67 réponses à ce sujet

[Theramond]

LynxAQ wrote...

I wouldnt stop so quickly. I have no doubt that the graph making, research etc took some time but the data on the graphs are meaningless and leave out so many variables that they can only be taken with a grain of salt.

I both agree, and disagree. Variables must be controlled, otherwise you would recieve highly inconsistent data that could not be interpreted. While it should be taken with a grain of salt, it is not useless information.

LynxAQ wrote...You have gone the wrong way completely with your tests. Which stat is the best to increase. Your tests say nothing and answer nothing of that question.

I disagree. Take a look at the results of the subsequent test. It supports your supposition that an all dex/def build is superior. There appears to be a cumulative effect proportional to the amount of dexterity attributed to the experimental builds, which only appears to be noticeable when defense reaches a certain threshold. In the subsequent experiment, a +50 defense buff clearly demonstrated this effect, which if taken into consideration, makes the all-dexterity build far superior over the dex-con build. A dex-con build is only superior when defense is below a certain threshold, which might be the reason why there is so much confusion as to which build is better. A player that does not take their defensive score into consideration, might experience the dex-con build as an undifferentiated or slightly superior option in comparison with the all-dex build.

A sidenote: with each additional defense score, this cumulative effect eventually turns into a run-away greenhouse effect with longevity far surpassing the designs of the experiment.

Chart demonstrating this effect:



Predictably, if I were to test the accumulation of dexterity versus constitution over time (starting at 10), it would probably look like the chart above, based on the previous experimental data. Health would see a steady increase, while defense would resemble a parabolic curve.

In addition, it appears that armor also fulfills its greater potential as defense increases. I believe the logic is as follows: with a greater evasiveness, abilities like "sunder armor" are more effectively mitigated, allowing the PC to maintain the benefits of armor throughout the fight more frequently. If so, this would be another bonus to going all-dex instead of using a dex-con build.

Modifié par Theramond, 31 décembre 2009 - 10:33 .

[WillieStyle]

A sidenote: with each additional defense score, this cumulative effect eventually turns into a run-away greenhouse effect with longevity far surpassing the designs of the experiment.

This is the key point here. Defence enjoys increasing returns with itself. The more defence you have, the more valuable defence becomes until you become unhitable. It's why many games either cap avoidance or encode diminishing returns to defence.



Some minor tweaks that would improve tanking in this game:

-Reduce the dex to defence conversion. Ideally it would be nonlinear so that it starts out at something like 1 dex to 1 defence but diminishes based on how much defence you already have.

-Make armor reduce a percentage of damage recieved. This would make armor scale better against tough enemies.

-Improve Con. This is a tough one because it's hard to balance hitpoints versus avoidance. Hitpoints are only important if a boss can 2 shot you or you have a hard time being healed. I'm not sure adding more Ser Cauthraines to the game would make it more popular (I would like it but I'm not sure others would); and healing is so easy to come by that coupled with the pause button, it's hard to see how Con can be valuabe against anything but instagibs.

-Change offensive tank talents. Tanks don't need 3 different ways to hit a single mob with a shield. They need more ways to deal small amounts of damage but large amounts of threat to large numbers of mobs around them. They need ways to apply AoE debuffs *cough*ThunderClap*cough*. They need ways to close range with ranged enemies either with a ranged stun or grab. Basically, the defensive portions of the champion/reaver/templar talents are more what sword&board talents should be.

-Fix threat. Taunt is too powerful and everything else is relatively useless. If sword and shield tanks had more AoE abilities, you could nerf taunt while still making tanking viable and interactive. This would also make a Shield tank actually better than a dual wield tank. Right now, you're probably better off dual-wielding daggers; Far more damage, almost the same damage mitigation.

[Lord Phoebus]

Another factor that hasn't been mentioned in the dex/con debate is healing. The dex tank is always going to take damage at a slower rate than the con tank, but healing spells/poultices will heal the same amount of damage to each, making it easier for a healer to keep a dex tank topped up than a con tank as the battle continues.

[shree420]

Good suggestions.

WillieStyle wrote...
This is the key point here. Defence enjoys increasing returns with itself. The more defence you have, the more valuable defence becomes until you become unhitable. It's why many games either cap avoidance or encode diminishing returns to defence.

Some minor tweaks that would improve tanking in this game:
-Reduce the dex to defence conversion. Ideally it would be nonlinear so that it starts out at something like 1 dex to 1 defence but diminishes based on how much defence you already have.

I don't think we need to nonlinearize this; it could be taken care of by the enemy leveling algorithm. Better attack values for tougher enemies.

WillieStyle wrote...
-Make armor reduce a percentage of damage recieved. This would make armor scale better against tough enemies.

That's an interesting suggestion. Armor is currently useless against bosses and elites.

WillieStyle wrote...
-Improve Con. This is a tough one because it's hard to balance hitpoints versus avoidance. Hitpoints are only important if a boss can 2 shot you or you have a hard time being healed. I'm not sure adding more Ser Cauthraines to the game would make it more popular (I would like it but I'm not sure others would); and healing is so easy to come by that coupled with the pause button, it's hard to see how Con can be valuabe against anything but instagibs.

I believe the whole point of CON should not be increasing HPs - that should be a function of leveling alone - but increasing the physical resistance. Sort of a natural armor that reduces damage. In real life, we don't say that a tough guy has more health than a wimp; it's just that the musculature and hardiness prevent damage from soaking through. That's what would make a CON build really viable. A bonus to health and stamina regen is another valid option.

Fixing the CON build by simply increasing the HP each point gives isn't valid IMO because it requires the tank - who is now getting hit more - to quaff potions/ get mage to heal/ and/or wait after the battle a long time to recover lost HP.

WillieStyle wrote...
-Change offensive tank talents. Tanks don't need 3 different ways to hit a single mob with a shield. They need more ways to deal small amounts of damage but large amounts of threat to large numbers of mobs around them. They need ways to apply AoE debuffs *cough*ThunderClap*cough*. They need ways to close range with ranged enemies either with a ranged stun or grab. Basically, the defensive portions of the champion/reaver/templar talents are more what sword&board talents should be.

Agree. S&S rely on their specialization talents for AoE control, unlike DW and 2h. To be fair to Bioware, I can't imagine a shield usage talent that could do AoE damage/stun; I got a vision of the warrior swinging his shield around. Bonk! Bonk!

Maybe a simultaneous bash on one enemy and a strike at another?

If I'm not mistaken, ( and I'm no grognard, so please excuse any ignorance) historically the advantage of a sword & shield route was only obvious in a phalanx when you had shield-mates on either side. Your shield protected the person on your right, and you were protected in turn by the shield on your left.

WillieStyle wrote...
-Fix threat. Taunt is too powerful and everything else is relatively useless. If sword and shield tanks had more AoE abilities, you could nerf taunt while still making tanking viable and interactive. This would also make a Shield tank actually better than a dual wield tank. Right now, you're probably better off dual-wielding daggers; Far more damage, almost the same damage mitigation.

Also true. Tanks deal too little damage to keep threat unless you use Taunt/ end the battle with spell damage before they reach the mages.

[LynxAQ]

I agree with basically what WillieStyle (Haha, the name 3> ) said.



Threat needs to be more interactive, at the moment, all you do is press taunt and everything sticks to you like glue until they or you are dead. The entire threat system (if you can call it that) is very static and boring.



I also find it odd that you get a 1 to 1 conversion for agility to defense but get a 1 to 0.5 conversion for strength to attack. It doesnt make sense. But it is a single player game so in a way it doesnt matter that much.



I do believe there should be a hard cap on avoidance. I personally dont believe you should go above 75 or 80%. This would then indirectly make constitution a viable stat for tanks at least. But constitution needs to be buffed, cause other than someone playing a full time Blood Mage, there is no reason to put anything in constitution for any class...



I agree that the shield and weopon offensive skills should have been debuffer abilities. Like one reduces armour, another reduces defense etc, make them not stackable so you have to choose which was better to apply to the enemy etc.

[menasure]

some very good hints in here for humanoid tanks. i wonder if shale or dog can work as good while build as a more con based tank. they get more health, though they might lack a high threat.
as for the threat generation: there's always a tendency for mobs to hit whatever hits them first ... problematic for mages if they AoE in the first second of a fight because threaten will not easily persuade mobs to go for the tank instead even if it's a heavily armored one. taunt is more reliable though it works with the same principle of adding a value to the threat. the best way is still to attack with the tank first imo and if that fails the interrupt skills combined with threat generating skills might tip the threat balance towards the tank.
it might be boring sometimes if your threat generation really works all the time but i still prefer that compared to having to run with mage or healer just to stay alive.

Modifié par menasure, 31 décembre 2009 - 04:34 .

[Hulk Hsieh]

DragoonKain3 wrote...
26 base STR
all the rest in DEX (+50)
Fade (+4)
Warrior bonus (+4)
Human (+1)
Level 20 (+20)
Helm of Honnleath (+2)
Dalish gloves/boots (+5)
Felon's Coat (+15)
Andruil's Blessing (+2)
Magister's Shield (+6)
Key to the City (+2)
Lifegiver
Rose of Thorns (+2)
Champion's Shield (+15)
Shield Defense on (+15)
Rally on (+10)

It seems only 30 DEF of them comes form the shield...which is exactly the same as the archery DEF bonus.
Maybe an Archer Tank is also a viable option. No anti-knockdown anti-flank though.

[Haplose]

menasure wrote...

some very good hints in here for humanoid tanks. i wonder if shale or dog can work as good while build as a more con based tank. they get more health, though they might lack a high threat.
as for the threat generation: there's always a tendency for mobs to hit whatever hits them first ... problematic for mages if they AoE in the first second of a fight because threaten will not easily persuade mobs to go for the tank instead even if it's a heavily armored one. taunt is more reliable though it works with the same principle of adding a value to the threat. the best way is still to attack with the tank first imo and if that fails the interrupt skills combined with threat generating skills might tip the threat balance towards the tank.
it might be boring sometimes if your threat generation really works all the time but i still prefer that compared to having to run with mage or healer just to stay alive.

Actually Shale has no issues with Threat generation. He has full access to General (non-weapon style) Warrior talents, including Taunt and Threat. On top of those his "Stoneheart" stance turns him into a huge aggro magnet, with each talent in the tree adding to the Threat he generates - both passively and actively, when he uses those talents.

So he's a kind of Supertank - or rather would be, if his avoidance wasn't so sucky. And at higher levels avoidance is much better then lots of Con and soak.

[tranj84cl]

Regarding the thread title, and OP, I find it a bit arrogant to call this the 'Art of Tanking' when it seems to be more of a 'How I built My Tank' thread. Opposing schools of thought are hardly addressed outside of saying, 'this can be changed depending on how you play.'



The rest of the thread is much more worthy of the original intent. I am in the Dex camp, for reasons listed by other people. I've found 150-160 defense makes you pretty much unhittable to the degree that 100 def + 150 hp pales in comparison. Of course, going the avoidance route means you have to build for it. Magic is your weakness, attaining 100% spell resistance isn't that hard provided you *built for it*.

[shree420]

menasure wrote...

some very good hints in here for humanoid tanks. i wonder if shale or dog can work as good while build as a more con based tank. they get more health, though they might lack a high threat.
as for the threat generation: there's always a tendency for mobs to hit whatever hits them first ... problematic for mages if they AoE in the first second of a fight because threaten will not easily persuade mobs to go for the tank instead even if it's a heavily armored one. taunt is more reliable though it works with the same principle of adding a value to the threat. the best way is still to attack with the tank first imo and if that fails the interrupt skills combined with threat generating skills might tip the threat balance towards the tank.
it might be boring sometimes if your threat generation really works all the time but i still prefer that compared to having to run with mage or healer just to stay alive.

Taunt is all you need, and it works just fine to get everybody's attention - sometimes too well! The only issue is if the enemies you've just Blizzarded with your mage come singly rather than as a group, since you can't Taunt repeatedly. In which case you'd want to set up a Glyph of Repulsion or Mind Blast stuns while your tank builds aggro.

I think Mind Blast has the interesting property of resetting threat
on enemies. So when enemies come out of MB, if they're being taunted or
attacked by your tank, they'll focus on the tank.

Shale's a good tank because she can soak damage and generate threat - she appears to have been built as a tank first and foremost. Dog is NOT a good tank - he doesn't have the damage absorption abilities Shale does, you can't give him any armor save any bonuses from warpaint and collar, nor can he sustain aggro as he doesn't have the Taunt equivalent. Use him as an off-tank and he works very well.

[LynxAQ]

Depends what platform you play on. On the consoles you could kinda use the dog for tanking at lower difficulty levels because the dog gets the warrior talents. On the PC the dog does not get the warrior talents and therefore cannot get taunt and therefore definately not a tank candidate.



Not sure why dog doesnt get warrior talents on PC but does on the console.

[shree420]

LynxAQ wrote...

Depends what platform you play on. On the consoles you could kinda use the dog for tanking at lower difficulty levels because the dog gets the warrior talents. On the PC the dog does not get the warrior talents and therefore cannot get taunt and therefore definately not a tank candidate.

Not sure why dog doesnt get warrior talents on PC but does on the console.

I forgot that - I play it on the PC. And you're right, Dog should have gotten his own special tree, if not Warrior. And he does function quite well as a tank on lower difficulty levels. As you get better gear for Alistair( or get Shale with her special tanking skills) Dog fades as a tank in viability.

I plan to use him as an off-tank in my next playthrough as a Rogue with Alistair(tank) and Morrigan.

Re: Taunt, that would help Dog build aggro, but it doesn't solve the armor problem. He's no pansy, but using him as tank can't work too well for those reasons.

[menasure]

Lord Phoebus wrote...

Another factor that hasn't been mentioned in the dex/con debate is healing. The dex tank is always going to take damage at a slower rate than the con tank, but healing spells/poultices will heal the same amount of damage to each, making it easier for a healer to keep a dex tank topped up than a con tank as the battle continues.

as far as i know the magic stat is meant to alter the level of healing a character receives. so how viable would a combo of healer with a high magic stat tank be, does anyone have a clue about this?

Modifié par menasure, 31 décembre 2009 - 09:33 .

[WillieStyle]

menasure wrote...

as far as i know the magic stat is meant to alter the level of healing a character receives. so how viable would a combo of healer with a high magic stat tank be, does anyone have a clue about this?

This is incorrect. A dev claimed this once, but subsequently corrected himself.  Magic affects healing from potions and affects the strength of healing spells CAST by the character.  Thankfully, it does not affect healing spells recieved by the character.  The only effective magic tank is an arcane warrior.

[maudlin27]

Great graphs, another factor to consider in surviability though is healing. Defence makes healing MUCH more useful than con, since the healing effect will have a much higher proportional effect with high def-low health than high health-low def.



Disappointed in the original post for this thread, which isn't an 'art to tanking' or tanking guide, but rather 'look at this tank I built'. Fortunately the graphs/figures provided later on help give an insight into the type of tank build to look at

[dkjestrup]

Healing should be like pokemon, a percentage heal, rather than a number.



Con should increase armor and physical resistance. Dex tanks should be hit for 100 damage every 50 attacks. Con tanks should be attacked for 5 damage every 2-3 attacks or so.

[Erakleitos]

My next playtrough will be an all naked party with all points into con

[Theramond]

maudlin27 wrote...

Great graphs, another factor to consider in surviability though is healing. Defence makes healing MUCH more useful than con, since the healing effect will have a much higher proportional effect with high def-low health than high health-low def.

Hmmm, that's interesting. I never thought of that.