League of Legends on-hit items: three items buffed in Patch 25.14, and a question still unanswered

On a League of Legends community thread, a Riot Games designer left a few lines about Gươm Vô Danh. The content centred on the fact that the item needs a buff, that the entire on-hit item group has a problem, and that any adjustment would have to account for both melee and ranged users, with the ranged side prioritised. No patch accompanied it. No release date. No win rate, no pick rate, no gold efficiency per item slot. The original piece itself notes that the statement has not been officially confirmed by Riot and that it reached Vietnamese readers through a Reddit round-up.

That is the part worth pausing on. A balance debate that ran across multiple forums, drawing everyone from bronze players to Challengers, closed without a single quantitative metric being used to defend any position. The entire weight of the argument rested on feel. And feel cannot be verified.

I have followed debates like this long enough to recognise a repeating pattern. When a community says an item is weak, it is usually right about the conclusion and wrong about the cause. Confirming that an item is weak is relatively easy. Establishing why it is weak is the time-consuming part. In this case, that time-consuming part has not been done by anyone.

What the on-hit item group actually is

The on-hit item group is the set of items that trigger effects on each basic attack: bonus magic or true damage per hit, attack speed, and stacking mechanics. On the Vietnamese server, the group includes Móc Diệt, Gươm Vô Danh, Vũ Điệu Tử Thần and a few other items whose English equivalents still need to be checked against the original client before being used as a basis for anything.

The champions who build around this group are those dependent on basic attacks: Irelia, Yasuo and Yone on the melee side, and auto-attack marksmen on the ranged side. These are two sets of users with different damage profiles, sharing a single stat frame.

The difference lies in how they deal damage. A melee champion must close the distance, absorb direct damage, and trade health to maintain pressure. A ranged marksman stands outside, needs a stretch of uninterrupted firing time, and dies the moment they are caught. One item serving both groups means every stat change pushes the two sides in opposite directions. The Riot designer called this a difficult problem, and his phrasing suggests Riot has not found the answer.

Why the previous buff round was not enough

The nearest precedent is Patch 25.14. In that patch, Riot buffed Móc Diệt, Gươm Vô Danh and Vũ Điệu Tử Thần to lift this item group. As of today, community feedback still describes the on-hit group as the weakest in the item system. That is the most important fact in the whole story, and it does not come from the designer's statement. It comes from the game's own balance history in the first half of the season.

A single buff failing to fix an item group usually has two explanations. The first: the buff was too small relative to the gap it needed to close. The second: the group is weak because the environment around it changed, not because of its own stats. The second explanation is far more serious, because it means any further stat buff will keep being swallowed up.

The Riot designer raised one notable point: Gươm Vô Danh has a very strong interaction with certain champions that have special mechanics tied to basic attacks, so even a small stat change can significantly shift those champions' power. In analytical terms, this is the stat-amplification problem. An item balanced for the median user can push outlier users very high or very low. Irelia, Yasuo and Yone are precisely those outlier users, and the balance history of all three champions is a chain of oscillations tied to items.

Numbers never panic – people panic, and people are the variable.

That line applies here in an uncomfortable way: there is no number to panic over, so only pure panic remains. Nobody in the debate could produce Gươm Vô Danh's pick rate by rank bracket, its gold efficiency relative to competing items, or the win rate of on-hit champions before and after Patch 25.14. If those figures existed and supported the claim that the group is weak, they would have been cited long ago.

I spend roughly thirty per cent of my writing time cross-checking data from two or more sources, and in this case both sources came back empty. That is a result, not a dead end. When a community debates balance without balance data, what is really being debated is the play experience, not the strength of the item.

Three scenarios and the price of each

If Riot buffs Gươm Vô Danh moderately, the on-hit group could return to broader build diversity, and the current balance state barely shifts. This is the best scenario, and also the least dramatic, which is why it gets discussed the least.

If Riot buffs hard, Irelia, Yasuo and Yone will rise in solo-queue presence, dragging a follow-up nerf into the next balance cycle. The buff, overperform, revert loop is a pattern the on-hit group has already been through several times. Each round costs two to three patch cycles and leaves behind a layer of players who grew used to temporary power and then lost it.

If Riot fixes a single item only, the group-level problem remains untouched. The original piece acknowledged exactly this, noting that the Gươm Vô Danh issue is one part of a larger picture and that several items in the same group are in a similar state. Fixing one item while the whole group is weak is the familiar way of handling it, and also the surest way to guarantee the problem returns.

Before trusting your eyes, check what your eyes have already decided to trust.

Here, the community's eyes settled on a conclusion before any data appeared. This is the point I want to separate from everything else, because it concerns how an unofficial statement gets converted into expectation.

A Riot designer posts on Reddit. The community reads it as a promise. In Riot's operating history, exchanges like this usually function as reaction-testing before a decision is made, not as a commitment. The gap between testing and commitment is exactly where expectations get inflated. When the actual patch turns out smaller than expected, the next reaction is disappointment, and disappointment generates content, not change.

A hypothesis nobody has put on the table

There is a possibility no one in the debate considered: the on-hit group may be weak for a reason outside its own stats. In an environment where burst damage and durability are tuned toward shortening fights, a playstyle that needs sustained damage over time loses ground systematically. If that is true, raising Gươm Vô Danh's stats is just patching a symptom at the end of a causal chain. The correlation between the item being weak and the item needing a buff is obvious; the causal relationship between the two has never been demonstrated.

I do not have enough data to assert this hypothesis. But I have enough data to say it has never been tested, and a balance debate that ignores the environmental hypothesis is a debate missing half of itself.

Two things never lie: data and time.

That Riot designer's statement is not data. It is a signal, and a signal needs to be placed beside numbers before it becomes the basis for any decision.

Signals to watch in the next balance cycle

Three things will answer this more clearly than any comment. First, the scope of the next patch: if Riot touches only Gươm Vô Danh, the group-level problem stays intact; if Riot touches several items in the group together, that signals an admission of a structural problem. Second, the presence rate of Irelia, Yasuo and Yone across the following two patches: a sharp climb would confirm the stat-amplification hypothesis. Third, and least noticed, is how quickly this debate cools on its own. If it fades even without a patch landing, most of the recent heat came from a demand for discussion rather than from the item group's actual strength.

A weak item group can be fixed in a few weeks. A community accustomed to debating without numbers takes far longer.