Consistency Patch Reception

[[TECH]Gorb]

(less facetiously, all these changes were a waste of time compared to actually examining undesirable outliers and modifying them as desired. In fact, I would prefer that (almost) all skills scale linearly, but just going 'ok all skills scale linearly now' is not the way to do it)

"I would rather they changed almost all skills to scale linearly but I'm not happy that they changed almost all skills to scale linearly".

You have a problem, I think, with wording your posts. The "keep it civil" in Ekamo's OP also still applies.

[changlingbob]

It is not that they now scale linearly that I object to; more it is that the decision appears to have been made to make everything scale linearly and just done it. I mean, pebbles took a nerf, but as noted a couple of posts above, was the lesser growth on rank 4 an actual balance consideration originally that, in going 'I want the line on this graph to be straight', has now been overlooked? Is this outlier desirable or undesirable? Is it there for a purpose? Too bad, now its a straight line.

Now I actually misspoke when I said I would prefer skills to scale linearly; I would actually prefer skills to be regular. Old stalegmites is obviously irregular: 100/180/260/300. Linearly is obviously regular: 100/160/220/280, increase by 60/level. So however is something like the following: 100/180/240/280. This latter example increases by 80/60/40 each level, and very clearly determines whether this skill is a skill to maximise as fast as possible or a skill that has diminishing returns (if this is indeed the intent of the 40 at last level).

We already have skills that are one-point-wonders (valkyrie's javelin does not change stun duration with levels, for example), and there are skills that just get better each level with improved everything (damage, mana cost, range, whatever; I can't think of any off the top of my head though!). There are even skills such as demented shaman's entangle that have other optimal points. To me, examining why a skill is irregular and making it regular in such a way that fits the design and balance intent is more important than just making everything that currently is irregular and jamming it into a straight line distribution, as was done.

[[IaF]Skyve]

I believe status quo should be retained unless there are solid arguments for tampering with it. And I have yet to see any solid reasoning why a nice looking number progression is directly beneficial to the game.

So you would rather still have two decimal points on some heroes statgains? Because what was the purpose of that change other than to make the numbers look "nicer"?

[[eR]Ekamo]

There are some reasons for why I think consistency as a reason in itself shouldn't be striven for. It should not, in my opinion, be the only reason for a change, and should always be subordinate to two other much more important considerations, balance and design.

A neat scaling of numbers might look nice for someone who loves linearity and straight graphs (hrmhrm) but from a design perspective it is as arbitrary as it gets, and so is any scaling until you figured out what effect you want out of it. This effect is what you want to achieve during different stages of the game with the skill, and ultimately the hero, and can be very deliberate in its implementation. Demented Shaman's Entangle is one of the few extremes out of this, but there are/were/could be several more.

Let me take Pebbles previous scaling for example. I could easily make the argument that the design intent behind his "odd" scaling on Stalagmites was to quickly make him a force to be reckoned with early game, while keeping his overall max potential slightly when he had his skills maxed out capped.
Or why not Arachna? She had a longer cooldown on level 1 to slightly try to keep her already ridiculous ganking in place, while later on when people group up more and she loses most of her solo ganking advantage it transitions into a more effective and consistent weapon to be used every teamfight/encounter which retains her relevancy.
What about Forsaken Archer? They were designed that way to retain their high effectiveness to harass, but being harder to lasthit with, just like any hero is with their differing damage ranges. Same thing could be applied to Tundra and Coeurl.

I could probably make this case for five more of the changes in the latest patch, where I think the good design that the previous "odd" scaling promotes way outweighs any arbitrary, neat number scalings. Don't take me wrong though, I do realize that some of those previous numbers merely were Dota fossils with no real reasoning for their odd scaling, such as Blood Hunter or Devourer and were changing them was probably the right thing to do, because in this case the benefits previously outlined by some of you with applying the neatly scaling numbers outweighed the non-existing good design that was gained.

What I'm asking for is thought behind the possible implications behind every change, and not indiscriminately approving of everything that removes RNG, weird scalings and odd Dota mechanics. There might be a reason for why they are the way they are, but then again, there might not. Either way, that possibility should not be ruled out until a proper analysis has been conducted, and that is all I'm asking for.

So you would rather still have two decimal points on some heroes statgains? Because what was the purpose of that change other than to make the numbers look "nicer"?

Those were hidden numbers, only accessible by going into the game code itself, and thus a completely different matter that required its own discussion. Which it probably received.

[[TECH]Gorb]

@Ekamo: while I feel Pebbles should be debated in his thread as his change was more of a balance-affector than most of the others, I will say that a 20 40 damage loss is not going to harm a top-tier hero that has a good fifty twenty-five times that number in (regular) burst.

With regards to your examples, do you think Arachna needed the buff? Is Arachna in a decent place at the moment? Would this change push her into an overly-strong position? As a counter-argument I could argue that her early-game harass comes from the ability to (ab)use her orb attack to bypass creep aggro and slow the target for further (hard-hitting) auto-attacks. If you want to nerf her early-game, perhaps nerf the scaling of her orb? Or look into other options.

The same goes for Forsaken Archer (beyond the fact that your primary last-hitting control comes from your autoattacks and thus additional help from Skeletons is simply block damage added to your autoattacks, accounting for the delay between auto and Skellie attack).

You seem to be resisting the changes on the ground that they are change rather than opening up avenues of discussion on the heroes themselves (perhaps in their own threads) and seeing what impact the changes have on the hero.

It is not that they now scale linearly that I object to; more it is that the decision appears to have been made to make everything scale linearly and just done it. I mean, pebbles took a nerf, but as noted a couple of posts above, was the lesser growth on rank 4 an actual balance consideration originally that, in going 'I want the line on this graph to be straight', has now been overlooked? Is this outlier desirable or undesirable? Is it there for a purpose? Too bad, now its a straight line.

Now I actually misspoke when I said I would prefer skills to scale linearly; I would actually prefer skills to be regular. Old stalegmites is obviously irregular: 100/180/260/300. Linearly is obviously regular: 100/160/220/280, increase by 60/level. So however is something like the following: 100/180/240/280. This latter example increases by 80/60/40 each level, and very clearly determines whether this skill is a skill to maximise as fast as possible or a skill that has diminishing returns (if this is indeed the intent of the 40 at last level).

We already have skills that are one-point-wonders (valkyrie's javelin does not change stun duration with levels, for example), and there are skills that just get better each level with improved everything (damage, mana cost, range, whatever; I can't think of any off the top of my head though!). There are even skills such as demented shaman's entangle that have other optimal points. To me, examining why a skill is irregular and making it regular in such a way that fits the design and balance intent is more important than just making everything that currently is irregular and jamming it into a straight line distribution, as was done.

Does everything now scale linearly? Have any abilities been left out?

Is it impossible to assume that the consistency patch could be used at a basis for further variance (based on regularity, as per your approach - which I approve of, btw)?

I want to thank you for responding and I agree with what you are saying, however regardless of their intentions/inspiration behind this patch, I am confident that it allows a stable platform for further experimentation and scaling (even if that wasn't the particular aim).

[changlingbob]

Does everything now scale linearly? Have any abilities been left out?

Is it impossible to assume that the consistency patch could be used at a basis for further variance (based on regularity, as per your approach - which I approve of, btw)?

I want to thank you for responding and I agree with what you are saying, however regardless of their intentions/inspiration behind this patch, I am confident that it allows a stable platform for further experimentation and scaling (even if that wasn't the particular aim).

I'm not sure if things have been left out, but most heroes appear to have been considered.

With that said, I actually disagree about the patch giving a stable platform; I mean, it does, but now if we want to consider non-linear but regular scaling we have to go back a couple of revisions before we can go forward. Its not a huge thing, but it does get in the way of just examining what the current numbers are and what the state of balance is and determining a solution. Its a stopgap to that aim, rather than actually trying to achieve it (assuming that is the aim). Making everything linear for now certainly aids ease of use, especially for new players, in the short term, but I bet if some of them get semi-reverted to a scaling more similar to the irregular changes people will get confused, especially those newer players who are just getting the hang of things.

[[TECH]Gorb]

The thing is, you would have had to enforce consistency as a whole (i.e. removing a layer of design complexity that is the outliers of DotA-ported values and also irregularities of S2 hero design) before applying another layer of design complexity (that is regularity in scaling for a few heroes that need it), simply because you need to evaluate which heroes actually need it and which would rather benefit from different tweaks (see the suggestions/issues in the OP of the Pebbles thread).

Consistency makes thing easier to read, it allows people to make better sense of hero strength progression and finally eliminates any confusion when attempting to pinpoint the cause of over (or under) performance. This is why I feel that it provides a solid base for further work.

If the regularity was introduced (per hero that may require it) within a short timeframe to 2.6.10, I'd agree about the confusion. However I would argue that this theoretical second balance patch would be a substantial amount of time away to allow the standardisation changes to be tested and observed. This gives players time to acclimatise, in addition to incremental scaling being easier to get your head around than seemingly-random values, a la Pebbles' old Stalagmites damage steps.

Addendum: Perhaps ironically, DSham's Entangle doesn't seem to have been touched (unless it was standardised prior to this patch?)

[[eR]Ekamo]

@Ekamo: while I feel Pebbles should be debated in his thread as his change was more of a balance-affector than most of the others, I will say that a 20 40 damage loss is not going to harm a top-tier hero that has a good fifty twenty-five times that number in (regular) burst.

With regards to your examples, do you think Arachna needed the buff? Is Arachna in a decent place at the moment? Would this change push her into an overly-strong position? As a counter-argument I could argue that her early-game harass comes from the ability to (ab)use her orb attack to bypass creep aggro and slow the target for further (hard-hitting) auto-attacks. If you want to nerf her early-game, perhaps nerf the scaling of her orb? Or look into other options.

The same goes for Forsaken Archer (beyond the fact that your primary last-hitting control comes from your autoattacks and thus additional help from Skeletons is simply block damage added to your autoattacks, accounting for the delay between auto and Skellie attack).

Did I even once talk about the state of the current balance of these heroes? Because I'm fairly sure I was just talking about the design aspect, so your comment about Pebbles was badly placed.

TL;DR version for what I think the old versions of the three heroes' skills design intent was:

Pebbles: Quickly scaling early on to be able to deal among the highest early game burst damage in the game, while later tapering off with a minor gain at level 4 of Stalagmites. This could also be tied into his hard time with positioning early on which makes it a high risk/reward skill then, but later on with mobility items when that is not as much of a problem, the skill isn't extremely dangerous percentage wise anymore.

Arachna: Weaken her already ridiculous early ganking while still allowing her ultimate to be a good teamfight tool later on. She might be on the weaker side currently, but I do not think the right way to balance her is make her even more of a threat during her early ganking phase, that is after her ridiculous laning (which I think is justified but that is a completely different debate) currently her strongest card.

Forsaken: Niche her Skeletons into more being used as a harass tool early rather than a lasthitting tool. Gives her strong lane presence but irregular skeletons combined with her damage spread of 6 on her regular attacks gives her more of a trouble lasthitting vs. strong lane opponents.

All of these three mainly has to do with design intent, and not as much with balance to do. I am not saying said heroes can't be balanced unless they stayed as they were, I am merely stating that I think they are more dull now, i.e. anti-fun.

You seem to be resisting the changes on the ground that they are change rather than opening up avenues of discussion on the heroes themselves (perhaps in their own threads) and seeing what impact the changes have on the hero.

Oh, I don't mind change the slightest. Change is good for the game, if it is implemented to achieve a specific goal that outweighs the possible downsides. The change has to be undertaken with a deliberate and focused intent behind it, and I'm currently arguing that most of the changes implemented in this patch lacked that.

[[TECH]Gorb]

My point was that you didn't talk once about the current balance of the heroes. So, um, thanks for proving that

You cannot argue from a purely design standpoint when discussing balance changes, especially given how linked balance and design are.

You are as welcome to your opinions on the heroes as I am on mine, however that doesn't make them right (by definition, opinions are not right nor wrong, but eh). If Pebbles' Stalagmites was meant to taper off, why were the values so random? Why wasn't a model adopted a la changlingbob's whereby you see diminishing returns on leveling the ability?

If Arachna's ult facilitated "ridiculously easy ganking" (again, I'd argue it was more her orb) then why were the gank-responsible aspects of the ult not nerfed (not even looking at Q for now)? A cooldown is a good tweak for balancing the recurrence of the ability, but her absurd ganking power comes from the effects of Spider's Sting combined with her slow/dot (which negates healing to boot, now). I could quite easily argue that the cooldown increase at level 7 isn't an as effective way of nerfing her early game harass/ganking potential as looking into her Q or other aspects of her R.

Now, FA. I could go on about this hero for a long time (I quite like her, Arach and Pebbles ironically enough. Not to say that I'm any good, but I've certainly played them quite a bit). Skeletons are inherently flawed in that they fold to a single attack and provide gold (/+ negligible XP) to the attacker. This promotes harassing the (rather slow) Skeletons inbetween last-hitting (which as an Agility ranged hero FA doesn't do that badly at, barring going up against, say, Flint or CD).

I also fail to see how normalising the damage scaling per level on Stalagmites, standardising the cooldown on Spider's String and regulating Skeleton autoattack damage counts as making the heroes more "dull". For starters, "dull" implies a lack of enjoyment and is therefore an incredibly subjective term (which doesn't make for objective commentary, imo). Secondly, you could make Pebbles' Stalagmites more "interesting" by adopting the method changlingbob suggested above, however in order to do this I feel normalising the damage steps is a good way to eliminate any confusion with regards to the numbers themselves (and their effects on balance). Finally (as I've covered Arach a coupla times now), FA's Skellies dealing a possible +/-(6 x variance) per combined set of attacks (assuming no harass from the support/opposing carry, assuming 6 Skeletons, etc) is not going to make or break anything but the most luck-based of situations (which in themselves have a oh God don't make me do the maths stupid probability of occurring).

I'm pretty sure at least the Pebbles change was undertaken, originally, to make the numbers look nice - however, the decision to lower the ability's overall damage probably came as a result of balance brainstorming (before even hitting SBT or whatever). However, that's just like, my opinion, man