same roll all together rather than spacing them out.

Just a thought, I don't want to dash your hopes and dreams, and your method (if true) would be far more fair than if the method I suggest were instead. I'm just throwing out alternative possible means of programing the system.

And that doesn't even begin to touch the idea that our tokens could be predetermined before we collect them.

Yeggy

Pwuz_ wrote:

I'm not convinced that there really is a way to properly collect this data. I've been wondering if the tokens we pull are from a set queue that is shared across all users on a 1st come 1st serve basis. If that is the case, without having the data from all users & times pulled, the results may not be consistent. In which case, you would increase your odds of pulling 3* tickets by stockpiling your tokens and using them in groups. That wouldn't prevent someone else from pulling one from between yours, but it would group them . Sorta like when playing scratch off lotto tickets; you have a better chance of getting a winner if you take a bunch from the same roll all together rather than spacing them out.

Just a thought, I don't want to dash your hopes and dreams, and your method (if true) would be far more fair than if the method I suggest were instead. I'm just throwing out alternative possible means of programing the system.

And that doesn't even begin to touch the idea that our tokens could be predetermined before we collect them.

If you see the data collected so far - particularly in standard tokens, which has the greatest amount collected - you can see a hint of proper statistical distribution. And as with many other game, d3 likely uses a pseudo-random number generator.

Having pre-determined covers for each player is silly and is harder to implement, anyway. Same goes for having a pull queue. Some form of randomness is easier to implement.

The one thing that is diffucult to ascert are the skewed probabilities in special token events towards a particular 3* or 2* as usually not enough covers are opened for those tokens.

There is also the problem with the introduction of new characters, which changes the chances and screws with the previously accumulated data. You can see the example of ARES in standard tokens, which was introduced later and has a lower % chance than other 2*. A lot of packs were opened before this Dark Avenger's time.

I guess what I want to say is: it's easier to implement some form of random generator for tokens than having any sort of "pull queue" or having pre-determined covers for each player. Also the data collected so far simply points away from this conclusion. Honestly, I see no good reason why anyone would do anything like that in the first place.

Pwuz_

True. It's just in my nature to look for alternatives when finding logic in situations.

Yeggy

Pwuz_ wrote:

True. It's just in my nature to look for alternatives when finding logic in situations.

That's a good habit to have, but I hope that House M.D. taught you as much as it did to me. Occam's razor.

EDIT: Also, it's never lupus.

Mawtful

Is there any clue regarding the contents of the current Vibranium Token. Generally there are more than just the pictured heroes available.

The tab shows Black Panther, Thor, Wolverine and Bullseye. Currently my Thor is at 2 covers only (sold my old one off for some quick ISO post-balance) and I'm trying to decide if I go for a 10 pack. If those are the only heroes, then 6 out of 11 covers will go to good use. Of course 3 of those, BP, are less likely to show up than the other 8, and I'll definitely be annoyed with any Bullseye covers I pull.

I guess my other option is to wait a few more days for the start of the next PvE event, which will undoubtedly feature BP, and might also feature IM40 who I also need covers for (he's a reward in the current PvP).

mischiefmaker

The text says "contains a ** character or Black Panther", which is consistent with event tokens from other PvP events. We have pretty good reason to believe that the way it works is that there is a roughly 84% chance you get a 2* cover, with all 2* covers being equally likely (so Bullseye and Daken are 33% less likely than the other characters, as they only have 2 covers to others' 3), and a roughly 16% chance to get a 3* cover -- in this case, Black Panther.

Mawtful

mischiefmaker wrote:

The text says "contains a ** character or Black Panther", which is consistent with event tokens from other PvP events. We have pretty good reason to believe that the way it works is that there is a roughly 84% chance you get a 2* cover, with all 2* covers being equally likely (so Bullseye and Daken are 33% less likely than the other characters, as they only have 2 covers to others' 3), and a roughly 16% chance to get a 3* cover -- in this case, Black Panther.

Oh, so this could contain any 2* character? (Daken, Captain America, etc)?

That's much less appealing. I think I can wait a few days until the next PvE event - since PvE tokens tend to have a much more limited pool of possible covers, right? (Powered-Up characters)

Unknown

I created a tab for Pure Vibranium, and also a tab for standard tokens after patch 48.0. It's a better idea to fill your standard tokens in this new tab since the data in the other standard tab is getting irrelevant due to the new items in the game such has new heroes and 25hp.

mischiefmaker

Mawtful wrote:

Oh, so this could contain any 2* character? (Daken, Captain America, etc)?

That's much less appealing. I think I can wait a few days until the next PvE event - since PvE tokens tend to have a much more limited pool of possible covers, right? (Powered-Up characters)

Yes, typically PvP tokens are any 2* cover (sometimes with increased chance of a particular 2*, like Magneto), and usually limited to one 3* (though there are exceptions to this too, like Divine Champions, which could have been Loki or Ragnarok). PvE tokens tend to be limited selection for both 2* and 3*, but then of course you're at the whim of which ones are boosted for that event.

Microtom wrote:

I created a tab for Pure Vibranium, and also a tab for standard tokens after patch 48.0. It's a better idea to fill your standard tokens in this new tab since the data in the other standard tab is getting irrelevant due to the new items in the game such has new heroes and 25hp.

Thanks for the help! I'm going to combine these sheets -- I actually already created a mechanism in the standard tab to track pulls since the recent change; it's just hidden. Having everything in one sheet should be much less confusing.

Mawtful

Thanks for the info.

After looking at Microtom's data on the Vibranium tokens I went ahead and grabbed a 10 pack anyway. I've added my results to the page and I'll keep a track of my future covers there.

Unknown

This is a very nice spreadsheet, especially after I found all the other tabs, grats.

It would be also nice to know and gather data about the color spread. Is that completely random? Or some skills are more rare than others? I know it would triple the size of the sheets, and it is kinda late to rework all the tables, but it would be still nice-to-know.

mischiefmaker

I didn't include colors in the original sheet because it seemed like too much information and not enough sample size. IceIX has since hinted that all covers are equal probability, which is supported by our data (take a look at IM35/Storm vs. the other 1*s, or Daken/Bullseye v. the other 2*), so we haven't bothered adding it.

Unknown

Hey Mischief: the spreadsheet gave me a notice this morning that it's close to reaching the limit of formulae in one document. Any plan for what to do about that? There may be "empty" formulae sitting out on the right end of many tabs, where there were blank columns left waiting for players who never logged anything. I'll start by going through old events from farthest right tab in and deleting any unused columns.

EDIT: Done. Every tab further right than Diabolical has been cleaned up. I left one empty column at the far right of each tab. Some tabs had 26+ empty columns off to the right, so in theory, I just reduced the number of formulae in the document by around half. That should keep us set for another few months.

mischiefmaker

Excellent, thanks for taking care of that.

Unknown

Opened 100 Standard Tokens:

* (81)
Iron Man: 6/4/5
Storm: 6/4/3
Juggernaut: 6/6
Hawkeye: 5/7
Yelena: 8/5
Venom: 4/2
Black Widow: 3/7

** (10)
Storm: 1/0/0
Spider-Man: 0/0/1
Ares: 1/1/0
Hawkeye: 1/0/1
Black Widow: 0/1/0
Magneto: 0/0/1
Wolverine: 1/0/0
Captain America: 0/1/0

*** (4)
Black Widow: 0/1/0
The Hood: 1/0/0
Ragnarok: 1/0
Punisher: 0/0/1

Hero Points: 5

Unknown

I must've got the luckiest ever. I was spending 4000 ISO on standard tokens as I had no-one to level up, and I just kept getting the mundane Black Widows and Yelena Belovas, until my last 2 tokens. I actually managed to get 2 Patch Wolverines in a row. I couldn't believe it
However, those are the last rare things I've got, the best since have been a couple of Dakens and a Captain America.

Unknown

Just made my first contribution to the spreadsheet, 30 Standard tokens, very odd pulls. A whopping *10* Modern Storms, and 3 lots of HP, even though I only got 1 2* character and no 3*... I've opened quite a few before now, hundreds probably, and I haven't gotten any 3/4*, but I do seem to keep getting hero points...

Unknown

I have gotten nearly identical results...