The interesting discussion to me would be to look at the bottom end of AC.
How much can you mitigate for having a low AC? Like as a pf1 caster you could rely upon mirror image, displacement, and other such defensive abilities instead of ac. Is that still a valid tactic?
If you were playing an unarmored character, would dex be more important than con for survivability?
At what point does your AC hit too low? Would a dex 10 unarmored character (making a 13ac at level 1) be not viable, like playing a con 8 character in pf1?
Mistform Elixir/Blur is equivalent to 2 AC. Invisibility to 5, roughly.
The problem is what happens before you cast these spells... Mountain Monks are notorious for going down at round 1 before their initiative arrives.Also, an important thing to take into account is that a glaring weakness will attract blows. If one character has such a low AC, monsters will often move out of their way to attack him as it's an easy target. Such a character can completely nullify the work of a Cleric/Champion by his sole presence (unless the Cleric/Champion just let him down).
I did find something odd, though. I switched combatant one from using his trident (1-handed +1 striking for 2d8+4) to using his katana 2-handed (2d10+4) against the light shield wielding second combatant still using the trident, and somehow the 2nd combatant wins more often when the 1st combatant is using a weapon with a bigger damage die. I don't know why that would be, but I consistently get the 2nd combatant winning 70% of the time when Combatant 1 deals 2d10+4 and only 57% of the time when combatant 1 deals 2d8+4 instead, which seems like it's moving in the wrong direction.
Hey, I really appreciate you letting me know about that issue. I thought I had already changed this, but it should be fixed now. It was not properly adding the damage on your 2d10+4. It was rolling 2d10 and ignoring the +4 portion.
I had this:
if (x[3] or x[4] or x[5]) == "+":
Needed this:
if x[3] == "+" or x[4] == "+" or x[5] == "+":
Oddly enough, probably the most survivable "min AC" character would be the monk. Since you have the action economy efficiency to "run up to someone, hit them the useful number of times, run away to an inconvenient distance (possibly behind something blocking line of sight)."
Low levels would be rough though.
Had a dragon monk trying to do precisely this in a game as an attempt to deal with AC issues at lower to mid levels.
It worked okay against a lot of enemies and slow melee things were kinda helpless, but he went down in pretty much every encounter that had archers and didn't have something he could hide behind.
During the playtest Michael Sayre had a monk in this vein which apparently outlived the party's champion. Now he's probably got a pretty dang good tactical mastery, but it was interesting to hear about.
During the playtest Michael Sayre had a monk in this vein which apparently outlived the party's champion. Now he's probably got a pretty dang good tactical mastery, but it was interesting to hear about.Oddly enough, probably the most survivable "min AC" character would be the monk. Since you have the action economy efficiency to "run up to someone, hit them the useful number of times, run away to an inconvenient distance (possibly behind something blocking line of sight)."
Low levels would be rough though.
Had a dragon monk trying to do precisely this in a game as an attempt to deal with AC issues at lower to mid levels.
It worked okay against a lot of enemies and slow melee things were kinda helpless, but he went down in pretty much every encounter that had archers and didn't have something he could hide behind.
I believe Michael Sayre's Monk was a dex 10 Dragon Monk, that relied almost entirely on mobility to avoid hits. Lots of move, flurry, move with a high speed.
|
| 1 person marked this as a favorite. |
I believe Michael Sayre's Monk was a dex 10 Dragon Monk, that relied almost entirely on mobility to avoid hits. Lots of move, flurry, move with a high speed.
That sounds correct, and that was under the playtest rules, where Monk AC was worse to start with. I believe he once referred to that character as something like 'a series of bad decisions that kept working out surprisingly well'.
The conclusion was that Monks probably shouldn't all have to be quite that mobile to survive, which was part of why Monk AC got a much needed buff...but it is notable that the mobility allowed his survival even with really bad AC.
"Playing a really bad AC character on purpose and trying to make it work" is really more of a "stress testing the system" issue.
Playtest monks had an issue without going all in on dex, but that's not to say they weren't somewhat survivable with good tactics.
|
I did find something odd, though. I switched combatant one from using his trident (1-handed +1 striking for 2d8+4) to using his katana 2-handed (2d10+4) against the light shield wielding second combatant still using the trident, and somehow the 2nd combatant wins more often when the 1st combatant is using a weapon with a bigger damage die. I don't know why that would be, but I consistently get the 2nd combatant winning 70% of the time when Combatant 1 deals 2d10+4 and only 57% of the time when combatant 1 deals 2d8+4 instead, which seems like it's moving in the wrong direction.Hey, I really appreciate you letting me know about that issue. I thought I had already changed this, but it should be fixed now. It was not properly adding the damage on your 2d10+4. It was rolling 2d10 and ignoring the +4 portion.
I had this:
if (x[3] or x[4] or x[5]) == "+":Needed this:
if x[3] == "+" or x[4] == "+" or x[5] == "+":
The numbers seem to be making more sense now. I had also forgotten to put in the deadly damage for the Katana. so 2-handing the Katana vs trident and shield goes in the Katana's favor.
Of course, what's not accounted for in this is the utility of having a free hand or being able to throw the trident with returning for my Retributive Strikes to get extra attacks off if an enemy doesn't attack me. But all of this is showing me that there is a significant boost for having a regular shield over a buckler.
EDIT: Although here's something curious... raising a shield only grants an advantage vs attacking 3 times if you also shield block.
Trident w/ 3 attacks: 69.9%
Trident w/ 2 attacks and buckler: 30.1%
Trident w/ 3 attacks: 63.8%
Trident w/ 2 attacks and shield: 36.2%
Now, granted, chances are if I've got a shield raised and the enemy is attacking me, I'm going to shield block if I haven't already used retributive strike. But bucklers break easily, so I may not be able to do that more than once or twice in a fight (I wear two bucklers, so could take an action to switch hands. Eventually I'll have the Viking Second Shield feat to be able to pull out another shield or grab something to use as a shield as well).
Trident w/ 3 attacks: 69.9%
Trident w/ 2 attacks and buckler: 30.1%Trident w/ 3 attacks: 63.8%
Trident w/ 2 attacks and shield: 36.2%Now, granted, chances are if I've got a shield raised and the enemy is attacking me, I'm going to shield block if I haven't already used retributive strike. But bucklers break easily, so I may not be able to do that more than once or twice in a...
These numbers do surprise me a bit honestly. I don't know exactly what stats your champion has, but I assume the 3rd strike with the katana would only be hitting roughly on a 20. I'd have assumed a third strike with a 5% chance of success would be inferior to +2 AC from a shield.
Granted when that 5% happens it would possibly swing the fight by itself, especially with deadly in play. Most of the posters in this thread have put a very high value on AC, and I must say I agree with them. But maybe I am putting a bit too much value there.
As for the shield block, the program currently does not track shield HP. So it is slightly inaccurate there. From what I've seen a lot of fights end before a shield would break. Especially in the case of a sturdy shield. Bucklers are another matter, but as your numbers show, using a buckler might not be very beneficial anyway.
Haven't run the numbers, but +1 AC (and +1 Ref) should be better than +1hp/level (and +1 Fort) right?
As there is a "cap" on AC (Item Bonus + Dex = 5 for light and medium armor, 6 for heavy) I think the answer is very much "it depends". Most characters are going to aim for that AC target, and at that point more dex isn't really increasing AC (just moving into lighter armor).
The same isn't true for Con (hp/fort) I believe.
|
|
| 2 people marked this as a favorite. |
If Dex actually provides you with AC it is better than Con, yes. Even on an Elf with a net Con of 10 or 12.
But due to Dex caps that is true only until you hit Dex 20 with no armor proficiency, Dex 16 if you have Light Armor Proficiency, Dex 12 if you have Medium Armor Proficiency, and Dex 10 if you have (and wish to use) Heavy Armor Proficiency.
After that point, the comparison becomes Reflex Save and three Skills (Acrobatics, Thievery, and Stealth) vs. Fortitude Save and HP. I think, at that point, it depends on how much you care about the Skills in question which is better.
So whether Dex is more important than Con varies a lot from character to character based on Class and what kind of armor they have access to. Both are usually more important (at least until you hit diminishing returns when they hit 18) than any other stats except Wisdom and whatever stat you use to attack, so both will almost universally be raised to 18 over the course of a character's career (unless you're going Full Plate and relying on Bulwark...you can ditch Dex then and do okay).
|
|
Trident w/ 3 attacks: 69.9%
Trident w/ 2 attacks and buckler: 30.1%Trident w/ 3 attacks: 63.8%
Trident w/ 2 attacks and shield: 36.2%Now, granted, chances are if I've got a shield raised and the enemy is attacking me, I'm going to shield block if I haven't already used retributive strike. But bucklers break easily, so I may not be able to do that more than once or twice in a...
These numbers do surprise me a bit honestly. I don't know exactly what stats your champion has, but I assume the 3rd strike with the katana would only be hitting roughly on a 20. I'd have assumed a third strike with a 5% chance of success would be inferior to +2 AC from a shield.
Granted when that 5% happens it would possibly swing the fight by itself, especially with deadly in play. Most of the posters in this thread have put a very high value on AC, and I must say I agree with them. But maybe I am putting a bit too much value there.
As for the shield block, the program currently does not track shield HP. So it is slightly inaccurate there. From what I've seen a lot of fights end before a shield would break. Especially in the case of a sturdy shield. Bucklers are another matter, but as your numbers show, using a buckler might not be very beneficial anyway.
Keep in mind, those numbers are without using shield block. They are, however, with the trident, not the Katana, so Deadly isn't factoring in.
I ran them again just to be sure. Here's what I'm using:
AC: 23
HP: 78
Initiative: 8
Attack Bonus: 14
Extra Attacks: 2
MAP: 5
Damage: 2d8+4
Deadly Weapon: 0d0+0
Rounds buffing: 0
Miss chance: 0
Damage reduction: 0
Fortification Chance: 0
Reaction: none
Reaction Value: 0
AC: 25
HP: 78
Initiative: 8
Attack Bonus: 14
Extra Attacks: 1
MAP: 5
Damage: 2d8+4
Deadly Weapon: 0d0+0
Rounds buffing: 0
Miss chance: 0
Damage reduction: 0
Fortification Chance: 0
Reaction: none
Reaction Value: 0
Now, this is fighting a clone of himself. I need to look up the average creature stats for a 5th level encounter. This also doesn't account for tactics, having to move, flanking, etc.
Expected damage for each should be:
Combatant 1: .45x13 + .05x26 + .2x13 + .05x26 + .05x13 = 11.7
Combatant 2: .5x13 + .1*26 + .3x13 + .05x26 = 14.3
That does seem to say that Combatant 2 should be expected to win more often, so I'm not sure where the difference in the methods is coming in.
The tool is also showing combat taking an average of 3.28 rounds (give or take), while going off of expected damage, it would seem like Combatant 1 would need 6.67 rounds to deal 78 damage on average.
Oof, I'm an idiot. Turns out I was resetting the damage stat between rounds instead of between attacks. So if the first attack hit for your character with 3 attacks, the second/third would do the same amount of damage even on a miss.
It should all be set up correctly now. I ran your numbers again and I got:
Combat took an average of 5.65042 rounds.
Your first combatant won 30.506 percent of the time!
Your second combatant won 69.494 percent of the time!
These numbers are a lot more inline what the value I would expect out of 2 AC from raising your shield. Just to be sure I ran the same numbers through the first duel script and got the same win percentage. Oddly the average rounds was 6.27318 but I suspect that is because PF1 has crit confirms and only crits on a 20 vs also critting on 10+ AC which makes the fights slightly longer.
So yet again thank you for bringing the error to my attention.
|
|
Now the results are making more sense. Thanks for fixing that, and for putting this tool together.
So now without using shield block, I get:
Attacking 3 times: 31.0%
Attacking 2 times w/shield: 69.0%
Combat took: 5.6 rounds
Attacking 3 times: 46.2%
Attacking 2 times w/buckler: 53.8%
Combat took: 5.2 rounds
When you have to take 1 action for something else:
Attacking 2 times: 64.5%
Attacking 1 time w/shield: 35.5%
Combat took: 7.1 rounds
Attacking 2 times: 73.1%
Attacking 1 time w/buckler: 26.9%
Combat took: 6.6 rounds
That lines up, and it's interesting to see that AC loses its advantage when each combatant only has two actions left.
Leitner, would you mind sharing some of the logic behind the tool? I've tried it some for simulating an upcoming duel (8 lvl flurry ranger with bear AC vs dragon barbarian) and I'm wondering about the results, or rather how to interpret them and play around with the values. Thanks.
Leitner, would you mind sharing some of the logic behind the tool? I've tried it some for simulating an upcoming duel (8 lvl flurry ranger with bear AC vs dragon barbarian) and I'm wondering about the results, or rather how to interpret them and play around with the values. Thanks.
Certainly, any particular part you'd like to know more about? I somewhat doubt people here would like me to paste the couple hundred lines of code. For your specific duel I unfortunately don't have a good way to represent combat with an animal companion involved.
I would definitely love to add it someday, but all of the additional combatents, targetting priority, team designiations, etc would be a ton of work and is potentially currently beyond our capability. Neither my wife nor I are professional programmers, doing this mostly as a hobby.
Certainly, any particular part you'd like to know more about?
No code please. Just the logic. So the sheet you enter has values, do you just run the steps from there?
D20 on both sides to determine initiative.Winner starts, roll attack vs defense. Determine hit/miss, substract from HP value, possible second attack, third attack. Same for other combatant, and then repeat. Fairly straightforward I guess? Store values until HP < 0 after x rounds. And run that same a couple of times?
Other questions: multiple attacks accept higher values, 4 or 5 for a flurry ranger, @ max MAP -4? Or does the MAP keep counting?
Adding the bear support option is possible, by the way. Instead of 2d6+6 + 2d8, I just changed that to 4d6+7, which is around the same. There's also a primary and secondary weapon eith different dice d8/d6, so that will just be averages.
No code please. Just the logic. So the sheet you enter has values, do you just run the steps from there?
Not sure I follow your meaning. The program takes all the stats for each character and puts them into a combatent class.
D20 on both sides to determine initiative.
Winner starts, roll attack vs defense. Determine hit/miss, substract from HP value, possible second attack, third attack. Same for other combatant, and then repeat. Fairly straightforward I guess? Store values until HP < 0 after x rounds. And run that same a couple of times?
Yep. Each side rolls initative(rerolls ties). And then take turns attacking each other till one player has 0 hp. This can result in a tie via something like iron command(At least I think the reaction rules work like this). It runs through the fight 50,000 times. So the numbers you get will vary slightly, but mostly by a tenth of a percent or so.
Other questions: multiple attacks accept higher values, 4 or 5 for a flurry ranger, @ max MAP -4? Or does the MAP keep counting?
So for something like a flurry hunter using an agile weapon you would put 2 for your MAP. If you are attacking 3 times it would roll your first attack normally, your second at a -2, and third at a -4(Extra attack #*MAP penalty). I could add a value to cap your MAP in cases where players get an extra attack action from class feats or such, but currently it would just keep subtracting. If you told the program you had a MAP of 2 and 50 attacks it would try rolling all 50 adding a further 2 penalty to each roll.
Adding the bear support option is possible, by the way. Instead of 2d6+6 + 2d8, I just changed that to 4d6+7, which is around the same. There's also a primary and secondary weapon eith different dice d8/d6, so that will just be averages.
Something like that definitely could work. I used the idea for testing TWF rogues in the PF1 duel version. It gets a bit iffy with stuff like HP, AC, etc. But might still work to approximate your results.
Ok, clear what it does. Tried some combi's, don't have a way to verify if the numbers are anywhere near correct. For flurry rangers, the MAP increase can actually hurt, as from lvl 5 up, there is an easy way to get 4 attacks at decent MAP with haste. Or 4 attacks with an action for support of the AC.
Shield block and a d8 weapon works well for a barbarian in the runs I did. More chance of winning as the dice doesn't add much to the total damage dealt and the AC gain plus block does prevent a lot of damage.
While having a shield can significantly reduce your possible damage in a turn, generally speaking you would be better off using that third action to do something else.
Moving away, even just a Step, completely eliminates one attack action. If it is a slow monster, or you are have above average speed, it could negate two or more whole attacks. Consider a Tiger Stance Monk. Come in Strike Twice with flurry and then step away. Now the Creature can move 10 feet to reach you, either as two Steps or a Stride which provokes an AoO, OR ignore the squishy Monk and strike the high AC Champion or the Fighter with Reactive Shield.
I don't have time to run all the numbers, but it would be interesting to see the survivability of the rogue run above raising a shield vs just striking 3 times OR using an action to feint instead, or just having the Dodge reaction.
While having a shield can significantly reduce your possible damage in a turn, generally speaking you would be better off using that third action to do something else.
Moving away, even just a Step, completely eliminates one attack action. If it is a slow monster, or you are have above average speed, it could negate two or more whole attacks. Consider a Tiger Stance Monk. Come in Strike Twice with flurry and then step away. Now the Creature can move 10 feet to reach you, either as two Steps or a Stride which provokes an AoO, OR ignore the squishy Monk and strike the high AC Champion or the Fighter with Reactive Shield.
I don't have time to run all the numbers, but it would be interesting to see the survivability of the rogue run above raising a shield vs just striking 3 times OR using an action to feint instead, or just having the Dodge reaction.
Yes, but moving away and negating one action is eliminating your foes WORST attack, whereas +2 AC is reducing the quality of all their attacks, including their best.
I think that if you do the math, you'll find that Forcing an opponent to make 2 attacks instead of 3 is rarely better than Raising a Shield.
Ok, clear what it does. Tried some combi's, don't have a way to verify if the numbers are anywhere near correct. For flurry rangers, the MAP increase can actually hurt, as from lvl 5 up, there is an easy way to get 4 attacks at decent MAP with haste. Or 4 attacks with an action for support of the AC.
Well, hopefully the numbers are accurate. If your instinct tells you they are wrong please let me know. I've tested various portions of the code quite a bit. And FuriousThrune has helped point out a couple of issues that are now fixed.
Capping the MAP would be easy enough to do for the program and I will probably add that feature sometime. But unfortunately dealing with every permutation of attack possibilities would be impractical without forcing people to manually input each attack/bonus.
A fighter using double-slice for the first two actions and then a shield block with the 3rd could be represented with 1 additional attack and 0 map. But as soon as you were using a 3rd/hasted action to attack this no longer works.
Adding a field for each attack/bonus could be done, but I was somewhat assuming people would prefer something that works for most situations and is less tedious. If that is not the preference let me know.
Yes, but moving away and negating one action is eliminating your foes WORST attack, whereas +2 AC is reducing the quality of all their attacks, including their best.I think that if you do the math, you'll find that Forcing an opponent to make 2 attacks instead of 3 is rarely better than Raising a Shield.
I ran a level 2 goblin rogue "Breadcrumbs" taken off of Ravingdork's character emporium against a CR 2 Giant badger.
Attacking twice and raising a shield led to the goblin rogue winning 26% of the time vs the badger.
Attacking twice and then running lowered the winning chance down to 16%, however, if I give the rogue fleet so one move action takes the badger two to catch up, that increases the win % for the rogue up to 39%.
EDIT: Shield block was not used here. But presumably if you took that instead of Fleet, it would likewise increase your survival chance.
Yes, but moving away and negating one action is eliminating your foes WORST attack, whereas +2 AC is reducing the quality of all their attacks, including their best.I think that if you do the math, you'll find that Forcing an opponent to make 2 attacks instead of 3 is rarely better than Raising a Shield.
I ran a level 2 goblin rogue "Breadcrumbs" taken off of Ravingdork's character emporium against a CR 2 Giant badger.
Attacking twice and raising a shield led to the goblin rogue winning 26% of the time vs the badger.
Attacking twice and then running lowered the winning chance down to 16%, however, if I give the rogue fleet so one move action takes the badger two to catch up, that increases the win % for the rogue up to 39%.
Sounds about as expected. If you can eliminate multiple actions by spending one, you're doing pretty well. But eliminating only your opponents worst attack for one action isn't worth it.
The other possibility is that moving away may eliminate none of the opponent's attacks. But redirect them to an ally who may or may not be in a better position to handle it.
On the plus side, it may prevent the enemy from using a great 3 action ability.
Hitting and moving can be a good idea in plenty of circumstances, but trying to tie it down to being always better or worse than using a shield is a pretty gross oversimplification. Having the shield, and therefore the option, is always going to have some value.
Yes, but moving away and negating one action is eliminating your foes WORST attack, whereas +2 AC is reducing the quality of all their attacks, including their best.
I think that if you do the math, you'll find that Forcing an opponent to make 2 attacks instead of 3 is rarely better than Raising a Shield.
Is it though?
Using the numbers run above by voideternal-
No shield, third attack:
Hits on 17, Crit on 20
Average 5 damage
With shield, third attack:
Hits on 19, Crit on 20
Average 3 damage
Total damage 2 attacks without shield: 28
Total damage 3 attacks without shield: 33
Total damage 3 attacks with shield: 25
Difference between moving away and keeping a shield up is 3 hp on average, but there is also a much higher chance of getting an "unlucky" round and the creature rolling a 20 on the third attack.
And as HammerJack points out, it might redirect the creature to a harder to hit target, thus reducing your damage to 0.
| 1 person marked this as a favorite. |
Just going off of experience here. I find that you pretty much want to optimize AC while still doing the things your character was designed to do. You probably do not want to carry a shield on a rogue, ranger, or barbarian.
I do think the mitigation from a higher Dexterity is worth quite a bit more than extra hit points from Constitution particularly given that we get the whole die worth of hp now.
Thanks for all the informative answers ;)
However, my question was more in the line of: "in actual play, without this +2 to AC, can you withstand the onslaught of a same level monster ?"
In other words, you're a rogue level 10 without a shield, you're in the reach of a level 10 monster, what are your odds of surviving the turn (without a Champion helping, of course) ? And even if you survive, will you have to move away next turn and/or need a heal spell, thus being more a liability than an asset ?
I appreciate that a game should be about what I want to play, not what I have to play, but I also don't want to drag the party down.
Yes you can, but sometimes you will be unlucky and need to recover.
What is the rest of the party like? If there are 2 melee specialists in a party of 4-5, then you are fine. If its 0 or 1 then you will need to be tougher (barring specialist parties).Take Nimble Dodge, for the times that you need it.
While having a shield can significantly reduce your possible damage in a turn, generally speaking you would be better off using that third action to do something else.
Moving away, even just a Step, completely eliminates one attack action. If it is a slow monster, or you are have above average speed, it could negate two or more whole attacks. Consider a Tiger Stance Monk. Come in Strike Twice with flurry and then step away. Now the Creature can move 10 feet to reach you, either as two Steps or a Stride which provokes an AoO, OR ignore the squishy Monk and strike the high AC Champion or the Fighter with Reactive Shield.
Curiously, in the short introductive adventure I'm making for my friends, using pregenerated characters, Tiger style Monk and Liberator Champion are the group frontliners.
The Monk is all but squishy: as an Unbreakable Goblin, he ends up having both more AC and more HP than the Champion (at level 1, the lack of a proper armor hurts!), and he's got a shield to raise too. In the combat simulations I made, the Champion was more likely to go down than the Monk even without the latter using evasive maneuvers like stepping away; actually, the best tactic seemed to be abusing Goblin Scuttle + Tiger Stance to get into flanking position for free.
The main advantage of moving, or other action denial, lies more in not letting the opponent use attacks that require multiple actions. For example, if you're fighting some horrid tentacle monster, their optimal round probably looks something like: Attack, Grab, Constrict. If you make them move, they won't be able to constrict which in turn gives you the possibility to get out of the grapple first.Yes, but moving away and negating one action is eliminating your foes WORST attack, whereas +2 AC is reducing the quality of all their attacks, including their best.
I think that if you do the math, you'll find that Forcing an opponent to make 2 attacks instead of 3 is rarely better than Raising a Shield.
Or Breath weapon, Strike, Or Spell, Strike.