The simplest method to balance all Classes and Specs immediately for instanced PvP and PvE

Community General Discussion
402

“But the point is they wouldn’t. Again, you’re wanting to ignore the building blocks that lead to whether a key is completed and only focusing on the end result. But that’s quite literally an important task because different combinations of building blocks will never have the same completion rates if you only focus on the end product.”

Of course it would result in equal completion, because it will increase the absolute value of the four modifiers until it is achieved. That’s the whole point.

"You seriously think the top 10% of players are absolutely perfect? "

No. But I know that the rate of errors committed by the top 10% is normally distributed, so again this is moot.

403

I really don’t think it would though, which is the main issue. You haven’t demonstrated how tweaking a small number of factors will account for differences in completion rate brought about by a larger pool of factors. There is no guarantee that you will be able to find such a fine grain change in output with fine grain adjustment to an arbitrary input.

If the reason completion rate for an output spec is lower than a utility spec because the group is because the group can’t survive overlap X, the only way the output spec can overcome that pull via damage output is to give them enough damage to kill the enemy before overlap X. But that change in damage to overcome that pull will have downstream effects for the rest of the dungeon. Now the wall boss of the dungeon that contributed to completion rate is also easier for that output spec. The completion rate won’t change for the output spec until the binary outcome of overlap X flips from fail to pass, but then you’ve overshot the target completion rate because the rest of the dungeon became easier.

This cycle can exist with every spec that your algorithm needs to tweak. Finding a perfect balance will be truly impossible and I’ve already highlighted how miserable the transition period would be to play through. We’d see all the downsides but realize much fewer upsides.

Do you have any proof of this claim?

404

Trust him, bro. After all,

3 Likes
405

Absolutely!! Because then I could play the class/spec/talents I want and still be viable. An equal slate for everyone means no more FOTM classes taking every slot, increased PUG potential, larger player pool for every activity. Speaking as a ret pally main this would be a HUGE plus to me, because certain specs have suffered greatly under Daddy Blizz’s “kind” ministrations. :hear_no_evil::see_no_evil::speak_no_evil:

1 Like
406

Well no, you’d still have to play the talents Blizzard optimized your spec for

407

Nah, cuz a bad player will still do less damage.

408

No Sh!t Sherlock!! I know I’m a filthy casual.
But if all specs were better balanced then maybe when I apply for my M+3/Normal PUG I wouldn’t get the standard “LOL Ret bAd!” denial. I’m not asking to balance a filthy casual like me to match a KSM hero, just enough that I’m not automatically denied on my Ret/kitty/frost DK solely based on my spec. :speak_no_evil::see_no_evil::hear_no_evil:

1 Like
409

No one is sending you a “LOL RET BAD” message for your +3/N.

1 Like
410

Not every time but 3 times so far this expac words to that effect. :hear_no_evil::see_no_evil::speak_no_evil:

411

No one is wasting their time doing that.

Esp since ret arent even bad for that content.

1 Like
412

So … three times in a year.

413

Sort of.

The idea is that underused talents would also get incremental buffs until they were chosen equally by the top 10%.

However it would still require you to recognize which talents to choose together to make the use of them all.

And since incremental talent buffs will still apply uniformly in PvE and PvP, in 5 man and 20 man, the algorithm for talent balancing (which would be an independent neural network operation from the proposed M+ neural operation), would also have to take this into consideration.

The best way to program this would be two ensure that at least two talent builds are optimal for PvE and at least two are optimal for PvP (which does not imply mutual exclusion).

The best way to isolate such unused talents would be to look at the top decile of that class and spec and ask “is talent X be used regularly in PvP or PvE 5man or PvE 20 man?”

If talent X is not used instanced PvP or M+ or raids, then clearly talent X needs some sort of help.

We would then have the Neural Network reach the following conclusion via similar cost function on talent usage (instead of completion rates for keys): “Since Talent X is in Row N of this talent tree, and PvE players already have diversity in Talents Y and Z in this same Nth row, then let us buff talent X for PvP, since all PvP players choose Talent W, and don’t choose talent X,Y or Z.”

So overall, you wouldn’t be able to choose any talents you want for any game mode, nor do we want that(since certain talents have strong synergy with others). However, being able to choose between at least optimal talent paths for the gamemode (PvP or PvE), would overall be healthier for all classes and therefore healthier for the game overall.

The change in talent X could be the proc rate, the cooldown, the duration or the raw output (it would seek the balance changes in these four categories until it was considered viable).

414

Maybe I’m reading it wrong, but the way I interpret your solution is to have all classes doing the same dmg output in an optimal fight. Who would play class A if class B does the same dmg but has ability xyz for utility/support. There is more to the game than damage numbers.

415

You are reading it wrong. Classes with more utility and support would have less damage and healing output compared to those that don’t, through negative output modifiers, balanced by a superhuman neural network, instead of “developers” that obviously don’t even play their own game and takes months to make any changes at all (which lead to different FoTM classes and specs, instead of eliminating FoTM).

In short, low utility classes would be monster damage machines, because they would need that extra damage (or healing) to complete X level keys at the same rate as high utility classes.