Development Manifesto: Harvest Crafting

"

We're going to be making some further changes to Harvest Crafting with the release of 3.14.0 in April. In short, we're concerned by how deterministic some Harvest Crafts are and how easily players can craft near-perfect items. We don't feel that this is consistent with how Path of Exile's itemisation should work. Before we detail what changes are coming, we want to explain a bit about Harvest's initial implementation and our design philosophy around items and crafting.

Our Initial Plan for Harvest Crafting
Using currency items like Exalted Orbs or Chaos Orbs on your gear is a powerful, risky and exciting way to improve it. But because these items work equally well on both low- and high-level equipment, it's generally considered best to save them until you're crafting your end-game gear. Many players would even argue that it's best to save them to trade for already-crafted items. In either case, using valuable currency items on your levelling gear is generally perceived to be non-optimal.

When we were designing the crafting mechanics that would be the reward system for Harvest League, we wanted to experiment with a system that allowed you to craft items while levelling (using the equivalent of powerful currency like Exalted Orbs), so that players could get the experience of using valuable currency items without feeling bad that they were missing out on future value. Our goal was to make a system that was so attractive that players would try to improve their items by throwing a few crafts at them in almost every area that they play.

Towards the end of Harvest's development, it became clear that this was a very powerful crafting system. We had hit our goals and then some. The question was whether we should preemptively nerf it before players saw it, or try it out for a league at full strength. We decided to try it out. The data we'd get from players interacting with a very powerful crafting mechanic would be very helpful with our ongoing work deciding the appropriate power level of crafting opportunities in Path of Exile 2.

So we launched with the powerful version of the system, including some crafts that modified items in more specific, deterministic ways than we had allowed before.

Upon playing the league, the crafting side of it was an immediate hit. Players easily filled gaps in their builds while levelling, and made many, many mirror-worthy items. As our concern over this grew, discussion started about how to adjust things when we eventually integrate it into the core game.

Integrating Harvest Crafting into the Core Game (in 3.13.0)
We decided to integrate Harvest into the core game as map-only content. Harvest Crafting while levelling went fine during the Harvest league, but pretty much every league has some item acquisition mechanic that occurs while levelling so we didn't think it was necessary to have Harvest on top.

In the end-game, Harvest Crafting allowed the creation of some ridiculous items, mostly via a set of deterministic crafts that interacted with specific types of mods. This had to be toned down, and we had to choose between two ways to do it: either by removing the dangerous crafts, or by keeping them in as rare possibilities and then balancing by rarity.

We decided to do the latter. The good crafts would stay in the game as very rare outcomes. We didn't remove any crafting options, other than ones that related to the process of growing a garden as they became redundant with the garden-maintenance mechanic being removed.

So in 3.13.0, we released an integrated version of Harvest where players stumble across groves full of crafts and hope they get the most valuable deterministic ones. We hoped that this would still keep the most valuable crafting options available while limiting the most abusive crafting to just very lucky or successful players.

Analysis of Integration into the Core Game
With Harvest integrated into the core game with 3.13.0, players still made ridiculous items. Even the crafts we made quite rare felt pretty common when the entire community was pooling them together and using them on the right items.

The second issue was that players felt overwhelmed to get so many crafting options. It was possible for up to 100 or so crafts to be given to the player per grove, and this caused people to feel obliged to consume them otherwise they would be wasted.

The result of both of these factors was that players had access to basically unlimited medium-power crafting and had to RNG-hunt the best crafts which they could then store for later.

Players also expressed frustration that the most effective way to get the best items in Path of Exile was to join a discord channel and try to trade for these incredibly crucial crafts.

The first part of this that concerned us was that Harvest was critical in making the best items (and hence made many other game systems obsolete). The second part, where it was inconvenient to trade them, becomes far less of a problem if we can solve the first issue. Our problem with all of this can be summarised with the following thought: "Why would I use a regular Exalted/Divine/Annul Orb when I can get one through Harvest that has a deterministic result?"

The entire rest of Path of Exile's crafting system is somewhat redundant with Harvest Crafting in its current form. While we are glad we tried the experiment of keeping all the crafts and balancing by rarity, it's unfortunately going to have to change.

Path of Exile's Item Philosophy
At its core, Path of Exile is a game about the acquisition of powerful items.

When we were designing Path of Exile, a critical aspect of item acquisition is that it is through random (rather than deterministic) means. When you defeat monsters and bosses, you receive random items. When you approach Cadiro for a valuable item, you are offered a random one rather than one that you can control. When you craft an item, you receive a random modifier. In limited cases where you can choose a specific modifier, it's usually worse than what could have been rolled randomly. Even game systems specifically intended to involve determinism such as Divination Cards and Incubators still have amounts of luck involved.

It's also important that items are hard to perfect. Ideally there's significant diminishing returns in the currency item crafting process, which lets most players get something good enough relatively easily, and the expert players can show off with really good items that took a lot of effort to make. Obtaining perfect items is ideally close to impossible, with very few players able to claim that they have such valuable treasures.

We feel that the current state of Harvest Crafting runs against both of these important philosophies. We know that many players would love us to keep deterministic crafting in the game because it enables them to complete their items far more quickly than they otherwise would. But then there would quickly be nothing left to achieve. It was an interesting experiment, and we understand that some players will likely be attached to this level of incredibly easy crafting, but it's just not the Path of Exile we set out to make.

This sentiment was summed up by a member of our design team who recently said "We don't want to take away the feeling of closing your eyes and Exalting an item, scared to see whether you ruined it or not."

Upcoming Changes in Path of Exile 3.14.0
So in 3.14.0, we're making some changes. The first three are direct nerfs to Harvest, and the second two are improvements. It's also worth noting that Harvest Atlas passives have been adjusted to accommodate these changes.

Previously, every seed in a patch granted an instance of that seed's craft. Now, only some of the seeds do (so you're getting far fewer of the crafts that were overwhelming people with their quantity). Higher-tier seeds are closer to the 1:1 ratio from before.
Some mods that had overly-deterministic behaviour have been removed. These include all annulment mods (other than the ones that remove a mod that isn't of a specific type before adding one of that type), and all type-specific divine mods.
Crafts that add mods of specific types (like Physical Modifiers, for example) to items can now only be applied to non-influenced items, except for the existing mod that applies an influenced mod to an influenced item.
The chance of encountering a portal to the Sacred Grove in a map has been increased by 60%.
The Heart of the Grove encounter is now a map fragment that sometimes drops from Tier 4 Harvest bosses, instead of randomly appearing in place of a normal Harvest grove. This allows you to trade the encounter if you don't feel up to it, and it means that finding The Heart of the Grove when you are in a map with difficult mods doesn't lead to an impossible encounter.

Overall, this is undoubtedly a heavy nerf to high-end Harvest Crafting, but we strongly believe that it is in Path of Exile's best interests going forward, and that there are still a lot of compelling Harvest crafts that make the grove worth running at any point in map progression.
Posted by Chris
on Mar 10, 2021, 1:37:34 PMGrinding Gear Games

"

Used to purchase entry level mapping gear (trade leagues), re-roll bad map mod combos before corrupting, spend on Zana maps, atlas/league mods and for the "Chaos Spam" method of crafting

"

Used to purchase mid-late game mapping gear. Used to create deterministic mods on weapons (looking at caster weapons here). Used to craft meta-mods on items to add more deterministic modifiers to items.

"

We decided to integrate Harvest into the core game as map-only content.

"

Our goal was to make a system that was so attractive that players would try to improve their items by throwing a few crafts at them in almost every area that they play.

"

This sentiment was summed up by a member of our design team who recently said "We don't want to take away the feeling of closing your eyes and Exalting an item, scared to see whether you ruined it or not."

"

Previously, every seed in a patch granted an instance of that seed's craft. Now, only some of the seeds do (so you're getting far fewer of the crafts that were overwhelming people with their quantity). Higher-tier seeds are closer to the 1:1 ratio from before.