Regardless of what anyone thinks of neutral items, what I have been missing in the discussion about them is that the discussion never goes beyond “RNG bad”. For instance rarely does anybody even state the question how much RNG are neutral items actually?

That is imo a much more interesting topic. What type of RNG is acceptable and what type of RNG isn’t? How can the effects of RNG be mitigated through gameplay or by designing the neutral item drop system? But most of those question don’t even pop up because the discussion does not progress that far. So I’ll do it here:)

In order to answer these more interesting questions I will go through:

1. The RNG Factor, RNG basics
2. Playing around RNG
3. The RNG factors of neutral items
4. How the neutral item system is designed with regards to these RNG factors/how it has been changed to accomodate for them
5. Conclusion: on the RNG element, what do neutral items add to the game?, were neutral items worth being implemented?

1. The RNG Factor, RNG basics

Essentially the RNG Factor is the proc chance with regards to the strength of the proc, in particular over a limited amount of throws with the dice.

Every attack, or in Multicasts’ case spellcast, is a throw with the dice with a chance to trigger the respective event, with enough throws the results will average out, naturally. If we are only talking 3 attacks though the probability to have reached the average are low. A low number of throws with the dice increases the variance/RNG factor.

An ability with a low proc rate and a high power level, like PA’s Coup de Grace, has a much higher RNG factor than Juggernaut’s crit which has a much higher proc rate and a much lower power level on top of that.

Generally speaking the abilities with a high RNG factor are much more likely to be perceived as bullshit due to the much higher variance in outcomes, especially if the number of dice-throws is limited. If a Juggernaut walks up to you and crits you 2 times in 3 attacks then the difference in outcome is not compareable to PA throwing a dagger, it critting, and then getting another crit on either her first or second attack from phantom strike, which also happens to proc almost simultaneously with the dagger crit proc.

PA’s item choices and mobility causes the feeling of RNG to be amplified further since the time Juggernaut spends walking up to you with his manta and butterfly, PA spends daggering and phantom striking you already with a deso and a basher, further fueling her crit.

As a result even assuming if both abilities had the same average damage amplification, which clearly they do not, PA’s crit would be considered much more lame because of its significantly higher variance which is supported by her gameplay, consequently the RNG factor on Coup de Grace is increasing the power of this ability which probably restricts the devs in how powerful PA and her crit can be before it starts causing problems on a pub game level, which is probably why it is this ability that is frequently tinkered around with, even on smaller scales, whenever PA is being changed, in particular when too strong.

An example of how Icefrog has handled things like that in the past is Ogre’s multicast from 6.82, in it Iceforg increased the proc rate of multicast significantly for a 2x proc, resulting in more multicasting, but at the same time the base strength of fire blast was reduced meaningfully, as a result the RNG Factor and variance went down. Multicast moved a bit away from being a Coup de Grace type ability towards being a bit more of a juggernaut crit ability, more of an average amplification rather than a super high proc event, without sacrificing the multicast-feel, possibly even adding to it. Good change imo.

Note: That ogre approach would not work for Coup de Grace, since PA is actually designed around having that murder crit/this mega amplification upon proc with short bursts of attacks and retreating. She would feel wrong if what was done to Ogre were done to her. If Icefrog ever were to tackle the RNG factor on this ability I could imagine it being done by adding counter, procs on every Xth attack, since PA is very well suited/designed to do that (and it fits thematically for a Coup de Grace), while Juggernaut is not designed to have a crit working like that.

2. Playing around RNG

The example Sumail brought up which I will expand upon was the rune spawn RNG of the game. Several people however brought up that these too are an example of RNG they dislike, even if to a lesser degree.

I will argue that regardless of the opinions on runes, they are a type of RNG that is meaningfully interacted with for a very simple reason: you can control it to a large degree. The term rune-control developed specifically for that reason.

You can can neither know where on the 2 spots the rune will spawn, nor can you know which rune will spawn, nor does the game consider that a regen rune for a mid core Skywrath is more likely going to be far more powerful than it is for the enemy meepo mid laner (random example, feel free to dislike it^), in other words the effect of a rune can vary drastically depending on heroes and game situation.

However what you CAN control is vision over those areas, vision over the enemy midlaner, you can control lane equilibrium, you can control your own actions with regards to the enemy hero and potentially cut him off and you can know when the rune will spawn and plan around that accordingly. You can have the map awareness to keep track of the enemy supports and whether or not they are missing. And you can communicate with your own team and you can decide on what priority you give the rune depending on whether or not someone has a bottle or not or what your sustain situation is like.

Assuming you properly control the runes you can even wait and see which rune spawns and then base your next actions according to the rune spawn. Is it a haste rune to gank a sidelane or invade the jungle? Is it a DD/Illusion to dominate mid? Is it a regen which makes you quickly spam your abilities and potentially harass the enemy hero at his own tower?

Long story short: The RNG factor of rune spawns can be be greatly mitigated/controlled by either team. In fact, it forces you to do so.

Despite that though rune spawns are still an element that is up to discussion and evaluation, I personally preferred bounties as alternatives to the other rune spawn for those draw scenarios in even lanes where one will get the rune and the other won’t purely based on luck due to how even the lane is, which was especially fair once bottles were involved or when those bounties enabled the purchase of a bottle.

However runes currently are not an immediate concern and not game ruining, and clearly it adds meaningful levels of depth to mid-laning based on the factors which I listed that are within your control. Mid-laning would be much emptier, brain-dead and easier without runes, hence nobody ever seriously suggests actually removing runes rather than compensating for the RNG element in their problem solving suggestions.

The take-away here is, RNG that can be interacted with, that can be played around with, can not only be a non critical issue, it can add meaningful depth, which DotA prides itself on.

(felt like I had to generally defend RNG a bit here considering the last couple of weeks on reddit)

3. The RNG elements of neutral items, evaluating the RNG factor

Trying to keep this shorter, the RNG element:

a) whether or not a neutral item drops is somewhat random

b) you can not influence which neutral item drops

c) similar to runes the game does not consider what heroes are being picked, thus neutral item X can be significantly more powerful on hero Y (like Essence Ring was on Io or Clumsy Net was/is? on Skywrath)

d) a neutral item can be disproportionally more powerful in a losing or winning game (like repair kit)

e) a neutral item in the same tier can simply be worse than a other one

f) a neutral item can have almost blink-like reveal consequences

Going through the RNG factor of each:

a) pseudo random and farming alone makes neutral items actually dropping based on RNG a complete non-issue, could argue for T5 if you are turtling in your base this is a factor, since you don’t have access to the jungle, but regardless of whether or not that is a good or bad thing, it is not at all RNG related.

b)c)d) these are clearly RNG elements with a high RNG factor, beyond moving the items around in your team you can not influence this, those are legitimate RNG concerns by design.

e) this is a balance issue not unlike basic hero balance, not a design or RNG issue, we can ignore it

f) this is another design issue

4. How the neutral item system is designed with regards to these RNG factors/how it has been changed to accomodate for them

If we want to answer the question of how much RNG are neutral items actually?, we now have to compare the RNG issues with the design of neutral item drops and the changes so far.

b) you can not influence which neutral items drop

In order to compensate for this the entire neutral item drop system needs to be set up for it to not particularly matter which neutral item drops. Probably the biggest restriction on the design and it will force neutral items to be mostly relatively basic and generic stat sticks. Or it will force nerfs on those outlier/unique neutral item drops.

There have been several changes that have compensated for this RNG element since 7.23.:

• The increase from 3 to 4 items reduces the RNG factor in the same way increasing the proc chance of an ability or item would, like it was done to ogre multicast in 6.82, it reduces variance.
• The restriction to only being able to carry 1 neutral item further reduces the RNG factor of this element, no longer all item drops are potentially being used, they now compete amongst each other on per hero basis thus reducing the impact of favourable/unfavourable drops, if the increase from 3 to 4 neutral item drops per tier is analogue to an increase in the proc chance of a passive critical strike, then this is analogue to reducing the power of the proc, a doublewhammy identical to the Ogre fireblast and multicast change in 6.82
• while this seems a bit like a cheap point, again balance amongst the items and with regards to RNG element c) will reduce the RNG factor because it will no longer matter what item drops (or barely)

c) disproportionally powerful items on hero X

This is a design issue on a per item basis, for instance Vampire fangs spell lifesteal is pretty average on most heroes, but then there is a Bristleback or an OD. Similar to Essence Ring, which was changed to no longer heal because of Io or heal amplification heroes, these features need to be either removed or nerfed. Which of course happened to vampire fangs in 7.24, reducing the spell lifesteal from 8% to 6%. This is interesting in so far that Icefrog likes to do the opposite with heroes, he pronounces their uniqueness strengths very frequently, these neutral items need to work exactly the other way round, not entirely but mostly generic (like League of Legends, lul, and no, I am not fucking saying DotA = LoL now, not at all, I am saying that’s what League by design needs to do with their entire game).

In extreme case scenarios restriction or changes like for essence ring need to occur or have occurred,in worst case scenarios these items need to be removed, like Book of Aghanim was gutted instantly.

As a result, as mentioned, this forces most of the neutral item drops to become rather basic stat sticks. Total elimination of unique features like the vampire fangs spell lifesteal will not be necessary, but several neutral items, like Ballista in 7.24, will have that unique feature nerfed while gaining another side feature. Grove bow for instance gained a universal feature on top of its rather unique core feature for similar reasons I presume.

A minor element of RNG will always remain with regards to item X on hero Y, but we have plenty of evidence by now that not only will this be mitigated quickly, since this is a design and not a balance issue and those need to be hit quickly, the RNG factor of this element has already reduced drastically.

Imo this is a very solveable issue.

d) neutral item being stronger when winning or losing

Those are extremely limited, repair kit is basically the only one, could argue the leveller is an offender here too, but similar to grove bow or ballista again, the leveller has such a solid base secondary stat that it is generic/basic enough except maybe when it happens to drop for Tiny or Spirit Bear, which seems awkward. This could require demolish abilities like spirit bear/Tree Grab/Leveller to not stack with each other for instance.

Similar to c) this seems entirely solveable, actually much less of an issue. I’d argue repair kit is much closer to being removed than it is to being acceptable, but that’s really the only item that is relevant here. And even repair kit is not even compareable to its splitshot iteration from the last patch.

f) a neutral item can have almost blink-like reveal consequences

The poster boy for this right now I’d say is mindbreaker. Book of aghanim is a perfect example of this too. There are barely any of these left, most can be reacted to during combat. Mindbreaker clearly is an issue here atm. These type of items basically should not exist imo. Iron Talon is an invisible but annoying offender to this rule aswell, cause the jungle AM which you beat up on lane might suddenly show up wiht battlefury and manta at minute 22 despite getting his ass kicked hard on lane and you are just like “wtf?”.

Almost all changes since introduction of the neutral item drop system follow the pattern of either removing/nerfing a trademark feature, removing an item that offends against RNG in any way too much, and go for balance beyond that.

It is imo being made VERY clear with those changes that RNG is not meant to be, and not going to be, relevant with regards to neutral item drops.

5. Conclusion:

• on the RNG element
• what do neutral items add to the game?
• were neutral items worth being implemented?

Whatever the RNG element of neutral drops originally was, by now it is a fraction of what it used to be, that much is certain. That however might not mean much though, because even while it got reduced massively maybe the RNG factor is still so high that it is game-deciding/ruining, right?

At this point I’d say though that within the next patches RNG will basically not be a particularly meaningful factor anymore, if it isn’t already. Not any more meaningful than that of regular proc based items and abilities, except for tier 5 maybe. The reasons for me stating this are:

• the improving balance among the tiers negating 1 of the RNG elements
• the removal of offenders like Book of Aghanim or recently Fusion Rune (like repair kit eventually?), reducing another massive source of RNG
• the nerfs of disproportionally powerful features like spell lifesteal on vampire fangs for instance making the extreme outlier items less of a problem, including something like clumsy net on skywrath since 7.24
• the reduction of the RNG factor by increasing the drop rate from 3 to 4 on top of the recent 1 item only restriction really damn near negated any other RNG factor beyond those I just listed that have been improved upon

I’d expect tier 5 to take longer than that purely because of the low sample size to get these into a better place. However based on 7.24 it might not take long either.

I’d be curious and interested if someone could actually argue how they still think neutral item drops are deciding games based on RNG. Because even though the drops essentially are random, it really damn near does not matter anymore by design. Even if currently the RNG factor of a specific drop combination is still an issue, there is so much evidence by now that this will most likely be removed/changed sooner than later, that I do not consider RNG an issue in the discussion about neutral items anymore, or at least not for much longer, it’s merely a shortterm issue.

What neutral items add to the game on the other hand, that is a different topic, since something not being a problem does not mean it existing is a good thing for the game either. Especially considering that in order for them to exist, they must necessarily be relatively generic, as I have laid out.

Currently I’d argue that it is questionable if adding them was worth the effort, if it was worth adding them considering the initial damage, polarization and uproar they caused. I think an answer to this question would strongly depend on the results of the discussion on what neutral items actually add to the game, if we were to agree on a variety of upsides down the road, then they might have been worth the trouble, however this is something we will probably be able to reflect upon only further down the road.

And as a final word, at this point I personally think neutral items are a surprisingly small element when compared to what they were on 7.23 release, basic economy, tempo of the game and regular item balance, business as usual imo, are bigger deals than the neutral items at this point. Regardless of whether you love or hate them, the importance of them is just not on a tier high enough anymore to strongly influence my opinion on the game/meta and even if they still are and always will be, their existence is not something I’d judge based on RNG at all anymore.

But that’s just me, feel free to disagree:)