So, I made a thing... (a manifest completion badge)
#201
Posté 31 août 2012 - 12:05
#202
Posté 31 août 2012 - 12:14
http://www.blade.io/...818&invert
Seriously?? SEVEN? No, just...no...
Oh, now its a four!!...15 packs later...<_<
http://www.blade.io/...eUR&invert
Yeeeaaaahhhhh...I don't think i'll still be playing at that point...
Modifié par Yajuu Omoi, 10 septembre 2012 - 07:21 .
#203
Posté 31 août 2012 - 12:14
#204
Posté 31 août 2012 - 12:17
http://www.blade.io/...827&invert
http://www.blade.io/...827&invert
Time to go get 90+ packs.... >.>
Modifié par ChickenDownUnder, 31 août 2012 - 12:21 .
#205
Posté 31 août 2012 - 12:22
http://www.blade.io/...111&invert
Modifié par WaffleCrab, 31 août 2012 - 12:23 .
#206
Posté 31 août 2012 - 12:29
#207
Posté 31 août 2012 - 12:40
Rolenka wrote...
Awesome, thanks! This is amazing!
http://www.blade.io/...01067385
http://www.blade.io/...01067385
^ for my rares.
Does your completion estimator factor in that the closer you get to the goal, the more ammo you get and the slower it goes? Or is that not true?
And would it be possible to get a completion estimator for uncommons? I'm a greedy greedy noob, I know.
It doesn't take Ammo IV drops into account. It does seem to be true that they drop all too frequently, bassed on what I've heard from longterm players.
Uncommons are a tricky thing because their drop rates seem weird to me. Sometimes I'll get one in the Vet pack, sometimes I'll get two. I'm not sure anyone has reliable enough numbers to do a good estimator.
Modifié par GuyBlade, 31 août 2012 - 12:43 .
#208
Posté 31 août 2012 - 12:44
WaffleCrab wrote...
not sure if it works for me
http://www.blade.io/...111&invert
Your profile is private. I can't pull data on private manifests.
#209
Posté 31 août 2012 - 12:54
randomfoo wrote...
GuyBlade wrote...
randomfoo wrote...
Hey, first of all, this is awesome. Good job!
I think there's a bug w/ your til completion code though:
I have 10 UR drops left (so close!)
At 10% PSP UR drop rate, this should be 990K/UR drop, or 100 drops required (9900k total) left? Your code seems to give half of that. Maybe it's missing counting one of the guns? (I have a Harrier V and Typhoon V)
http://www.blade.io/...eUR&invert
BTW, a few of us have been tracking PSP UR drop rates (direct link to spreadsheet) and I'm fairly certain (statistically speaking!) the drop rate is higher than 10%.
Right now the drop rate is 13.5% with a sample size of 652 packs, so at a 95% confidence level, the +/- is 2.6%. The drop rate is at least 11%, and up to 16% (unlikely).
(feel free to pull the drop rates from this sheet. if you want to dynamically pull it, drop me a line, i've written a lot of gdata code in the past, should be trivial in any language. also, if anyone wants to help add their numbers to the sheet, drop me a pm)
I think pulling the empirical data from the sheet is a fine idea. I've usually found that the easiest way to pull from a sheet is to enable "Publish to Web" and just grab the CSV. If you wouldn't mind enabling it (I don't think there are any security concerns), I'll update my code to periodically pull from your sheet.
Unfortunately, I think doing so will make depressingly large numbers even larger.
Heh, well I noticed that bug was fixed so my numbers just doubled, but with the empirical data it should split the difference (about 730K/UR vs 990K/UR).
Here's the CSV link: https://docs.google....aMHc&output=csv
Thanks! I've updated it so that it pulls from your spreadsheet a few times a day with updated values. I also included some text in the description crediting the spreadsheet as the source.
#210
Posté 31 août 2012 - 01:29
(apologies if someone else has already asked) is there a link to raw stats without the extra code? RSS feed / raw XML data? I'd love to play with some slicing and dicing but lack time
#211
Posté 31 août 2012 - 02:28
Xenmir wrote...
A neat idea GB. Depressing to see how much is left for UR though. lol.
How'd you get yours looking like that?
#212
Posté 31 août 2012 - 02:53
#213
Posté 31 août 2012 - 03:20
dstrawberrygirl wrote...
This is fantastic, nice work! With this and Manifest Detailer, the community are doing a great job of analysing the raw data. Guessing your server is getting hit pretty hard with all these images in signatures?
(apologies if someone else has already asked) is there a link to raw stats without the extra code? RSS feed / raw XML data? I'd love to play with some slicing and dicing but lack time
The wonderful thing about png is that it conpresses very well when you limit the number of colors. Each badge is less than 1 kilobyte.
There's not currently a place with raw data. I do keep an intermediate data blob in json format, but I don't know that I want to be in the business of being a data supplier. If you're interested in what I keep, I've put a sample file here:
http://www.blade.io/...ample-data.json
If you have any good ideas of something to do with the data, feel free to suggest.
Modifié par GuyBlade, 31 août 2012 - 03:23 .
#214
Posté 31 août 2012 - 03:33
#215
Posté 31 août 2012 - 03:45
#216
Posté 31 août 2012 - 03:50
GuyBlade wrote...
dstrawberrygirl wrote...
This is fantastic, nice work! With this and Manifest Detailer, the community are doing a great job of analysing the raw data. Guessing your server is getting hit pretty hard with all these images in signatures?
(apologies if someone else has already asked) is there a link to raw stats without the extra code? RSS feed / raw XML data? I'd love to play with some slicing and dicing but lack time
The wonderful thing about png is that it conpresses very well when you limit the number of colors. Each badge is less than 1 kilobyte.
There's not currently a place with raw data. I do keep an intermediate data blob in json format, but I don't know that I want to be in the business of being a data supplier. If you're interested in what I keep, I've put a sample file here:
http://www.blade.io/...ample-data.json
If you have any good ideas of something to do with the data, feel free to suggest.
It might be neat to run some simple stats. Probably for all the weapons, gears, and consumables. So the average level (I-X) or amount, and the distribution (x % of the community has the Harrier I, II, III... X). Seeing the distributions for gear and consumables of course would be great (x% of the community has no medi-gel. would explain so much
Anyway, seems like you could basically have 2 functions and run them across every single stat to generate some neat graphs.
#217
Posté 31 août 2012 - 03:55
Modifié par Korisselys1, 31 août 2012 - 01:25 .
#218
Posté 31 août 2012 - 04:04
#219
Posté 31 août 2012 - 04:06
count me in
#220
Posté 31 août 2012 - 04:13
#221
Posté 31 août 2012 - 04:17
So without BW supplying more data I guess this is wishful thinking, but wouldn't it be great to have some recommendations for highest DPS Assault Rifle for your currently-specced out level 18 Turian Soldier, taking into consideration the options chosen for each power rank, or a selection of best power-to-weight ratio shotguns for your Vanguard etc.randomfoo wrote...
It might be neat to run some simple stats. Probably for all the weapons, gears, and consumables. So the average level (I-X) or amount, and the distribution (x % of the community has the Harrier I, II, III... X). Seeing the distributions for gear and consumables of course would be great (x% of the community has no medi-gel. would explain so much
Anyway, seems like you could basically have 2 functions and run them across every single stat to generate some neat graphs.
#222
Posté 31 août 2012 - 04:20
#223
Posté 31 août 2012 - 04:22
And it's unlikely I'll play long enough to get that many credits. Sheesh.
#224
Posté 31 août 2012 - 04:29
Modifié par solomsr, 31 août 2012 - 04:34 .
#225
Posté 31 août 2012 - 04:33
dstrawberrygirl wrote...
So without BW supplying more data I guess this is wishful thinking, but wouldn't it be great to have some recommendations for highest DPS Assault Rifle for your currently-specced out level 18 Turian Soldier, taking into consideration the options chosen for each power rank, or a selection of best power-to-weight ratio shotguns for your Vanguard etc.randomfoo wrote...
It might be neat to run some simple stats. Probably for all the weapons, gears, and consumables. So the average level (I-X) or amount, and the distribution (x % of the community has the Harrier I, II, III... X). Seeing the distributions for gear and consumables of course would be great (x% of the community has no medi-gel. would explain so much
Anyway, seems like you could basically have 2 functions and run them across every single stat to generate some neat graphs.
Actually we have all the weapon info (Cyanon maintains that spreadsheet), but the specific question you want answered is easy since the only decent AR is the Harrier
Anyway, in all seriousness, if you're min-maxing, there are only a few top-tier weapons (1 or 2 a class), so that sort of recommendation engine you're asking for is a bit pointless IMO.





Retour en haut




