Welcome to Crestfall Gaming

Register now to Crestfall Gaming. Once registered and logged in, you will be able to contribute to this site by submitting your own content or replying to existing content. You'll be able to customize your profile, receive reputation points as a reward for submitting content, while also communicating with other members via your own private inbox, plus much more! This message will be removed once you have signed in.

Sign in to follow this  
Followers 0
Darkrasp

Darkrasp's Update 12/5/2016

30 posts in this topic

Awww, get well soon, Rodeg! Get lots of rest and take it easy. 

And thanks again, as always, Darkrasp, for the update. The attached video was very nice to see as someone who gets frustrated when I'm in the middle of combat when those stupid ass disconnects happen and yet I'm penalized for it by sudden death when I wasn't even there to try and fix the situation -.- 

B-b-b-but...no more weekly updates? How will we live without you? *sniffle* Ah well, I suppose you'll have to save your next update for AFTER the first ten beta testers are selected. Y'know, to quell the masses ;) 

0

Share this post


Link to post
Share on other sites

Best wishes to Rodeg, thanks for the update and see you in two weeks.

0

Share this post


Link to post
Share on other sites

nice update! especially the formular thing with items! you einsteins :D

0

Share this post


Link to post
Share on other sites

Get well soon Rodeg! Also, if anyone got confused for a second I believe he meant switching to bi-monthly updates, not bi-weekly

0

Share this post


Link to post
Share on other sites

Lovely information, I just wish time to pass faster so one day soon to see the server live :)

0

Share this post


Link to post
Share on other sites

The amount of detail you go into blows my mind. And people are hyped about stock servers...

2

Share this post


Link to post
Share on other sites

the disconnect bug is present on every mangos server i've encountered so far, great to see it fixed yere :D

really looking forward to beta this at some point!

 

thanks for the update B|

0

Share this post


Link to post
Share on other sites
12 hours ago, Darkrasp said:

 You may have seen him post a small video clip in Discord (WARNING:  MUSIC NSFW, this is just what he happened to be listening to at the time) on session code that implements proper handling of Alt-F4/disconnects, where the character stays in the realm for a period of time before disconnecting, and remains alive and in combat (if they were in combat when they disconnected). 

 

From that video I do not see he has identified or fixed issue with disconnects mid fight when you just log on your account while you are in game, which removes your char from game world instantly. On retail you can't log in same account that's already logged, while on pservers I played same thing start new play session, while closing old one and so removing you instantly from game world.

0

Share this post


Link to post
Share on other sites

I've disconnected and relogged mid-fight in actual classic many times (Onyxia was really bad for me that way, I've "tanked" P3 for minutes while d/c - was playing on dialup at the time)

0

Share this post


Link to post
Share on other sites

To clarify, I meant open another game client session and try to log on same account you are already logged in and in game. Simple alt+F4 wouldn't remove you from game world for like 20 seconds or until relog on curent pservers either, so i do not see point of that video.

Makes me ask, what happens if you d/c, relog while char is still in game world, but try to enter game on another char from same account?

0

Share this post


Link to post
Share on other sites
1 hour ago, wingpuller said:

Makes me ask, what happens if you d/c, relog while char is still in game world, but try to enter game on another char from same account?

I'm not 100% sure, but the way I remember it, there was a period of time where you'd get a bug saying "A character with that name already exists" when you tried to log on after a d/c. And (correct me if I'm wrong here people) you were able to log on an alt, and even see your main character logged on in the guild list etc.

3

Share this post


Link to post
Share on other sites

First off: Great job! Really.

The "cracking" of the formula is amazing. I still don´t understand the logical mind behind such discoveries lol.

 

However, please don´t take this as an attack or anything, but if I am right, the mobs shouldn´t stand inside each other but gather around you in a (half)circle? Just to make sure! ;) 

 

Greetings, Buddhaseyes

0

Share this post


Link to post
Share on other sites
1 minute ago, buddhaseyes said:

 if I am right, the mobs shouldn´t stand inside each other but gather around you in a (half)circle? Just to make sure! ;)

You are right, we disabled mob fanning for now because the algorithm for it needs some tweaking.  Low priority at the moment but we'll get it working soon.

0

Share this post


Link to post
Share on other sites

Wow! Thanks for the (super) fast response!
Very good, very good indeed!
 

Greetings, Buddhaseyes

0

Share this post


Link to post
Share on other sites
On 6.12.2016 at 5:19 AM, Darkrasp said:

I managed to do some mathemagic and I'm pretty sure I cracked the code on Blizzard's itemization budgeting for items.  It seems that an item of a given iLevel had a certain amount of what we'll call "itemization points" that they could spend on stats, attack power, spell damage, dps, etc.  Items would have a base dps value, and a base point value, specific to their type and level, and then the designers would be able to spend those points to add stats/procs/etc.

Isn't the amount of different stats on an item a factor as well? It seems for some items that the more different stats it has the more itemization points it would get in total. I guess they did that in order to make gearing for hybrids easier as they usually benefit from more stats than other classes. Take for instance Avenger's Breastplate and Conqueror's Breastplate, the T2.5 chest pieces for paladins and warriors.

 

Avenger's Breastplate
Item Level 88
985 Armor

+12 Agility
+23 Strength
+24 Intellect
+11 Spirit
+24 Stamina

+18 spelldamage
+1% spellcrit
+1% meleecrit

 

Conqueror's Breastplate
Item Level 88
985 Armor

+24 Agility
+34 Strength
+38 Stamina

+6 Defense

 

So if you compare only stats Avenger's armor has 94 and Conqueror's Breastplate has 96. Thats almost the same amount but with armor being the same you basically put +6 defense on par with +18 spelldamage, 1% spellcrit and 1% meleecrit which does not seem right if you just take static itemization points for a given item level. You can see this on many items that have a large amount of different stats on them. How is this explainable?

Another example that is easier to compare is Penelope's Rose and Tome of Knowledge.

 

Penelope's Rose
Item Level 61
+11 Intellect
+11 Spirit
+10 Stamina

 

Tome of Knowledge
Item Level 61
+8 Agility
+8 Strength
+8 Intellect
+8 Spirit
+8 Stamina

 

So Peneope's Rose has 32 stat points and Tome of Knowledge has 40. Thats a 20% difference even though their item type, rarity and level is the same.

 

Edited by fruitsalad
0

Share this post


Link to post
Share on other sites
45 minutes ago, fruitsalad said:

Isn't the amount of different stats on an item a factor as well? It seems for some items that the more different stats it has the more itemization points it would get in total. I guess they did that in order to make gearing for hybrids easier as they usually benefit from more stats than other classes. Take for instance Avenger's Breastplate and Conqueror's Breastplate, the T2.5 chest pieces for paladins and warriors.

<snip>

So if you compare only stats Avenger's armor has 94 and Conqueror's Breastplate has 96. Thats almost the same amount but with armor being the same you basically put +6 defense on par with +18 spelldamage, 1% spellcrit and 1% meleecrit which does not seem right if you just take static itemization points for a given item level. You can see this on many items that have a large amount of different stats on them. How is this explainable?

 

Stat budget has never had an exact correlation with player power for Blizzard and even through all expansions and how they've been gradually simplifying everything to being a function of internal itemLevel (to the point of abolishing quality modifiers) it still isn't.

Or you wouldn't have patch changes like these to this day (patch 7.1.5 notes):

Quote
  • The first step of these changes was to reduce the amount of secondary stat percentage you get per rating. The second step was to increase the budget of items across the board by 5% or so.
  • If you have 47% Crit right now, you are probably only going to lose a few percent after this change. This will be offset by the extra primary stats you get. Your overall power shouldn't be meaningfully affected.
  • Overall this change was aimed at shifting the balance of how much of an item's power comes from a secondary stat to help make item level a more consistent guide for what you should equip.
  • For the most part a 15 item level upgrade should be something that you can blindly equip and trust that it is the right thing to do. Most 10 item level upgrades should also be the same way. Once you get below that it may come down to the stats or sockets.

So item level > stat budget > player power has always been a moving target and more so the further back we go into WoW's development.

That's not to say that they randomly assigned budgets to items. There's specific stats that directly map to internal item level and have been since "forever". Base Armor Class is one such, Durability is another.

But even there before Bonus Armor began being denoted as an extra stat, some items will not follow the obvious pattern.

What I'm saying is.

  • Blizzard does have an internal valuation of stats, both primaries (like sta, agi, str, spi, int, def, hit, ac etc) and secondaries (crit, ap, sp, dodge, parry and so on) but we don't know those numbers.
  • Even if we knew those numbers, they don't always correlate 1:1 to player power, that's a moving target as evidenced by both a myriad balancing and re-balancing patches and a ton of exceptions for specific classes/specs.
  • There is no way to know how the stat budget is distributed in the case of set pieces, how does the whole set stat budget affect individual items in the set.
  • We don't know if each stat internal value in the item budget is a function of class (we do know that base stats have different effects by class, for example hp regen from spirit)

The best you can do is compare like for like (slot, armor class, item quality, player class, player role). Trying to extrapolate missing stats based on items with a large disparity in those attributes will not end well.

The items you picked have too many variables to consider, they're not a good example.

Conqueror's Battlegear

  • Armor: 3783 Strength: 137 Agility: 96 Stamina: 140 Defense: 26 Hit %: 2%
  • 3: Decreases the rage cost of all Warrior shouts by 35%.
  • 5: Increases the Slow effect and damage of Thunderclap by 50%.

Avenger's Battlegear

  • Armor: 3783 Strength: 100 Agility: 63 Stamina: 101 Intellect: 95 Spirit: 44 +3% Crit +1% Spell crit +85 Healing +85 Spell damage +11 Mana / 5 sec
  • 3: Increases the duration of your Judgements by 20%.
  • 5: Increases damage and healing done by magical spells and effects by up to 71.
0

Share this post


Link to post
Share on other sites
Quote

 

Another example that is easier to compare is Penelope's Rose and Tome of Knowledge.

Penelope's Rose
Item Level 61
+11 Intellect
+11 Spirit
+10 Stamina

 

Tome of Knowledge
Item Level 61
+8 Agility
+8 Strength
+8 Intellect
+8 Spirit
+8 Stamina

So Peneope's Rose has 32 stat points and Tome of Knowledge has 40. Thats a 20% difference even though their item type, rarity and level is the same.

 

 

20 minutes ago, Roadblock said:

<snip>

 

I edited my post and added two items that maybe are easier to compare. I just noticed that items with a lot of stats potentially had more "numbers" in total on them meaning that those itemization points mentioned where not static (or different stats were weighted differently) and only dependent on item level alone. I agree that comparing two set items wasn't the best example as most set items are class specific. But you can see that difference on items that don't have that restriction as well. And some ar even usable by all classes (without any implicit restrictions like armor class or weapon type). Darkrasp said he had figured Blizzard's itemization budgeting out, so I was curious how exactly it comes to that difference.

Edited by fruitsalad
0

Share this post


Link to post
Share on other sites

Are you assuming that all base stats have the same valuation? 1 point of stamina = 1 point of agility = 1 point of strength and so on?

I don't think that's the case. The easiest way to know this is not the case is by comparing the various green items with "of the x" suffixes. Those have the same armor class, the same item level, no set bonuses or equip effects. If anything is liable to be derived from a stat valuation function that's going to be the Bear / Boar / Eagle / Falcon etc items. A quick glance shows that stats don't match up 1:1 (especially comparing between the of 1 stat vs of combo stat variants).

 

I think Darkrasp was referring to a relatively small set of items with a clear purpose (weapon damage).

You have to keep in mind that the whole point of this exercise is to pare down a list of a couple hundred items as much as possible. After all the research effort by Darkrasp and the community that participated, items remain that have no historical data available. No one bothered in that era to do proper testing or run a proc counter and report findings.

So you either have to start flipping coins and assigning arbitrary numbers like most projects seem to be doing (from the few items that we managed to push the envelope and actually unearth historical data that was not previously known, see Ironfoe research on this forum for a prominent example) or find a pattern.

Extrapolating proc rates from dps for the "holes" in the set where we have similar items with good known values is definitely better than just throwing darts at a board with 10%, 20%, 50% and other random percentages. More so when it tracks ok for the items that we did manage to find good values (from classic era testing) for.

 

Edited by Roadblock
0

Share this post


Link to post
Share on other sites
8 hours ago, buddhaseyes said:

First off: Great job! Really.

The "cracking" of the formula is amazing. I still don´t understand the logical mind behind such discoveries lol.

 

However, please don´t take this as an attack or anything, but if I am right, the mobs shouldn´t stand inside each other but gather around you in a (half)circle? Just to make sure! ;) 

 

Greetings, Buddhaseyes

this system is important! atleast to me. its frustrating not seeing other pservers consider implementing this like it was in retail.

the fanning system made tanking a nightmare especially Grobbulus in Naxx because he kept moving all over the damn place and repositioning himself.

same deal with AOE packs surrounding the tank.

Edited by imbaslap
0

Share this post


Link to post
Share on other sites

@fruitsalad There is an *almost* exact formulation for the point value of stats.  I don't want to give away exact details on how it works, but suffice to say that the number of stats (talking specifically about STR, INT, SPI, etc.) on an item is relevant.  There is a multiplier to value based on the number of unique stats.  Single stat items are weighted more heavily than two-stat items, etc., however it does appear that all stats are weighted equally against each other.  One point of strength does have the same value as one point of intellect, they're just useful to different classes.  As far as I can tell, that's intentional.

There is also an *almost* exact formulation for the point value of 1% crit, 1 AP, 1 SP, etc.  1% crit has a much higher point value than 1 stat, so it follows that it won't start showing up on items until they are high enough level that they have the stat budget points available to afford it.

Item quality plays a role in how many points are available, but not the value of those points in terms of "purchasing power" of stats etc.

There are certainly some other modifiers beyond the relatively simple formula that I was able to derive (I suspect based on item type).  I had a few small anomalies, but these seemed to have a very slight overall effect, and I couldn't find any easy way of figuring out the direct cause.  Occasionally a weapon existed that was just better than it's contemporaries for no discernible reason other than perhaps the difficulty of acquiring it.

Things get hazy when you have to look at unique effects like that set bonus, "Decreases rage cost of all Warrior Shouts by 35%".. something like that is very difficult to estimate a stat- or dps-equivalent value for.  If it was on a single item, you could (and I did for some procs) figure out what Blizzard thought it was worth by reversing the formula and comparing it to a same-level-and-type item.  However many points that proc takes the place of can be directly converted into "expected dps".  However, being that it is a set bonus and not tied to an item, it gets to be nearly impossible to do for lack of adequate comparisons.

Edited by Darkrasp
5

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0