| Forum Index > Diablo 3 |
|
dAPhREAk
Nauru12397 Posts
On January 16 2013 09:56 ximae wrote: i gave u some links some posts above, there was also some spread sheet by ruler eric gathering info from various players but id really have to dig the battlenet forums for that link. from the posts you sent me: ![[image loading]](http://i.imgur.com/DMYYI.png) so, mp0 is best? lol. | ||
|
DODswe4
Sweden2157 Posts
On January 16 2013 10:01 dAPhREAk wrote: from the posts you sent me: so, mp0 is best? lol. I believe most people and stats show that the mp where you 1shot elites is the best for farming. this can obv be different for classes. | ||
|
Eschaton
United States1245 Posts
On January 16 2013 09:40 Infernal_dream wrote: The best MP level is one you can one shot all white mobs and burn elites in under 5-10 seconds. You can run a test 100 times. RNG is called RNG for a reason. There's nothing guaranteed in these runs (other than at least 1 yellow per elite at mp10). Law of averages won't work because you can get an extremely lucky mp1 run and an extremely unlucky mp9 that's going to skew the results. While the chances of it happening many times over aren't good it still can. Sorry, did you just say the Law of Large Numbers* won't work because Outliers exist? There is an Expectation Value associated with most** distributions of outcomes, and the average of all results will converge to this number as repetitions approach infinity. We do not know the expected values at this point, the data does not exist in our hands. No one is doing a single run and saying, holy fuck, I just got a double leg drop, MP0 is best. Nor are they saying, I just did 10 runs in a row and didn't get a single leg from MP10, MP10 is the worst. People are saying, hey, we don't know the expected value, but this is my anecdotal experience. Some on the b.net forums are making an attempt to use data but still the amount of data are very limited. Terms you may want to google are written with Caps. * "Law of averages" **I assume that Blizzard chose a non-pathological function for their loot generator. | ||
|
ximae
181 Posts
| ||
|
dAPhREAk
Nauru12397 Posts
On January 16 2013 10:09 ximae wrote: why are we discussing this when ptr is up? because im still at work. ;-) | ||
|
Arkani
Austria60 Posts
with the blizzard dmg buff i hope that my midbudget sorc is able to do faster mp0 runs than archom (boring as hell), can you give me some advice how the blizzard build should look and is there ana specific gear i should watch out for? black weapon + trivium, zuni and tal amu or still just get standard high crit dmg stuff? are frostburns worth it? thanks in advance!  | ||
|
dAPhREAk
Nauru12397 Posts
On January 16 2013 17:00 Arkani wrote: hey guys with the blizzard dmg buff i hope that my midbudget sorc is able to do faster mp0 runs than archom (boring as hell), can you give me some advice how the blizzard build should look and is there ana specific gear i should watch out for? black weapon + trivium, zuni and tal amu or still just get standard high crit dmg stuff? are frostburns worth it? thanks in advance! oculus with blizzard/teleport bonus, frostburn gauntlets, stone of jordan with blizzard bonus, maybe storm crow with blizzard bonus. not sure which have blizzard bonus, but those are probably going to be the best options. | ||
|
JerKy
Korea (South)3013 Posts
i tried using a mempo, but i find it a lot harder without the apoc, so i figured i'd have the same problem if i used a tal rasha's helm (i also don't have apoc on my offhand, which is something i'll be upgrading soon) here is my bnet profile if it'll help | ||
|
Doso
Germany769 Posts
| ||
|
JerKy
Korea (South)3013 Posts
| ||
|
Doso
Germany769 Posts
 You can sometimes can snipe cheap Storm Crows with crit. I got a 4CC one for like half the price it normaly costs. | ||
FallDownMarigold
United States3710 Posts
Nabbed it on buyout, and it was 1d11h. Surely if there were actually t3h_leet bot netz dominating the AH this would not have been possible for me, no? Bot would want to snipe at 1d12h ofc. So maybe there are or were some bots, but it can't be so bad given what happened with my tri-pox | ||
|
Infernal_dream
United States2359 Posts
On January 16 2013 10:04 Eschaton wrote: Sorry, did you just say the Law of Large Numbers* won't work because Outliers exist? There is an Expectation Value associated with most** distributions of outcomes, and the average of all results will converge to this number as repetitions approach infinity. We do not know the expected values at this point, the data does not exist in our hands. No one is doing a single run and saying, holy fuck, I just got a double leg drop, MP0 is best. Nor are they saying, I just did 10 runs in a row and didn't get a single leg from MP10, MP10 is the worst. People are saying, hey, we don't know the expected value, but this is my anecdotal experience. Some on the b.net forums are making an attempt to use data but still the amount of data are very limited. Terms you may want to google are written with Caps. * "Law of averages" **I assume that Blizzard chose a non-pathological function for their loot generator. So once again, we don't have enough proof. So nobody should be arguing anything. As I stated RNG will be RNG. We all know the theoretical best is the MP which you absolutely annihilate everything. Whether that be 0 or 5 depending on dps and armor ratings and such. Which is pretty much exactly what I was saying. You can run it a billion times. The number of outliers is still going to be way too high to get any reasonable results. Unless we get blizzard to put the algorithms for their loot generators on the internet then we will honestly never know. But as of now nobody can say an MP level where you take longer to kill mobs is actually more efficient than one where you one shot everything. | ||
|
ximae
181 Posts
On January 17 2013 08:10 FallDownMarigold wrote: I think the fear that bots heavily populate the AH is a tad overblown, because after monitoring tri-pox market for bout a week, last night I was lucky enough to do a search that must've just been a few minutes after a pretty sick one posted for way too low based on what I had been watching sell prior to its posting Nabbed it on buyout, and it was 1d11h. Surely if there were actually t3h_leet bot netz dominating the AH this would not have been possible for me, no? Bot would want to snipe at 1d12h ofc. So maybe there are or were some bots, but it can't be so bad given what happened with my tri-pox u said tripox? gona peek at ur gear righ now lol im trying to save up for a trihost, if i ever stop spending in stupid stuff edit: sick sick toon ur building man 5000 arm 47k life near 900 ar 187k dps wow. just missing that tricross man | ||
|
FallDownMarigold
United States3710 Posts
| ||
|
Cascade
Australia5405 Posts
On January 17 2013 08:14 Infernal_dream wrote: So once again, we don't have enough proof. So nobody should be arguing anything. As I stated RNG will be RNG. We all know the theoretical best is the MP which you absolutely annihilate everything. Whether that be 0 or 5 depending on dps and armor ratings and such. Which is pretty much exactly what I was saying. You can run it a billion times. The number of outliers is still going to be way too high to get any reasonable results. Unless we get blizzard to put the algorithms for their loot generators on the internet then we will honestly never know. But as of now nobody can say an MP level where you take longer to kill mobs is actually more efficient than one where you one shot everything. Don't want to get too entangled in this discussion, but... "...we don't have enough proof. [...] we all know that the theoretical best is [...]" I assume you mean that we don't have empirical proof, but that it is possible to make a convincing theoretical argument? What if increasing one MP takes you 5% more time, but gives you 10% more loot? Imo, it is tricky to theoretically argue in a "proof like" manner that the run-time will increase faster than the loot return for high MP, not least because how run time scales with MP depends a lot on your build. Sure, you can make estimates, and my guess (just looking at how health, damage and MF scales with MP) would also be that low MP is better, but I feel that "We all know the theoretical best is..." is a bit too strong statement. I think it is possible to make reasonable empirical tests for this, but you have to be a bit careful what you measure. If you try measure something like "gold per hour from drops" it is true that the data will be dominated by outliers (ie, most gold will come from that lucky 100M drop you get once every 50 runs). But if you at things like number of rares or number of legs (or both) it shouldn't fluctuate that incredibly much.The number of chances for a leg drop in an hour (depending on how fast you get your NVs, how much you die, and how many monsters/elites you happen to run into) shouldn't vary outside a factor 2, and after that it is just a counting exercise, so you will get a 10% error (from just the drop/no drop) after 100 legs. And 100 legs is so many hours of running, that the number of chances of drops shouldn't be a larger error. Anyway, as I said, I would also guess that lower MP is better, but maybe you are a bit to definite about how sure we are about it, and disregard the empirical data a bit too quickly. I haven't looked at their data though, so I shouldn't say too much, maybe their tests are bad. But I do not think it is impossible to test between a group of active players. Cheers. | ||
|
ximae
181 Posts
On January 17 2013 09:05 FallDownMarigold wrote: Ximae, I was watching them trying toget a good idea for how much I'd need to flip/farm/labor, not expecting to find one with good CC an just enough IAS for my BP priced remotely close to what I could nab. It caught me by surprise though! So underprice. I would recommend just camping that AH while you are on the comp for other things, browsing, movies, whatever. Good deals do happen on the AH - its not all insta-botted! i know, its how ive got most of my gear lol. I dont wield my cc and i dont really flip that much so its basically come to snipes since im not rich by any means. Prob is now i dont even have a minimum base of gold where i could snipe one as anything like that even underpriced is going for a few hundred mill | ||
|
.ImchEEzy
Canada123 Posts
http://us.battle.net/d3/en/profile/DeathRazR-1623/hero/30419233 about 20mil to use | ||
|
TheMooseHeed
United Kingdom535 Posts
On January 18 2013 05:33 .ImchEEzy wrote: hey, what items do I need to upgrade, http://us.battle.net/d3/en/profile/DeathRazR-1623/hero/30419233 about 20mil to use Replace your oculas ring with a zuni pox with average damage and vit. They go for super cheap now and you will end up with more damage and more int with the set bonus while maintaining your defensive stats. Depth diggers with vit also go for super cheap now. I am looking at one with more int then yours with vit for 2.5mil. On EU though but prices cant be that different | ||
|
b0yer8402
Canada17 Posts
next upgrade should be depth diggers? Should i go with tal rasha helm and armor? i really want to go up on mplvl but i barely do MP4. p.s : wanna stay with archon build for ever  | ||
| ||
|
|
Sparkling Tuna Cup
PiGosaur Monday
LiuLi Cup
Replay Cast
The PondCast
RSL Revival
Maru vs SHIN
MaNa vs MaxPax
OSC
MaNa vs SHIN
SKillous vs ShoWTimE
Bunny vs TBD
Cham vs TBD
RSL Revival
Reynor vs Astrea
Classic vs sOs
BSL Team Wars
Team Bonyth vs Team Dewalt
CranKy Ducklings
[ Show More ] |
|
|
EPT
