Updated Ultra Rare Drop Rate Data | All sample size goals reached | Will still do occasional updates | Links fixed 2015 April 8
#226
Posté 14 novembre 2013 - 07:47
Actually my theory is that there is a single probability for the UR drop-rate in a single gold-slot. That would mean, the drop-rate is the same for PSP - AP - RP, and the cost is higher for the SP due to the higher cost: 120K (2x 60K) vs 99K to purchase 2 of those slots.
#227
Posté 14 novembre 2013 - 07:56
Zso_Zso wrote...
It seems that the SP interval no longer intersects the other 3, so now some conclusion can be made, i.e. that getting URs via SPs costs more than if buying any of the 99K packs.
Actually my theory is that there is a single probability for the UR drop-rate in a single gold-slot. That would mean, the drop-rate is the same for PSP - AP - RP, and the cost is higher for the SP due to the higher cost: 120K (2x 60K) vs 99K to purchase 2 of those slots.
The conclusion about SPs having a higher long term cost per UR than the other packs was added to the document and the original post when I updated them on Tuesday. As for all the packs having the same per slot drop rates, the confidence intervals for the per slot drop rates all currently support this but i would like to get the ones for SPs, APs, and RPs narrowed down a bit more so that we will be able to tell if there are any small differences.
#228
Posté 14 novembre 2013 - 08:50
Does the evidence support something like this if the goal is to get the last of the UR weapons?
#229
Posté 14 novembre 2013 - 08:53
As far as I know, the UR drop rate is not affected by anything. So although it's good to be at cap on consumables, your friend is wasting money he could be spending on Cryo Rounds IV.1badmage wrote...
One of my Origin friends has the strategy: 1) first buy jumbo equipment packs until all consumables are maxed at 255, and then 2) buy PSPs.
Does the evidence support something like this if the goal is to get the last of the UR weapons?
Modifié par FatherOfPearl, 14 novembre 2013 - 08:53 .
#230
Posté 14 novembre 2013 - 10:18
#231
Posté 14 novembre 2013 - 10:41
#232
Posté 15 novembre 2013 - 01:34
#233
Posté 15 novembre 2013 - 08:50
#234
Posté 15 novembre 2013 - 08:57
1badmage wrote...
One of my Origin friends has the strategy: 1) first buy jumbo equipment packs until all consumables are maxed at 255, and then 2) buy PSPs.
Does the evidence support something like this if the goal is to get the last of the UR weapons?
That's the only way to maximize your chances. Don't waste money on Reserve and Arsenal Packs, not to mention simple Spectre Packs.
In fact , recruit packs are better to maximize your missiles, medigels , ops packs and ammo packs if you have just a few missing. My strategy was to play a platinum game, then with the 150 000 credits I bought 5-6 recruit packs to maximize my gear back to 255 , then 1 PSP.
Modifié par maaaad365, 15 novembre 2013 - 08:58 .
#235
Posté 15 novembre 2013 - 09:00
maaaad365 wrote...
1badmage wrote...
One of my Origin friends has the strategy: 1) first buy jumbo equipment packs until all consumables are maxed at 255, and then 2) buy PSPs.
Does the evidence support something like this if the goal is to get the last of the UR weapons?
That's the only way to maximize your chances. Don't waste money on Reserve and Arsenal Packs, not to mention simple Spectre Packs.
In fact , recruit packs are better to maximize your missiles, medigels , ops packs and ammo packs if you have just a few missing. My strategy was to play a platinum game, then with the 150 000 credits I bought 5-6 recruit packs to maximize my gear back to 255 , then 1 PSP.
Until I see evidence supporting this, I'd like to point out that the 'R' in 'RNG' stands for 'Random', so there is no way to alter your chances of getting a UR.
#236
Posté 15 novembre 2013 - 09:09
Then only valid evidence that would solve this problem is to get a new game, get the money cheat and start buying PSP , AR and RP. Change the type of pack you buy unce you reach 25 UR and count how many packs you needed to get there.
#237
Posté 15 novembre 2013 - 09:16
This very nice .pdf in the OP says that all the packs have similar drop rates, and that people will actually average more URs per credit by buying APs and RPs.maaaad365 wrote...
The PSP description states clearly : "higher chance for ultra-rares". In the long run you do get better results. Of course you can get lucky and get 2 UR in a Reserve Pack, but that's just random.
Then only valid evidence that would solve this problem is to get a new game, get the money cheat and start buying PSP , AR and RP. Change the type of pack you buy unce you reach 25 UR and count how many packs you needed to get there.
Modifié par FatherOfPearl, 15 novembre 2013 - 09:17 .
#238
Posté 15 novembre 2013 - 09:30
#239
Posté 15 novembre 2013 - 10:34
If it helps, we have tons of data points on packs that were collected here:
https://docs.google....c&rm=full#gid=0
I probably could have designed the spreadsheet better; as is, there are ambiguities in terms of actual drop rates.
The numbers from your report match up very closely with what I have reported. Mostly likely 7.5% UR drop rate per Rare slot in SPs and above. Arsenal Packs are probably best value for URs because they don't drop Common Character cards.
I also have very convincing evidence that they tinkered with UR drop rates in the past.
Modifié par InterwebNinja, 15 novembre 2013 - 10:40 .
#240
Posté 15 novembre 2013 - 07:13
InterwebNinja wrote...
I wasn't aware of this data collection effort on BSN, as I rarely read this forum, but I've pursued a similar effort on the reddit forum /r/MECoOP.
If it helps, we have tons of data points on packs that were collected here:
https://docs.google....c&rm=full#gid=0
I probably could have designed the spreadsheet better; as is, there are ambiguities in terms of actual drop rates.
The numbers from your report match up very closely with what I have reported. Mostly likely 7.5% UR drop rate per Rare slot in SPs and above. Arsenal Packs are probably best value for URs because they don't drop Common Character cards.
I also have very convincing evidence that they tinkered with UR drop rates in the past.
That data will help but what is this evidence of drop rate tinkering?
#241
Posté 15 novembre 2013 - 10:07
For the prior DLC, I showed a much higher drop rate for PSPs than had been previously reported - I showed 18% compared to something closer to 12% before that. Several hundred pack samples for each. Again... possible it's always been 15% and that was noise, but statistical comparisons of the two samples make it very unlikely. That said, I wasn't quite as thorough in those data collection efforts as I was with the spreadsheet I linked. That's the most recent and seemingly stable set of data I have.
My guess is that these were behind-the-scenes 'promotions', in order to boost spending on packs after new DLC was released. It would make sense - people who start getting a bunch of URs are probably more likely to keep spending money in the store. But, that's just my speculation.
Modifié par InterwebNinja, 15 novembre 2013 - 10:08 .
#242
Posté 15 novembre 2013 - 11:26
It would be so sweet to land a double pack to end my suffering...
#243
Posté 16 novembre 2013 - 12:19
InterwebNinja wrote...
It was only a couple short periods right after new DLC was released. After they dropped Arsenal Packs / Reserve Packs, I had data on several hundred of them that showed they were dropping at ~30%. Statistically it's possible that that was just noise... but very, very unlikely.
For the prior DLC, I showed a much higher drop rate for PSPs than had been previously reported - I showed 18% compared to something closer to 12% before that. Several hundred pack samples for each. Again... possible it's always been 15% and that was noise, but statistical comparisons of the two samples make it very unlikely. That said, I wasn't quite as thorough in those data collection efforts as I was with the spreadsheet I linked. That's the most recent and seemingly stable set of data I have.
My guess is that these were behind-the-scenes 'promotions', in order to boost spending on packs after new DLC was released. It would make sense - people who start getting a bunch of URs are probably more likely to keep spending money in the store. But, that's just my speculation.
Did you look at anything other than a point estimate? Either way could you give me this data so I can look at it myself.
#244
Posté 16 novembre 2013 - 12:34
#245
Posté 16 novembre 2013 - 12:45
ReaperRed wrote...
58 reserves packs, 12 URs. Manifest done. Well, I have a mostly empty Xbox and PS3 account, I suppose, so I guess I'll still provide those numbers periodically.
Nice!
Data:
5 rp 0 ur
5 sp 2 ur
#246
Posté 16 novembre 2013 - 01:36
ComradeShepard7 wrote...
Did you look at anything other than a point estimate? Either way could you give me this data so I can look at it myself.
I just did confidence intervals... (in the tab 'Other Unlocks' (which predated Arsenal and Reserve packs), the differences in the data we collected shortly after the DLC was sinificantly higher than what we collected prior to that (rows highlighted in blue). 18% unlock rate +/- 3% vs. 9% unlock rate +/- 3%. Very hard samples to reconcile.
Some other data had led me to believe drop rates were closer to 12% for PSPs during non-DLC periods, but I can't seem to find records of that.
As for the Arsenal and Reserve packs, in retrospect, I'm not sure my samples at the time were large enough to justify any strong claims about there being a boost shortly after they were released. While the first set of packs to come in did unlock at close to 30%, the sample was pretty small (65). The lower bound on the confidence interval is 17%, which is still higher than the ~15% we are seeing now, but I don't think it's nearly as compelling as the other sample disparity for PSPs.
Regardless, additional data is here. The data from 'My Unlocks' is what I started recording shortly after I switched over to PC earlier this year, and it is not reflected in any of the other tabs in this spreadsheet, so you may way to add that to your set as well.
Modifié par InterwebNinja, 16 novembre 2013 - 01:36 .
#247
Posté 16 novembre 2013 - 03:06
InterwebNinja wrote...
I just did confidence intervals... (in the tab 'Other Unlocks' (which predated Arsenal and Reserve packs), the differences in the data we collected shortly after the DLC was sinificantly higher than what we collected prior to that (rows highlighted in blue). 18% unlock rate +/- 3% vs. 9% unlock rate +/- 3%. Very hard samples to reconcile.
Some other data had led me to believe drop rates were closer to 12% for PSPs during non-DLC periods, but I can't seem to find records of that.
As for the Arsenal and Reserve packs, in retrospect, I'm not sure my samples at the time were large enough to justify any strong claims about there being a boost shortly after they were released. While the first set of packs to come in did unlock at close to 30%, the sample was pretty small (65). The lower bound on the confidence interval is 17%, which is still higher than the ~15% we are seeing now, but I don't think it's nearly as compelling as the other sample disparity for PSPs.
Regardless, additional data is here. The data from 'My Unlocks' is what I started recording shortly after I switched over to PC earlier this year, and it is not reflected in any of the other tabs in this spreadsheet, so you may way to add that to your set as well.
Just to make sure I'm reading things right before I continue, the blue highlighted rows are pre DLC data while all the other rows are post DLC data, and I'm guessing that the DLC in question is Retaliation due to the presence of UR characters afterwards and one person getting 5 UR character drops?
Also, are you including both UR weapons and UR characters in your post DLC calculation to find an "overall" UR drop rate?
#248
Posté 16 novembre 2013 - 03:13
The_Bastage wrote...
I have 2 UR cards left to go. I've spent 1.2 mil so far with nothing. Damn RNG!
It would be so sweet to land a double pack to end my suffering...
Well, I look forward to getting more data when you max your manifest in the, hopefully, near future.
Gr1ng0 wrote...
Data:
5 rp 0 ur
5 sp 2 ur
ReaperRed wrote...
58 reserves packs, 12 URs. Manifest done. Well, I have a mostly empty Xbox and PS3 account, I suppose, so I guess I'll still provide those numbers periodically.
And I still have 13 to go since I have instisted on continuing to take SP data. At least I have a decent number of Adrenaline Mod III's for once.
#249
Posté 16 novembre 2013 - 03:14
First 8 had nothing,
two in a row with a single UR,
one with nothing,
one with a single UR,
eight in a row with nothing,
one with a single UR,
the final thirteen had nothing.
Total, 34 AP's, 4 UR's, 0 doubles.
#250
Posté 16 novembre 2013 - 03:24
ComradeShepard7 wrote...
Just to make sure I'm reading things right before I continue, the blue highlighted rows are pre DLC data while all the other rows are post DLC data, and I'm guessing that the DLC in question is Retaliation due to the presence of UR characters afterwards and one person getting 5 UR character drops?
Also, are you including both UR weapons and UR characters in your post DLC calculation to find an "overall" UR drop rate?
That is the correct interpretation with the blue highlighted fields, and yes, that particular sheet was from Retaliation. My UR drop rate does not include UR characters. As far as I can tell, UR character drops are indepedent from UR weapons (i.e. UR weapons drop rate is the same whether you have a maxed UR character manifest or not).
edit: side note... I'm not enthralled with symmetric confidence intervals as a metric for measuring likelihoods (i.e probabilites between 0 and 1). The more likely (or unlikely) an event becomes, the more asymmetric the distribution becomes. That, is if something is measured at a 10% rate, you may get a confidence interval of +/- 5%, but it is much more likely the true rate is 15% than it is 5%. I tried to research this problem but didn't come up with much useful. I'm sure the problem has been studied, but I'm not a stats expert and maybe just don't know the right search terms.
Modifié par InterwebNinja, 16 novembre 2013 - 05:27 .





Retour en haut




