What Happened with Items

Lake of Kalandra saw a number of balance changes that were not properly communicated before release. After a week of addressing feedback with hotfixes, we have written this post to explain what our intention was, what went wrong, how we have fixed it, and to reassure you about the direction we intend to go in the future.

There's a bit of backstory to explain. I want to start by describing three philosophies that have been guiding our decisions recently:

Philosophy One: Reward mechanisms should scale properly with Item Quantity and Rarity bonuses

For the last few years, we have been using what we internally call item templates to control what drops from league content. This is where a monster (often with a reward symbol over its head) drops a specific type of item when it is killed.

But Path of Exile is a game about opting-in to more difficulty in exchange for more rewards. You can roll your maps to be harder or add sextants to them. You can play with additional party members. You can trigger additional stacking league content like Delirium. All of these things make the game harder in exchange for more and better rewards. The way we achieve more and better is through item quantity and item rarity bonuses. Item quantity means you directly find more stuff, and item rarity means that it has a higher chance of upgrading to magic, rare or unique. Item templates ignored quantity and rarity bonuses. A template of "drop four rare jewels" just did exactly that, regardless of how much extra difficulty you had stacked.

From the 3.19 expansion onwards, we are trying to make sure that reward systems scale with player item quantity and rarity bonuses. That's why the reward conversion system that higher-tier Archnemesis monsters have is so powerful. Any bonuses you have from additional difficulty will affect the rewards that the rare monster drops. Additional item quantity causes them to drop more items that are converted, and additional item rarity causes those items to upgrade, which also affects the converted one. For example if you upgrade a rare item to a unique item and it's then converted to a currency item, it'll drop as a Divine Orb, Exalted Orb or Orb of Annulment.

Going forward, we are trying to make sure that as much as possible, reward systems scale with the reward bonuses you get for playing difficult content.

Philosophy Two: Players should fight fewer Rare Monsters at once, but they should be more challenging and rewarding

In fights with a lot of Rare monsters on screen, you can't follow what modifiers they have, what skills they're using, and sometimes not even what type of monster they are. There's too much to pay attention to, with too much noise and screen pollution. You cannot use appropriate combat tactics, and instead have to just stutter step or be so powerful that it's inconsequential.

Fewer, more difficult rare monsters help you pay attention to what is happening, assess it, and act accordingly. It gives you an opportunity to employ counterplay and for your playskill to actually matter (rather than relying on pure character power). It is also a lot cleaner and far better for performance.

Rewards should be set appropriately for the increased difficulty of these rare monsters.

Philosophy Three: There shouldn't be a large gap between the difficulty and rewards of league content and base game content

Monsters added in leagues are more difficult to kill and drop better items than regular ones encountered in the base game. When those leagues become core, these properties carry across, creating two tiers of content, with one far more rewarding than the other.

We feel it's good for league content to be harder than the base game, and therefore more rewarding. But the difference should be approximately twice as rewarding. If the gap were any larger, then it would be less efficient to kill regular monsters and a player should spend all of their time focusing on repeating a small subset of content.

With those philosophies established, let's have a look at some changes we made in 3.19, and then examine what went wrong and what we're doing to address it in the future.

Lake of Kalandra Balance Change: Rare Monster Normalisation

A lot of league content was spawning way too many rare monsters compared to the rest of the game. In line with Philosophy Two, and general player concerns about being overwhelmed by too many hard Archnemesis monsters in some encounters, we reviewed most league content in Path of Exile with a goal of making the rate of encountering rare monsters consistent.

There are three changes that needed to happen at the same time as this:

1. The addition of interesting rewards to some Archnemesis Mods that scale with both Item Rarity/Quantity (Philosophy One) and yield very valuable outcomes if combined in the right combinations to create moments of excitement as valuable rewards drop.
2. An adjustment to the average number of Archnemesis Modifiers on rare monsters to increase difficulty, justify the higher rewards and create more random interesting encounters that add variance to gameplay.
3. A rebalance of Archnemesis Modifiers to account for the fact that rare monsters now have multiple modifiers more frequently. This step was not performed until after release feedback came in. It was not deemed necessary at the time, and required extensive community feedback before we did it. This was a mistake and we should not have been so stubborn about it.

Lake of Kalandra Balance Change: Monster Item Rarity and Quantity Normalisation

As described above, various valuable Archnemesis modifiers convert drops in a way that directly benefits from item rarity and item quantity bonuses. When we were balancing and testing this, we wondered why certain league monsters were dropping significantly more items than regular monsters. It turned out that this was due to item rarity or quantity bonuses that were historically applied to monsters to make leagues feel rewarding. When combined with the new drop conversion system, these bonuses stacked exponentially and caused far too many rewards.

In line with Philosophy Three, we rebalanced league monsters so that they were twice as rewarding as regular monsters and didn't have these existing bonuses. To be clear, the bonuses were inconsistent and arbitrary. For example, Yellow beasts dropped more items than Red beasts. Incursion monsters didn't have any Increased Quantity, just increased Rarity, but Harvest monsters had both. This change was not mentioned in the patch notes.

Now we get to Beyond. This was beyond broken for map juicing, sometimes spawning over 200 unique monsters in a map. The amount of items that came from Beyond was just ridiculous. It is not okay for fifteen thousand unique items to drop in the same map. The new version is more reasonable (allowing up to one unique beyond boss per map), which is honestly a gigantic nerf. But it was intentional, and we mentioned in the livestream it was reworked, with more details in the patch notes. While we took away the extreme juice opportunity, we added a dedicated reward for Beyond: Tainted Currency Items.

What went wrong

We didn't patch note the item rarity/quantity rebalance for league monsters. This was an oversight due to human error, but that's why I proofread the patch notes. Unfortunately, due to the next point, this wasn't caught during my proofread.

I… didn't actually understand the impact of the change. It was mentioned to me in passing (that we were removing the league monster bonuses and replacing with just quantity), and I didn't ask any more questions. I was busy, distracted, and should have sought more information. Had I understood the consequences, we likely would have still gone ahead with the change, but hopefully with better communication and maybe some pre- rather than post-release counterbalance elsewhere. This is a massive internal communication fuckup and I take full responsibility for it.

There was not sufficient time to playtest the change properly for feeling. It is unacceptable that I allowed a change like that to make it into the patch without a big chunk of time allocated to making sure the game still feels great afterwards.

I also overstated the impact of the change when communicating about it in this post. I said "we removed a massive historic bonus", and this caused the community to think the impact was larger than it was. The reason why I used the word "massive" was that the numbers sound big when viewed in isolation, but are less impactful when viewed in context. For example, the rarity bonus that was removed from a Red Beast was 750%. This sounds big, but a four-mod Archnemesis rare has a 41000% bonus. Players have been saying we reduced drops by 90% but in reality, that type of difference could only be true in the most extreme situations involving Beyond, Delirium and Incursion stacked with party quantity, rarity, sextants and scarabs and a dedicated MF culler (peak efficiency of every juice mechanism that exists). Every other player is unaffected on average. For example when playing Breach, the reduction in currency items found is around 7% (when comparing 3.19.0d to 3.18.1f). In 100% Delirium maps, the difference hits 17%. In Incursion and regular non-league content, you'll find 25% more.

The next mistake we made was related to item culling. I am pretty sure I spoke about this on a podcast at some stage, but a while ago we introduced a system that culls some percentage of irrelevant normal and magic items before the items drop, in higher-level areas. These are items that would almost certainly be filtered out by almost any item filter, and are almost never picked up. The intention is to reduce clutter substantially without actually affecting any items a player would pick up. We have been gradually raising this culling value over time as we try to find a sweet spot that has the best performance impact with no gameplay impact. To be clear, this system doesn't affect things like rare items, currency, maps, etc. A few weeks before Lake of Kalandra launched, we raised the rate again. This means that if you're counting the raw number of irrelevant equipment items on the ground, some of the reduction is due to this harmless culling system rather than actual drop nerfs.

In addition, Lake of Kalandra is an out-of-area league. Its rewards entirely come from the Lake itself, rather than from your maps. This is in stark contrast to Sentinel, our last league, which not only dropped rewards in your maps, but was honestly tuned higher than average in terms of league rewards. Players went from receiving masses of league rewards as they clear maps to receiving absolutely nothing from the league until they travel to the Lake. This is unfortunate timing and exacerbated the perception of drop reduction.

The Lake itself was also relatively unrewarding on release and this has since been massively increased since then.

The remaining things that went wrong pertain to post-release communication. It took us several days to hotfix many of the changes in, and while we have posted about it each day, this full explanation took almost a week. I wish we could have done it faster, but we have tried to prioritise working on the actual fixes as quickly as we can. As the confusion about our motivations has raised a lot of concern with the community, I should have found a way to prioritise writing this post.

Improvements to testing and communication in the future

There's a lot to unpack from the above pile of mistakes. I believe that the intention was good, but there were significant deficiencies in testing and communication. I take personal responsibility for those areas, because they happened on my watch. I'm the Game Director for Path of Exile 1, and it is absolutely unacceptable that I can miss a change that has the consequences that the league monster one did. Changes like that need to be very, very carefully tested, have their consequences fully understood, and then be communicated clearly. I have let you down and I will not allow it to happen again.

I want to emphasise that our Quality Assurance team are not to blame for the issues that were not discovered before release. They work really hard and have a lot of limitations that are outside of their control. For the next upcoming release, I am specifically trying to integrate them more into development so that we get their feedback earlier during the development of features.

The direction from here

So where does this leave us?

For players who are juicing their content to extreme levels with six-person parties, dedicated MF cullers and stacked league mechanics, they no longer have Beyond to push things over the edge. But they still find ridiculous amounts of stuff. I have seen parties in this league get multiple mirrors per day, or find over 50 Divine Orbs from a single monster.

For regular players who are just alching their maps and adding difficulty where they feel they can handle it, we think that drops are in a pretty good place after this week's changes. They should have been like this at release, and I am deeply sorry that they were not.

Our plan is not to gut the rewards out from Path of Exile. We play the game too and enjoy finding heaps of valuable items. Our "could an alternate version of the game with extreme item scarcity also be fun?" experiment, currently internally called Hard Mode, is an entirely separate thing and its changes have not been folded into regular Path of Exile.

Please keep the feedback coming. We are reading, discussing, and continuing to make changes. I'm very sorry for the rough start, but I hope you continue to enjoy the Lake of Kalandra, Atlas Memories, and other new content released in this expansion.