Improved Witch Tree - Graphical Description

KenshiID, you going to do something similar for the Ranger? :D

But yes, I like the graphical depiction you made to show how much better we can make the Witch tree. (going with 2nd picture) Witch may be one of the most played classes, but we all know that good Witches will escape her tree as soon as possible, much like the Ranger because her starting highways are weak thematically.

The biggest issues with Witch are early game, they're amazing endgame for most CI, double totem, or high crit spell builds (by getting lots of mana/ES/EB and invading Templar or Shadow areas).

Main issues for early game Witch (as it relates to races) are:

+ Spells are not "less powerful" than attacks, for most races they're far superior but there are minimal ways to boost your damage with items. You can't find a 80% increased spell damage blue wand in the same way you can roll a great 2H weapon or ele cleave weapon. You can only get spell damage on your weapons and cast speed on your jewelry. Attacks let you get some form of bonus damage on every piece of gear besides helmets, chests, and boots.

+ To get a good amount of any other stat than INT you are forced to travel stat highway to templar or shadow, or spend currency on amulets. In both cases you spend a LOT of points on only stats compared to other classes.

+ Limited life and resist options, no way to get BOTH damage and life in the same number of points as other classes. No way to get resists without specializing in an element (also somewhat expensive for points; you can't get life *and* the cold resist node for merveil)

+ Support gems are pretty lame. I'd like to see Added Lightning added to the Sharp and Cruel rewards so that you could get it in combination with Fork for builds other than LMP Freezing Pulse. Of course, all of the summon rewards are useless in racing, but many classes have a lot of useless gems compared to the top tier builds.

+ Cold is arguably the only good element, Fire is not too bad since it can kill bosses very quickly but it doesn't have the AOE clear speed that cold does (especially with LMP now). Lightning is of course the worst early game. I made a suggestion to rework Shock Nova that could help (I recommend making it like Razor's electric ring from DOTA; an expanding ring of lightning that deals MORE damage to enemies on the outer diameter, deals damage when it expands and contracts back, player can move during the nova duration), but Spark at lvl 1 is never going to win a race. Arc is acquired too late and is currently not usable for most races.

+ Most monsters and, more importantly, all bosses have elemental resistance to at least one element. By comparison, there's no bosses with substantial physical resistance (and relatively few monsters with it. the monsters that have it [rocks, bears] are often skipped in races.

butte / Sicily

"

Islidox wrote:

KenshiID, you going to do something similar for the Ranger? :D

But yes, I like the graphical depiction you made to show how much better we can make the Witch tree. (going with 2nd picture) Witch may be one of the most played classes, but we all know that good Witches will escape her tree as soon as possible, much like the Ranger because her starting highways are weak thematically.

Yes I will.
I already have the tree drawn, but I have to make it in PS later.
I just got home from work (where I did the witchtree since I am a media designer, no body knows what I am working on as long as I'm working with PS ;D). So I need to take a break first.

It's only true for default that there are the most witches.
But in HC Witches are even below Rangers.

I like the idea of the new life/ES pathing on the left side, as well as changing the 30 INT node to be more in line with Acceleration/Master of the Arena style of stat-focused passives. Buffs to the starting survivability of Witch are in order I think, especially considering the tendency of any ES-based build to dive for the body&soul and nullification clusters.

There are a few problems, however. Adding any new source of max elemental resistances has to be done very carefully. We've seen a handful of uniques do it, such as the recent phoenix shield and Alpha Howl, but the designers have had to delicately handle the values so as not to reach potential elemental immunity in a build that takes every boost to max res possible. For example, Rise of the Phoenix enables the creation of a pure life-based Righteous Fire build, since the combo of very high max fire res allows a player who invests significantly into life regeneration to completely negate the RF degen. Another 3% max fire res would make the breakpoints for such a build dramatically easier to reach, since that alleviates the equivalent amount in regen. That's just an example with a build I've planned out myself, but I'd imagine similar issues could arise with other unforeseen combinations.

And while I like the counterpart to the Alchemist notable, I really think systemic problems with mana flasks themselves should be addressed before values for buffs to mana flask nodes can be settled on. With auras as powerful as they are, most builds tend to invest in max mana only in order to reserve as much of it away as possible. In turn, most effective endgame builds/players focus on regeneration, leeching, or blood magic as their method of sustaining the primary skill. It's hard to imagine any change to mana flask passives, or even mana flasks themselves, making a large, unreserved mana pool + flask rotation an attractive option in comparison.

"

Uvne wrote:

There are a few problems, however. Adding any new source of max elemental resistances has to be done very carefully. [...] Another 3% max fire res would make the breakpoints for such a build dramatically easier to reach, since that alleviates the equivalent amount in regen.

Yupp that's right, that's why I made 2 versions. I think the curent 15% res are pretty good by them selves, but I also think that Elemental Adaptation is missplaced on the tree atm giving templars and Marauders one of the best node in the game.

The whole reason Duelist travel to the Marauder tree is because elemental adaptation, while on the ranger tree there is almost nothing to get.

"

Uvne wrote:

And while I like the counterpart to the Alchemist notable, I really think systemic problems with mana flasks themselves should be addressed before values for buffs to mana flask nodes can be settled on. With auras as powerful as they are, most builds tend to invest in max mana only in order to reserve as much of it away as possible. In turn, most effective endgame builds/players focus on regeneration, leeching, or blood magic as their method of sustaining the primary skill. It's hard to imagine any change to mana flask passives, or even mana flasks themselves, making a large, unreserved mana pool + flask rotation an attractive option in comparison.

I read that people hated that there was so few mana options in the tree, that you are mostly forced to go, EB, BM or BM gem.

One way to fix mana issues is to make manapots better. I always like using the Passive tree to work not only has a Life and Damage source, but also as a form of utitily booster.

I love the Alchemist node. Whether the Mana version will be as good, worse or better is questionable but I thought it fit in well. It's also surrounded by other nodes so it is not a node you have to pass.

Also hybridflasks do gain from this passive. I know people are using those instead of Manapots (I am, but I also use manapotions). And the 30% recoveryrate applies to all flasks.

Good suggestions, the ranger and witch were in need of rework in my mind. I very much agree with splitting the ES% and HP% leading up to the crosspoint and a buff to the notables (especially giving it some all %res); the one problematic factor is that I am never going to take mana flask nodes due to lategame POE meta with maxing out mana, as someone already mentioned. I don't think I've seen any top player have even one mana flask, save for no-regen maps.

The only thing I could contribute to the discussion as an HC 84 witch and top ~75-25 overall witch racer that hasn't been touched on yet is the need for at least one flat +HP or +ES toward the start; a lot of other classes seem to start with a lot more HP, and even with two corals and a leather belt I end up with roughly 200HP by Prisoners Gate, and somewhere around 250 in Act 2. Even a flat 12HP like the duelist's first physical node would help a lot.

"

Spikey00 wrote:

Good suggestions, the ranger and witch were in need of rework in my mind. I very much agree with splitting the ES% and HP% leading up to the crosspoint and a buff to the notables (especially giving it some all %res); the one problematic factor is that I am never going to take mana flask nodes due to lategame POE meta with maxing out mana, as someone already mentioned. I don't think I've seen any top player have even one mana flask, save for no-regen maps.

The only thing I could contribute to the discussion as an HC 84 witch and top ~75-25 overall witch racer that hasn't been touched on yet is the need for at least one flat +HP or +ES toward the start; a lot of other classes seem to start with a lot more HP, and even with two corals and a leather belt I end up with roughly 200HP by Prisoners Gate, and somewhere around 250 in Act 2. Even a flat 12HP like the duelist's first physical node would help a lot.

As you can see I included a flat +10 Life node in the witchtree.

About ManaAlchemist.

Well GGG decided to put in 3 really bad 8% manaflask nodes, that no one ever takes.
At least I improved them and added even a really good one in to make them even better.

But if people still ignore it, so be it.
If at least 1 person uses it for leveling and specs out later, it was good enough.

Ranger Tree Take one here >>> http://www.pathofexile.com/forum/view-thread/122071/page/5/#p3299904

Some solid ideas here.

Two points I especially agree with:

- The ele specific small nodes: I never understood why a +dmg % limited to a single element would be only 6% so close to the start. Does anyone take the 2 small? If it is worthwhile to take the Walker notable, seems almost always better to path in from the top.

- The witch ES cluster does not compare to the ES clusters for Templar & Shadow. This is especially odd given that ES is ostensibly the Int defense model. While Templar & Shadows get 8% ES + 8% armor/eva, the Witch gets 8% ES + have a nice day. The notable is similarly oddly underpowered: the Templar gets more ES off the bat with 20% armor & 10% res to boot; the Shadow trades 8% ES for 10% eva, 10% recovery, and 10% res. Still quite a good deal.
For the "pure" ES class, I think their ES got "cut" a bit.

Hope GGG takes a look.

Some great ideas floating around in this thread, although I'm pretty sure the GGG devs are probably already looking into some passive tree tweaks here and there(as they always are I would assume).

Either way I'm looking forward to some improvements to the Witch and possibly Ranger tree whenever GGG finds the time to look at them in-depth, especially for races, because they are rapidly becoming a big thing for the game.