1.07 News, Info and Gossip

A5 barb is better, and no shift+number doesn't work in 1.07.

You can avoid having a merc unless you plan to play long-term, in which case I would implement one eventually. They're tough to keep alive but with a sorc it's not nearly as hard as with other characters... not until they're well-geared, though.
 
maxicek said:
I didn’t bother with a merc at all.
Click to expand...
It would seem difficult to hit some of the higher breakpoints without it though? Though I can see that going for 556 % is probably not that useful, SoJ being the only unique worth mfing for I guess.

Still, going for 374 % requires a merc as far as I can tell. But for rack running and general play they do seem somewhat useless.
 
Oh you def want a merc for MFing but even then you need to gear carefully. Rackable items should be a much much higher priority than SoJ or other harder-to-find uniques, imo
 
The barb seems to have hit some kind of breakpoint at lvl 55 and started healing a lot more, but he's still sooo useless! :/ I managed to clear nm and hell and I think I'll go for lvl 67 for now, so I can equip tal's amu (nm andy yay).

I'm not sure what to do afterwards, since this was supposed to be a test run for a hc sorc. Everything seems to oneshot. You do act bosses once with max lvl static from far away, else you get oneshot. Teleporting is dangerous, because it's so easy to get stunlocked by some random mob pack and insta die. Just getting hit by an exploding a5 ice boar can kill you. Seems like cows are the way to go, safest place with pretty "good" xp from what I can tell.

I'm not really getting anywhere with perfect topazes. I think I have 4 or 5 after clearing normal, nm and hell and I've used a flawless topaz anytime I've found a gem shrine. And I need like 14 of those for me and my merc. Speaking of which, since you get 4 sockets in magic armor from larzuk q, am I correct to assume that you want to find or shop a good base with % def and + hp, socket it and fill it up with topazes?
 
yes you could do that with larzuk.

Also, you could run with both a barb and sorc next so that you have BO before NM. You'll probably want it!
 
@Shuffleblast, try to get gear with high resists, MDR and PDR to slightly delay the inevitable of the merc death. In NM they will tank ok but will need the occasional reposition with teleport if playing a sorc. If your char level is low then low level mercs are too weak to tank anything in Hell and will die to MSLE, conviction/LE, might/ES or fanat/ES, hydras. But try the build out on Hell Kaa and he can last just long enough to kill with anything expect conviction/LE or MSLE long enough then he is good enough
 
pharphis said:
yes you could do that with larzuk.

Also, you could run with both a barb and sorc next so that you have BO before NM. You'll probably want it!
Click to expand...
You mean playing a multiplayer game with myself playing 2 chars? This seems very popular on bnet, but I've never done it before. Maybe I thought of it as cheesing, I'm not sure. But isn't 1.07 BO still really short duration? That was quite annoying.

Though I could maybe do it, I need to think about it. Just the other day I've hex-edited the d2gfx.dll file to be able to play multiple 1.07 instances on one computer (I couldn't find any online).

This brings me up to another issue: are there any no-cd methods for 1.07? I couldn't find any cracks and even installing from and mounting an .iso image file still requires a physical cd in my cd drive. Most annoying.

Sorc is lvl 71 now after spamming Shenk ad nauseam. I've ditched my mf gear and equipped viperskin (nm baal) for faster static, found a +80 mana ring and... FROSTBURN (nm council)! Yay! 650 mana with max warmth, the game is finally playable. :)
 
Short duration is still better than nothing! You'll mostly just want it for teleporting/bosses anyway.

I used bliss version switcher and singling files that all for multi-instances and no-cd. I started with vanilla 1.07 installation but it should work from any version installation iirc. Details in the BVS thread
 
Has anyone ever looked into whether a rack drop's quality is influenced by the same nonrandomness that produces stable racks to begin with?
 
Do you mean whether it is set, unique, rare or magic?

I doubt it is because we would have found stable set or rare racks by now (maybe even unique). One of the coders can probably answer that directly, though.
 
Well, we have plenty of anecdotal accounts of stable racks that drop uniques at a rate that looks more like 1/3000 than the supposed 1/1000, and we also have seen racks that drop them at a higher rate; there's just not much incentive in running a rack for longer than 2-3 unique drops so the sample size is smaller on the beneficial side of the spectrum.
 
Well, to question the RNG we have to first understand how it works. Many computers do not actually have hardware to generate random numbers. Instead, they are pseudo random generated through either simple or complex algorithms (such as that of rand() used in the C library). In order to keep it from generating the same numbers over and over, time is often used as the rand seed.
srand(time(NULL));
Click to expand...
Another common way to generate random numbers is by measuring CPU cycles/ticks. I.e. GetTickCount. Because the processor is regulated frequently, this will almost always return a different number. Diablo II happens to use both methods for its' RNG.

First when the game is opened, the random seed is set based on the current time. Whenever you actually create a game, a "game seed" is generated for that session. That helps keep things even more random within the already random map seed. The next part is going to depend on the patch. Prior to 1.09, it appears only the map seed is used in determining monster spawn points. As we've known for awhile, racks are only influenced by rack location and sprite spawn points. Every time a monster spawns in its' "active" radius, the rack gets reseeded or "refreshed" by the code responsible for updating the sprite table. Going back even further, prior to 1.07 it appears loot dropping was also only affected by map seed. This allowed for legitimate "dupe" items a.k.a monsters can sometimes drop identical rares if killed the same way (time+location).

Now in the current patch, I've noticed more calls in various loops to return the processor speed. It seems whenever you make a game, the game seed, map seed, and processor speed are all factored. The latter is the most random of the three, and is likely responsible for the unpredictable nature of monster spawn points. This amounts to changing racks more often, of course some racks will be semi-stable on occasion due to monsters being disallowed to spawn on certain tiles.

With that being said, I've had some maps where it takes 10 runs to see a gold, others thousands with nothing to show for. In the process of searching for racks, I often get one or two random unique items--usually useless although I did rack a random Arreats' before. That alone would convince me that rack quality is much more random than the rack itself. But it's no doubt that time is likely the biggest factor here. If you could "freeze" your computer's CMOS and the rack dropped more consistent, that would give us a big clue.

TLDR -> Quantum computers for the win!! They react to noise, perceiving decibels, frequency, bit rate, and patterns as machine processor instructions! They would use white noise to generate random numbers, which is basically impossible to predict and can be intercepted in practically an infinite number of ways. Imagine your girlfriend walks in to yell at you for spending too much time on D2. Then in that moment, her voice triggers an RNG change which happens to leave a Zod lying in front of you.

Sorry, but I'm busy playing video games hahaha!
 
Grape said:
Agreed!

Maybe about an hour worth of runs, and with my multi-tasking runs (super chests, racks, shrines and occasional boss packs), that's probably around 60 runs. And the next one please!
Click to expand...
How long did it take you to find such a rack? I've only done like 1-2 hours of test running in softcore and it seemed like getting a stable rack with a good base item is the harder part than actually rolling a unique.
 
Log in to Reply / Create Site Account
Remove Ads - Go PurePremium

Latest posts

Back
Top