Thoughts on Warrior Passive Skill Tree and Shield Prefixes

I think a lot of people have already commented on the passive skill tree, but I wanted to share my thoughts.

1. There are not meaningful prefixes on shield, which means that the Warbringer block ascendancy nodes are not compelling or meaningful choices.

In certain interviews, Jonathan has commented that the early skill tree or parts of the skill tree should be obvious to players as being an objectively right choice. This means that certain passive skill points that you have available to spend should make sense without too much thought put into it. By this logic, "I am a warrior using a shield. It probably makes sense that with all of these block nodes available, I will want to take at least some of those block nodes. There are lots of chances to take block so even if I may not path through all travel nodes next to a block node, regardless of which way I path, I should path a direction that lets me take some block." Stuff like that. A veteran player or someone who has played warrior for even a small period of time will probably be able to tell you that you want to be block capped, which means that I know going into my build that I will be taking enough block nodes to do that which means that before I have even made my character I know that x number of my total passive skill points will be being spent on acquiring block nodes. I also know that my off hand will be a shield. I also kind of know that a shield with increased chance to block as a local modifier on the shield will be, if not mandatory, then "strongly encouraged". This is seen by the relative power and lack of power, depending on how you view affixes right now, of the Warbringer acsendancy. On one hand, it is awesome that the warbringer gets 40% base block chance while using a shield. However, there are not really that many other prefix modifiers that I would like to get on a shield in place of increased block chance. Even if I get flat armour and percentage increased armour and life as my three prefixes to a shield I go from 150 armour on a stone shield to like 300 armour - I could be wrong here, but this sucks. I would much rather have life, percent OR flat armour, and increased block chance on a shield and just spend my ascendancy points elsewhere. In short, the opportunity cost of using a shield with 40% base block chance is very low compared to using a shield with 26% base block chance but an additional 150 armour. So, the choice has been made for me that I will be using that increased block chance shield, saving my ascendancy points for something cooler, and pathing through nodes that get me to the 75% block cap. This is fine, but it is kind of boring. It is the expected META play which I am okay with and I want to be clear - I am happy to be block capped. It makes gameplay feel so much smoother. However, I want to feel like I am missing out on something. I want to feel like EITHER I choose a shield that has 40% block chance, but I miss out on a juicy alternative affix that I could have had for my three prefixes. OR I want to feel like "woah, the Warbringer is so freaking cool because I get to use my sweet, juicy, and scarce ascendancy points to get a near perfect block chance shield FOR FREE, then I now have the opportunity to craft, upgrade, gamble, or get this sick prefix modifier on my shield that I otherwise would not have room for." Because there are not alternative modifiers or relatively competitive prefix modifiers for shields, the simple answer becomes "get increased block chance, life, resistances, and % physical damage reduction on any amount of armour shield". This is boring. It is not boring because block is boring. It is boring because I have absolutely zero FOMO for going for an increased block chance shield. On Titan I know I am going for an increased block chance shield. On Warbringer I know I am going for an increased block chance shield and skipping the acsendancy point. Not because the acsendancy point for 40% base block chance isn't cool enough, but because it is pretty easy to get a 40%+ block chance shield anyway AND there is not a prefix that I would want MORE than increased block chance. Maybe add in a prefix on shield like "percent of physical taken as ele or chaos" or "percent of physical damage recouped" or "percent of physical or all damage taken is prevented, x% of prevented damage taken as DoT" - just something interesting that makes me think "I have to get that stat, but now I do not know if I will have room for increased block chance. Man I need to make a meaningful choice."

2. Warrior beginning passives or pathing on passive tree does not make sense

For the most part, you are coming off the warrior starting point and taking melee damage. Why? Isn't 20 armour good? 20 armour at level one is "okay" but even at level 1, this seems like such a garbage first point. It is even worse the later into the game you get. If I have a shield with 100 armour and take 1 minor passive that gives 25% increased defenses from equipped shield, then I am at least getting +25 flat armour. If I have a 300 armour shield then I am getting +75 flat armour. In both cases, those nodes are at a minimum 25% better than this flat node. The alternative is 8% melee damage, which isn't great and late game will be close to nothing compared to other small passives, but armour is in such a bad place right now. Either these flat armour nodes should be providing something compelling that increases early survivability more than that increased damage or should have additional modifiers on it to make the passive alright early and good enough late. The reason these are kind of bad is because if I go off the warrior starting location from the damage side, I can spend 4 points, get 30% - 44% increased stun buildup and 18% - 27% increased melee damage, or I can spend 5 points and get 63% - 90% increased melee damage (not Titan is the low end and Hulking form Titans are the high end). Alternatively, I can go off the tree on the bottom side and spend 4 points to get 40 - 60 flat armour and .4% - .6% life regen or 40 - 60 flat armour and 30 - 40% increased stun threshold. Maybe, stun threshold is more important than I realize. Life regen is good, until you realize that the 40 flat armour is completely eclipsed by a single increased armour node or the fact that armour is pretty bad overall right now or that a chest piece at the beginning of end game will realistically have 1,000 to 3,000 flat armour on it (depending on budget). This coupled with armour having diminishing returns for damage mitigation means that these first two nodes do virtually nothing for survivability early and late. The two first melee damage nodes do little to nothing late, but are quite good early because your first % increased damage nodes matter more than your last % increased damage nodes. The real difference between these pathing options becomes apparent with points 3 & 4. Stun build up is essential and every point of increased stun buildup is welcome. The more often I can boneshatter or have enemies that are not able to act is always welcome. Maybe I am over enthusiastic about stun buildup, but it is so freaking good. I will waste points on pathing to stun buildup and here is stun buildup at the beginning of the tree on my way to other stuff that I want. What is not to like? However, by relative comparison, I can get .4% life regen. This is really bad at all levels. For two pathing points, I can get .4% life regen, which translates to .4 life regenerated per second when I have 100 life, 4 life regenerated per second when I have 1,000 life, and 8 life regenerated when I have 2,000 life. It will take me 10 seconds to regenerate 4% of life, 100 seconds to regenerate 40% of life, and 250 seconds to regenerate my life bar. Or, I can press my life flask and get 60% - 100% of my life back in less than 3 seconds. I genuinely do not know why these nodes are in the game. Yes, with increased life regeneration rate passives and other sources of flat or % life regen I will have a meaningful amount of life regeneration, but I hate these nodes so much right now because they seem under tuned or nonsensical choices. For these to even be worth grabbing they should either be "% increased armour" and "% increased life regen" or they should be "flat armour & % increased armour" and "% life regen & percent increased life regen", "% physical damage reduction" and "% of damage recouped as life", or they should be something that could be moderately useful at levels 1- 10 and virtually worthless but still doing something at level 100. Even at level 100 I am certain that attributes, melee damage, or increased stun buildup will be more useful than those nodes, which means that at level 1 I should not be taking those nodes and at level 100 I should respec out of those nodes or never have taken those nodes to begin with. This kind of aligns with the idea of there should be obvious choices, but I do not think it should be that obvious of a choice to not take these.

3. Increased life regen nodes should be changed to increased life recovery. Life recoup should probably be buffed and this might be how to do it.

If the numbers need to be tweaked that is fine, but any node that currently gives increase life regen has anti synergy with life recoup effects. This would not be a problem but in the South West corner there is a node cluster that has increased armour, increased life regen nodes, leading into a notable passive skill that gives "% of physical damage prevented recouped as life". This node looks good, but is bad. If I take 1,000 damage from a hit and I am lucky enough to have 90% reduced physical damage, that would mean that I prevent 900 damage. Of the 900 damage prevented, I would then recoup 15% or 135 life would be recouped. I would recoup this 135 life over 8 seconds (recoup in PoE2 is different than PoE 1 - 8 seconds instead of 4 seconds). Meaning, I "regen" about 17 life per second on a 1,000 damage hit where I still took 100 flat damage. In a little bit less than 6 seconds I fully heal from this damage through recoup. However, this is not the actual armour calculation or how it would functionally work. In addition, this is the best case scenario. If I have 70% physical damage reduction, which is much more likely than 90% phys reduction, then I take a 1,000 damage hit, I prevent 700 damage, and take 300 damage. I then recoup 15% of that 700, or I recoup 105 life over 8 seconds, which means I "regen" 13 ish life per second. The total damage recoup is 195 life less than the hit of damage. Recoup should either be buffed pretty hard to be worthwhile, physical damage reduction should be buffed, and/or increased life regen should be chanced to increased life recovery or increased life regen should also apply to life recoup. This way, if I have 100% increased life recovery, then that notable passive would provide 26 - 34 life recovery per second. Either way, with some amount of investment, the amount of damage that node recoups should off set the damage taken over the 8 seconds by at least 40% of the damage. Otherwise, why wouldn't I just press my life flask and get 500 life regen per second. I get it that that one node should not automatically solve my life recovery or life regen, but it doesn't seem to do enough to provide a compelling reason as to why I should grab it on the way to something better. Made to last" costs 3 points to get and in my mind the life regen is not very good and the increased armour though decent is also not very good. So why would I spend any points on this at all? Why would I not just spend 4 points and get 'Hunker down", which I am also probably pathing by anyway, but the points leading up to it increase my block chance and the more I block, the more damage I mitigate and the more I block the more life I now "regen" through block recovery. If I get hit 100 times for 10 damage each and I block 75 times at block cap, then I will "regen" 1,500 life and I will take 250 damage for a net +1,250 life. If I get hit 100 times for 10 damage in the same window of time, with 90% physical damage reduction, then I will "regen" 135 life 8 seconds later and take 100 damage. 8 seconds later I will have gotten +35 life or I could have +1,250 life. These nodes, make no sense and they are 2 travel passives away from one another. I get it that I need additional investment to make block work, but I am melee, I am obviously going to get block capped. Maybe, blocked hits also apply to "made to last" so I would recoup 15% of the physical damage prevented of blocked hits as well? IDK. Even if it does, then I recoup 118 ish life over 8 seconds for a whopping 15 life regen per second that does not interact with increases to life regeneration. I am just not seeing a situation where I would want to spend any points of life recoup when life on block is right there, where a well rolled life flask is giving me 500 life regen per second, where the amount of life recouped is less than the total amount of damage of the hit, and where if I stop taking damage then I stop "regenerating" life. Life on block can theoretically make me go positive in net damage taken, but life recoup will almost always be negative, which just helps me die slower. Life on block could let me die, but it could also heal me to full life in a very brief window of time if I have enough monsters hitting me, say like in a breach or during a ritual. This does not even factor in active block, which has 100% block chance for hits from the front, but I think the point has been made pretty clearly that life recoup compared to life regen and compared to life on block is under tuned. That or life on block is way over tuned, but flat life on block though good does not feel that good. It does feel mandatory, but it doesn't feel good. Life regen does not feel good either because it is fairly hard to get 100% increased life regen and even when I have that 100% increased life regen, I have what 2.1% - 3% base life regen. Maybe I use an entire skill and support gem for holy descent so I have access to consecrated ground for a juicy 5% life regen, which becomes 7.1% - 8%. Except the duration is bad and the AoE is also bad. At 4,000 life, which is like a best case scenario, I have 80 - 120 life regen, which becomes 160 - 240 life regen with increased life regeneration nodes, with holy descent or some way of getting consecrated ground, my insane life pool gets me 280 - 320 base life regen, which then comes 560 to 640 life regen. OR for no investment, press your life flask and regen nearly the same amount. This is the best case scenario, but it still takes me 10 - 20 seconds to regen my full life pool. This just helps me die slower. But maybe if I have 100% increased life regen, an insane life pool, an insane armour value, max block, increased life regen rate, and a way of getting consecrated ground then all of my ways of dying slower will sum total into infinite defenses and an immortal character. Except, the one design constraint that GGG will always build around is that there should never be an immortal character. This means that even taking no damage nodes and never taking defense, these nodes will still be bad or not good enough to make it worth taking them instead of something, anything else. Besides that, just to get what I described (980% physical damage reduction, life recoup, life on block, increased life regen or life recovery) you are investing SO many points. Instead, why not just get block capped, get life on block, have some life flask charges gained, and press your life flask more often and then with the 10 - 25 points that you just saved get some damage. Then kill stuff faster so that it does not present as much of a threat to your character. Better yet, why not just play basically any ES character, go chaos inoculation so chaos damage does nothing, and then get really low energy shield recharge start time, and then get increased energy shield recharge rate so that you go from 20% base energy shield recharge to 40% or 60% ES regen per second every time you stop taking damage for 1 - 2 seconds. Then if you stop taking damage for 1 second you can regen to full in 2.5 - 5 seconds. Compared to a lot of investment to get to a situation where it still take at best 10 seconds to regen to full life.

Even though it sounds like I am complaining a lot, I am having fun. I really like the warrior class and find some of the spell casters very difficult or near unplayable. I do not dislike this class by any means. I am pointing out that some of the nodes are obvious choices and other nodes are "obvious" duds. I am not entirely sure if they are bad or if there is some secret tech in the layers of these nodes in sum total together that makes them really good - honestly. I just think that having notable passives that synergize with each other better, or as good as the block cap + life on block nodes, or reducing the synergy of those life on block nodes so that the relative power I gain on my character for spending 5 passive skill points for one defensive layer is more in line with the power budget I get from investing 5 passive skill points on a different type of defensive layer. There are also all of the other complaints that players have with armour in general, but I feel like some of the relative power or lack thereof for prefixes on shield and the effective life "regen" from life recoup effects compared to the entire north and east sides of the passive skill tree really demonstrate that there are disparities in the overall design process - like I genuinely feel like one team of devs worked on the warrior's passives area and an entirely different team of devs worked on the monk's passives area. Then even within that it seems like one team of devs "balanced" block nodes and another team of devs "balanced" recoup nodes. I am just left wondering, "is life recoup even worth taking" and most if not all of the time, I think the answer is a resounding no, but I also do not know if maybe it is good under the right conditions. Just right now I can not think of what those conditions are, but I can always think "block cap is good" and "active block + life on block is also really good" and I KNOW that they are good without life recoup. So, why bother spending points on life recoup, life regen, and increased life regen when I already know that life on block and block cap just work and I could spend those points somewhere else on damage or more survivability or getting to block cap.

With all of that being said, to all you warrior haters, it is way better than you think! Even if it is way worse than CI + a huge ES pool. I am still having fun, TY GGG!

Shield passives need some major rework. There's already a downside with the movement penalty and some of the passives don't do it justice. i.e. 4% block chance? bleh this is bad, you can get a jewel socket with more than that. The best they got is like every 1% block chance is 1% attack increase, but it's so far away that you won't be able to get that until later on.

The shield skills themselves aren't useful outside of the shield charge and block. Shield charge just gets interrupted by enemies too in the sense of running into them. You can't plow your way out of crowded situations with it and blocking will just put you on stun if you are surrounded. In a game where movement is key the armor, life, and weight of a shield isn't really beneficial until later in game.

Yeah - I did not even account for the small passive skills. The jewel affixes for block chance are a lot more efficient by comparison, like you said. The major notables are where most of the benefit seems to be, but they still only do something worthwhile when you have a shield that has 40%+ base block chance on it.

I had not even thought about the movement speed penalty imposed by armour and shields. I agree that movement speed is one of the core drivers of defense, but in the end game, it is not as if you can choose to go without a shield. The difference between dual wielding two two-handers vs a two - hander and a shield is insta death vs alright dps. Regardless, it feels a little weird that the Giant's Blood keystone is basically mandatory because the relative dps of a one-handed weapon compared to a two-hander is ridiculous. It doesn't matter if it has 3x attribute requirements, it is practically mandatory to use a shield and to use a two-hander with it right now.

As for the campaign, I don't think you need to be block capped or that you will be able to get block capped until at least act 4, 5, or 6 which is practically coming up on the end game.

Either way, it just seems like for a lot of passive nodes (small and notable), there is either a lack of synergy or a lack of tangible benefits from speccing those points, which means that even though it reads as something that could be cool, you quickly find out that it isn't really providing as much power budget or benefit to your character as simply speccing into something closer or something better or some other class / ascendancy. I am not sure what needs to change but melee needs some love or just something in the form of improved utility or core mechanics of the class.