Last month we gave you a glimpse into how we were taking in initial feedback on the game, and ideas for some potential system changes we could make. Today we wanted to follow up and provide concrete info on some of the changes we’ve been working on, and give everyone a heads up before the 1.0.3 patch hits later this month.

Bridging the Item Gap

The ilvl (item level) of an item determines the statistical budget for its power. The way the game is currently set up, Act I drops ilvl 61 gear and below, Act II drops ilvl 62 gear and below, and Act III and IV drop ilvl 63 and below. Unfortunately this has caused two main issues. The first is players who find an Act too difficult feel compelled to use the auction house in order to progress. The second is that certain classes, skills, and play styles are less gear dependent than others, so although great items are making their way into the game economy, people feel pigeonholed into a handful of viable strategies. For a lot of people they would rather do something frustrating or boring in Hell Act IV (such as having Tyrael fight for them or breaking vases) for a chance at a "top-tier" upgrade, rather than fight hordes of monsters in Inferno Act I. We’re shifting to a philosophy where the best items in the game can drop from many different places, so a wider variety of play styles are viable.  If you would rather chain-pull elite packs in Act I rather than 3 minute cat-and-mouse in Act IV, we'd like you to be able to do that and know you can still find the best items in the game.

Nothing would explain it as well as just sharing the intended drop rates coming in the next patch, so here they are. Note that the drop rates vary slightly by item type; the table below represents an approximate aggregated rate of all item types:

New drop rates for 1.0.3

Item	Hell Act III/IV	Inferno Act I	Inferno Act II	Inferno Act III/IV
iLvl 61	9 %	18%	19%	24%
iLvl 62	2%	8%	12%	16%
iLvl 63	0%	2%	4%	8%

As you can see, players who would rather murder monsters 4x as fast in Inferno Act I can do so knowing they have a chance at amazing items, and players who want a challenge can kill in Acts III and IV in Inferno and be rewarded with a higher drop rate.

You Keep Using That Word

As previously mentioned, we’re going to be reviewing Legendary items in a future patch. Legendaries won’t change in 1.0.3, but it’s still something we’re actively working on. When we’re done, high level Legendaries should be flat out better than blue items, they’ll carry a good amount of power with them, and they should also be distinctive or memorable in the benefits they provide. We’ll be able to share more information on the specific changes we’re making after 1.0.3 launches.

The Nephalem Difference

It’s no secret that our goal for the end-game item hunt is players hunting monsters packs, building to five stacks of Nephalem Valor, and then killing a boss. While we’re seeing a lot of that occurring, what we’re missing is people feeling like it’s worthwhile to continue onward after killing a boss. To help hit that goal we’re lowering the number of guaranteed Rare items on bosses when you have your full five stacks of Nephalem Valor from two guaranteed Rares to one guaranteed Rare (you still have a very good chance at multiple rares, it's just no longer guaranteed). In exchange, all champion and rare packs will now drop a bonus guaranteed Rare item when you have your full five stacks of Nephalem Valor. The change benefits players with more overall drops, and a reason to push to continue progressing.

You Into the Group Thing?

We’re removing the bonus monster damage per additional player in a coop game. Our design goal is for players who prefer to play solo to be able to play solo, and players who prefer to play in groups to be able to play in a group. We feel the bonus monster damage per additional player is one of the biggest inhibitors to wanting to play with your friends. In a perfect world, single player and co-op would be absolutely equal, but that’s not attainable when you consider item properties such as “Life on Kill” or skills such as Archon which simply scale better when you are solo. Since the variety and breadth of game mechanics essentially dictate that solo vs. group play will never be 100% equal, our goal is to make them as close as possible but err on the side of coop in cases where we need to make adjustments. The inherent logistical requirements when forming up with other players and attempting to work together effectively warrants some added benefits.

Oh Yeah!

Inferno balance right now has a difficulty gap in which Act I feels about right, but Act II feels like trying to bust through a brick wall. In patch 1.0.3 we’re going to be lowering that wall by adjusting the damage and health of monsters in Inferno Act II, III and IV. We feel like Act I Inferno is in a pretty good place. Our design goal with Acts II, III and IV is to keep them challenging, but smooth the difficulty ramp out a bit. If a monk or barbarian is geared well enough that they can use a heavily offensive build and murder everything in Act I, they should be able to swap to a more defensive build and do okay in Act II. As they gear up they can begin adjusting back to becoming offensive in Act II, at which point they can jump into Act III with a focus on defense, and so on. Difficulty certainly ties into itemization, encounter and enemy tuning, and class balance, and all of these things together are going to paint a more reasonable difficulty curve as you hit Inferno in 1.0.3.

Paying for Your Mistakes>

Current repair costs at level 60 are barely noticeable, and because of that we see a lot of people wonder if “graveyard zerging” tough enemies or “chain rezzing allies on a boss” is intended gameplay – it definitely is not. To help solve the issue we evaluated a number of new death mechanics, such as just allowing the resurrection timer to increase even higher, disallowing resurrection during boss fights, or putting a debuff on you when you resurrect (such as reduced combat effectiveness). Ultimately we felt that increasing repair costs was the best solution that preserves the fast paced style of the game. Repair costs on level 60 items are going to go up a lot. Our goal is the next time a player is graveyard zerging a boss, it should occur to them that “this is probably not an efficient way to go about things”. We’re currently evaluating repair costs between 4x and 6x their current values. In the face of increasing costs, we recommend listening to the Hardcore players out there as they probably have some helpful advice on how to minimize repair costs. Following this change zerging a boss will still be possible, but our intent is that it won’t be optimal, and players who are seeking to be as efficient as possible will adjust their item hunting routes accordingly.

Whoa, Whoa, Nice Shootin’ Tex

We’re fixing a number of bugs with Attack Speed, mainly related to the stat not working on some items, but we’ve also decided we need to reduce the effectiveness of Increased Attack Speed overall. Many players have commented that Increased Attack Speed is such a dominant stat they feel it’s required. While we don’t have an issue with there being important stats, Increased Attack Speed in particular has secondary effects on mobility in combat, resource generation and resource consumption. We want there to be options and considerations for how you gear up, and one uber trump-everything stat can really work against choice and options. There are two different solutions we’re considering to reduce the effectiveness of Increased Attack Speed. The first is to simply reduce the value on all the items to their desired values. In general our desire is to never change items as that makes them feel less concrete, but the upside is you would still be able to look at an item and know exactly what you are getting. The other approach is to change the formula used for attack speed aggregation so that stacking attack speed from multiple slots suffers from diminishing returns. The downside of that approach is that it introduces yet another hidden modifier on an item property (and many people dislike hidden modifiers), and complicates the already difficult decision of item gearing. We’re currently leaning toward the first solution, to simply reduce the value on items, but we’d be interested to read people’s thoughts on the problem.

Just Three Two Easy Payments

We previously hinted that Blacksmith and Jeweler costs are coming down, and overall it will be far more reasonable to train them up and craft items. The most dramatic reduction is on the combine costs for tier 2-8 gems.

Quality	Previous Cost	New Cost
Flawed	3 Chipped + 500 gold	2 Chipped + 10 gold
Normal	3 Flawed + 750 gold	2 Flawed + 25 gold
Flawless	3 Normal + 1250 gold	2 Normal + 40 gold
Perfect	3 Flawless + 2000 gold + 1 Page	2 Flawless + 55 gold + 1 Page
Radiant	3 Perfect + 3500 gold + 2 Pages	2 Perfect + 70 gold + 2 Pages
Square	3 Radiant + 7500 gold + 1 Tome	2 Radiant + 85 gold + 1 Tome
Flawless Square	3 Square + 20,000 gold + 2 Tomes	2 square + 100 gold + 2 Tomes

The gem combine costs for Perfect Square and above will remain unchanged.

Nerf Them, Buff Me

Class tuning is not a major focus for 1.0.3. There will be a small number of skills changes, but for the most part we want people to continue experimenting and enjoy their skills for a while. Our goal was and continues to be build diversity, and though we see quite a bit of build diversity, we think we can do much better. Class tuning will be an ongoing process, and we’re targeting the 1.1 patch for most class tweaks, with a focus on punching build diversity up a few more notches.

But What About…

While these are a few of the larger systems adjustments we’re making, the 1.0.3 patch will include many fixes, quality of life enhancements, Auction House improvements, and other changes. We hope you look forward to the patch as much as we do getting it out there, and again we appreciate your continued feedback. See you in-game!

Wyatt Cheng is a Senior Technical Game Designer for Diablo III. He kindly requests that if you see him in Development Hell you spare his life and move on.