trying to figure out when GWM feat is worth it

[Chugger]

I realize a lot of players take a great deal of emotional stock in a "big hit" - "omg I just did 23 pts damage in one blow at lvl 1!" Craps players I've known also put emotional stock in certain things around the craps table, and last I saw the casinos are still paying their light bills, ahem. What this means is if a character is in reality pulling a whole ton of swish but occasionally landing a monster hit (say a HuV w/ GWM and a 2h sword) but another is hitting far more often but for less damage - which has a higher dpr and is contributing more? I realize there are situational uses for GWM, like the war cleric's limited +10 to hit channeling feature - a meleer who has bardic insp, bless and the BM's Presicion all lined up for a special fight. And yes, that boss fight is a special thing where burning it down fast is very good. So what I'm talking about here isn't all cut and dried - there is a big variability or subjective element to it.

Okay, let's say the monster has AC 11 and I have a standard +5 to hit and +3 to damage w/ my 2h greatsword. Average damage is 2d6 + 3 normally or 7+3=10 and with GWM feat is 20.

W/out feat I will on average hit AC 11 15 out of 20 times. That's 150 total damage in a cycle and an average dpr of 7.5, if I'm doing this right (pls help me if I'm not - I'm so rusty at this it isn't funny).

Okay, with the feat I will on average hit AC 11 only 10 out of 20 times. That's 200 total damage in a cycle and an average dpr of 10, if I'm doing this right. Okay, worth it.

Now let's up the AC to 16. At +5 to hit the featless guy will hit 10 out of 20 times for a total pool of 100 and a dpr of only 5.

The feat guy only hits AC 16 5 out of 20 times, for a total pool of 100 and a dpr of only 5.

Whoah. Fighting AC 16 two lvl 1 fighters, one with GWM and one with no GWM feat, do the _same DPR_, if I"m doing this right.

That's not good. Well, let's say I always knew GWM could have a problem like this - lots and lots of missing or swish. So I talk my friend into playing a wolf barb (and now we're lvl 3) - and I have _advantage_ on every hit (cuz I'm always gonna be 5 feet from my wolf-barb bud - we're like conjoined dudes or sumpin).

With advantage the GWM guy gets two dice. So the first dice cycle of 20 he will hit only 5 times. And the second dice cycle he will hit only 5 times, but 1/4 of them will be wasted and paired with a first-die-hit. So if I'm doing this right, he'll get a pool of 100 from the first and 75 from the second die - for a total of 175, or an average DPR of 8.75.

But let's say the no-feat guy also has a wolf barb friend and "always" attacks at advantage. He has a 150 pool or an average DPR of 7.5.

So I'm seeing, if I'm doing this right, a significant decline in value for GWM feat the higher the AC goes. Yes, we can do much better than +5 to hit, but it's hard to get those consistently.

Let's see if we can do AC 18 at lvl 5 (ignoring extra attacks) with +7 to hit. The char with no GWM hits half the time. We're back to a 100 pool or a 5 dpr (for one attack per round at this point).

At AC 18 taking a -5 from the GWM feat means hitting 5 times or again 100 - a dpr of 5. Same as above.

Now at lvl 5 and up, when we have typically +7 to hit, we can be fighting all sorts of things. Some giant-type creatures (hill, ettin, ogre) have very bad ACs. Frost Gi is ac 15 but is a way off (fire giant has higher ac). Orc and gnoll and the like have "champions" who wear armor and may have 15 ac or more.

Okay, where GWM might really shines maybe is with a Battle Master fighter. He has precision 4 times per rest, iirc, which is plus 1-8 after you see the d20 roll. Bardic insp is plus 1d6, after you see the d20 roll, too. Bless - not too hard to get - any cleric or pal can cast - is plus 1d4 and comes with the 1d20 roll. So this fighter can see the d20 roll and on average add to it 10.5 to hit. If he's rolled a 5 on the d20 and needs to hit ac15, there's a good chance he'll do it - probably worth a risk. If he's rolled a 3, maybe not. Maybe better to say the insp and sup dice (a BM point so to speak) for the next attack. With all that plus his barb friend (or some way of getting advantage), maybe he can make GWM shine.

I'm starting to sense that GWM may only be worth the risk in certain fights, especially low AC high-HP encounters - or ones with lots of pluses to the roll in ones hip pocket, ready to use. You can do a lot of other things with that feat - there are other great feats out there. But the GWM can be huV to get the feat at lvl 1 then instead of an asi at 4 get Lucky and...yeah, but skipping an asi, there is a price for that. Lucky is great, but so is getting a core stat up 2 to 18. Not being capped out at 20 in a main stat by a certain lvl hurts. Of course, this is assuming I'm getting this right. And a lot of our fights, especially the early ones, are with low AC creatures (zombies - a gwm wants to fight _zombies_). I'm guessing I got some of this wrong, however - anyone have a good link on this topic? I've found some but they aren't really on the mark - maybe I've had bad luck or need to learn to think up better search terms. But if you can help me that would be great.

(edit, just caught a mistake - I left out GWM's other feature, that you can attack again if you crit or zero your target. For simplicity's sake I'm going to have to ignore that for now. The first will give a bit more DPR, esp if you're a fighter champion - the second (dropping a target to zero) is highly subjective and will vary a lot - just know the numbers I gave for GWM above are at least a bit too low)

[Armored Walrus]

Pretty sure if you search a little there are multiple threads on this and other forums where all of that math has been done already, and the question you're asking has been answered.

[Sans.]

http://www.giantitp.com/forums/showthread.php?373572-GWM-Reference-Table

[qube]

considering,
X in {0,20}, the hit chance; and B the avr base damage
GWM is useful if:
(X/20) * B < ((X-5)/20) * (B+10)
(XB/20) < ((X-5) * (B+10) /20)
XB < (X-5)(B+10)
XB < (X-5)B+(X-5)10
XB < BX-5B+10X-50
0 < -5B+10X-50
0 < -B+2X-10
10+B < 2X
In your two examples (and a third one),
AC 11, +5 attack (15/20 hits, so X = 15), 2d6 + 3(B=10) damage.
20 < 30
GWM is useful
AC 16, +5 attack (10/20 hits, so X = 10), 2d6 + 3(B=10) damage.
20 = 20
GWM makes little difference
AC 17, +5 attack (9/20 hits, so X = 9), 2d6 + 3(B=10) damage.
20 > 18
no GWM here

-------------------------------------
Advantage is a bit more tricky ... actually, no, it's mathematical fun:
On 400 rolls, there's 1 '1', 3 '2's (namely {1,2}, {2,1}, and {2,2}, 5 '3's,
... basically 2N-1 accounts of N.
On 400 rolls, there's 1 '1 or less', 4 '2 or less', 9 '3 or less',
... basically N² accounts of 'N or less'.
So, if only a 20 hits (X=1), there are 19² misses, so (400-19²)/400 = 39/400 hit chances.
if only a 19 or 20 hits (X=2), (400-18²)/400 = 76/400 hit chances.


Now, dumping this into the damage compare formula
((400 - (20-X)² )/400) * B < ((400 - (20-(X-5) )² )/400) * (B+10)
(400 - (20-X)² )*B/400 < ((400 - (20-(X-5) )² ) * (B+10)) /400
400 - (20-X)² < ((400 - (20-(X-5) )² ) * (B+10))
400 - (20-X)² < ((400 - (25-X)² ) * (B+10))
400 - (400-40X+X²) < ((400 - (625-50X+X²) ) * (B+10))
40X-X² < ((-225+50X-X²) * (B+10))
40X-X² < ((-225+50X-X²)B+(-225+50X-X²)10)
40X-X² < -225B+50BX-BX² -2250+500X-10X²
0 < -225B +50BX - BX² -2250 + 460X - 9X²
which is a much uglier formula, but it would mean

AC 11, +5 attack (15/20 hits, so X = 15), 2d6 + 3(B=10) damage.
0 < 5625, so GWM is useful
AC 16, +5 attack (10/20 hits, so X = 10), 2d6 + 3(B=10) damage.
0 < 3200, so GWM is useful
AC 21, +5 attack (10/20 hits, so X = 5), 2d6 + 3(B=10) damage.
0 > -175, so GWM is no longer useful

I ran that formula through Excell, thinking I'd get some special curve, but the difference when to use and when not to use GWM seems to be between when X > 5
(so, somebody please verify my maths)

[Sirdar]

Desamir has made a comprehensible analysis that you may like:

http://www.giantitp.com/forums/showthread.php?472938-Great-Weapon-Mastery-How-to-5-10-Like-a-Pro&highlight=Gwm+like+Pro+desamir

[smcmike]

As others have said, the math has been done in great detail by others.

At the table, though, I think it's best to just have a rough rule of thumb. So:

Do Not Use GMW if:
- the enemy is so weak that 1 hit will almost certainly kill it
- the enemy has a noticeably high AC

Do Use GWM
- just about any other scenario. Also, try to get advantage.
-Actually, just play a barbarian, and never stop using it.

[Chugger]

Thanks everyone. Great thoughts - really good links! I simply was not finding these links - other ones that were wonky and not really on the mark - these are spot on! Thanks.

I'm seeing that my intuitive suspicion was correct, that GWM or SS is only situationally useful. I'm guessing that if you take it, say, at lvl 1 as a huV it will become less and less useful as lvls go up and monster ACs tend to go way up - but of course there are low and low-ish AC monsters.

I do appreciate the help with the formulas and numbers a LOT - been so many years since I took statistics, and I was never "fluent" at it. I can discuss cardinality, for example, and why Cantor's naive theories broke down and why Zermelo/Frankel sort of "took over" - I can even discuss G's IT to some extent - but my brain just doesn't like the rigor I require to actually do statistics because my instinct is so strong - it wants to take over and just guess - and I have to constantly fight it away and focus on the part of my mind that does statistics - and it wears me out (and yes, I'm a terrible programmer!) ... all of which is probably very good for "at the table" issues, as some brought up - because a mostly-good instinct will serve a player well at a table - and most dnd situations have some tolerance for mistake (mistakes inevitably come with a reliance on instinct - you are hoping the speed at which you came up with mostly right and sometimes inspired solutions overcomes that I suppose). You can see me trying to guess my way through with a shorthand in the OP - but the real approach is nice. Thanks again.

So, going to "at the table" situations, I might be able to memorize the bulk of these math tables on which ac to gwm/ss when to attack normally, but I think most players can't. My instinct tells me that at low levels (at least) GWM/ss would be useful _if_ one can count on rests to bring back certain "big plus to hit" features - so that a character can have one or two rounds of burst power when it is really needed.

A BM fighter HuV (with SS) could have +7 base to hit at lvl 5 (dx +4, pb +3, iirc), +2 from arch skill, and 4 times per rest a plus 1-8 from precision usable after seeing the d20 roll. A 2-lvl war cleric dip = bless (plus 1-4) and a +10 feature usable once per rest usable after seeing the d20 roll (to get the fighter lvl 5 features obviously he'd have to dip for lvls 6 and 7, if that's worth it). A war cleric in the party could bless him and grant him a +10 with his lvl 6 feature, instead - when he needs it. A bard friend = plus 1-6 in the hip pocket, usable after seeing the d20 roll. Div Wiz could help out on an important roll. The fighter could sacrifice his lvl 4 asi for Lucky and have 3 chances to fix bad rolls at the cost of 2 dex, which is significant.

It's harder to get adv. for a ranged attack but possible, of course - let's not even discuss that for now.

Some of the above (like the div reroll) is very hard to factor, but basically this fighter has a +9 to hit on each arrow - but an average +11.5 to hit if blessed. And he has all these hip pocket add-on numbers to use or not use. Let's say he needs to hit AC 20 and rolls a 9 on the d20 and a 3 on Bless. He's got a 12 so far. Except he's got a 7, cuz he said he was taking the -5. He needs 13 more to hit. If he has access to a war cleric's +10 bonus he can do it with very good chances of success. Now he has 17. He rolls his BM Precision but gets a one. Now he has to use his Bard Insp but can't roll a 1. He rolls a 2 or more and hits - or a one and misses. If he "had" to hit and get that extra damage, he (5/6 chance) did it.

Even better, let's say he rolls an 18 on his first shot (that hits even with a -5 for him) - he doesn't have to burn any of his hip pocket bonuses. They're there for the next shot.

Now the question is - is gwm worth all this effort and focus? If that extra 10 killed the monster and stopped it from doing great harm to the party, sure. But as we level up and monsters have 100+ ... 200+ (more) hitpoints, it seems to me that a steady damage rate is going to take over. If the fighter just hits with most of his arrows and does 20 pts - and if a 5-person party is each doing about 20 pts - that's 100 a round, which is hurting this monster badly. And that 10 point bonus - which won't hit much on a high AC monster - hasn't really "scaled" up, not for a high ac/high hp fight. BUT it would have been very useful on the journey going up all those levels (earlier) to get to this fight - and how many of us get to level whatever, anyway?

I see other posters saying essentially just run a barb, use gwm all the time, and just roll with the hits and misses. It'll be fun. Maybe. I think it's important to know there are ways to make gwm work better, if you can get the right party composition and a little consistent help.

[BW022]

You are forgetting that with GWF you can choose when to use it and do things to increase chances of hitting. Maybe you only use it when the wizard casts sleep, fighting zombies, or when the monk keeps tripping folks? Or you wait until higher levels.

[smcmike]

I'm guessing that if you take it, say, at lvl 1 as a huV it will become less and less useful as lvls go up and monster ACs tend to go way up -.

This doesn't really happen in 5e. As you go up in level, GWM will likely become more and more useful. Proficiency bonus increases, ASI's, better access to advantage, and other accuracy-related class features mean you will probably hit more late in the game than early, and access to multiple attacks attacks reduces the negative effects of swinginess.



I see other posters saying essentially just run a barb, use gwm all the time, and just roll with the hits and misses. It'll be fun. Maybe. I think it's important to know there are ways to make gwm work better, if you can get the right party composition and a little consistent help.

Running a Barb and just using GWM more or less all the time is one of the most effective damage-dealing options in the game.

[Chugger]

Thanks. I still want to know if there is a way to substantially up my odds of landing GWM hits in a key fight. I think I have a sense of that - and when not to use gwm. Against pretty high ac. I can load a number off that formula into my head before I go to a table - and from a RP standpoint it would represent my character's intuitive knowledge of the skill.

[djreynolds]

You know a fighter could take magic initiate or a level of cleric for bless, you already have a high con save.
But bless, precision, and advantage. Also using the trip maneuver on your first attack without GWM and then on your next GWM, assuming level 5.

[Foxhound438]

(math stuff)

Your math is, on initial glance, on point, but the takeaway is that something needs to have really high AC to make your GWM attack not worthwhile if you have advantage from something, so unless your DM is intentionally screwing you by using things with AC way out of line (AC 19 is expectable on things with CR 20, while +5 to attacks only goes until level 4), if you have advantage GWM is worth using.

[SharkForce]

Thanks. I still want to know if there is a way to substantially up my odds of landing GWM hits in a key fight. I think I have a sense of that - and when not to use gwm. Against pretty high ac. I can load a number off that formula into my head before I go to a table - and from a RP standpoint it would represent my character's intuitive knowledge of the skill.

well, there are a few ways. advantage is very helpful as noted. one of the easiest ways to get that is to shove something prone.

other ways could include (but are not limited to)...

- having devil's sight or darkvision while in darkness that your opponent cannot see through.
- having a foresight spell on you.
- having a wolf (level 3) totem barbarian nearby.
- your enemy being paralyzed, incapacitated, or blind (for example, by a hold person spell)
- taking the mounted combat feat and riding a large enough mount.
- having your foe be restrained (for example, from the entangle or web spells).
- pack tactics (this is mostly only accessible to kobolds, who can't use heavy weapons effectively... but they can use sharpshooter with bows, which is fairly similar).
- someone else could use a help action (particularly useful if you have relatively weak summons that are capable of taking orders to ready help actions in between each of your attacks, for example)

apart from that, you can look for bonuses to hit. they're not common, but they do exist. for example...

- archery style gives you a bonus to attacks with ranged weapons (for sharpshooter, not great weapons)
- a level 20 barbarian can get +4 to strength, which is worth +2 to hit.
- battlemaster fighters can get a maneuver that grants a bonus to hit.
- bardic inspiration can be used to improve your chance to hit.
- a bless spell can increase your chance to hit.
- magic weapons often (but not always) confer a bonus to hit.
- magic items that can increase your attributes beyond 20, such as books that increase attributes or belts of giant strength, can increase your chance to hit.
- there is an ioun stone that can give you +1 to your proficiency bonus. since your proficiency bonus is added to your attack rolls with weapons you're proficient in (which i'm presuming you intend to be), it will add a bonus to your attack rolls... however, it is probably many times more powerful in the hands of a spellcaster because they can target save DCs that are already low... and it can get to a point where you have spells that never fail and cause extremely powerful negative status effects. however, it can also boost your chance to hit, so it's on this list.

[qube]

@Foxhound438 thx for checking :)
@SharkForce: the magic item oathbow (for sharpshooter, obviously)

Now the question is - is gwm worth all this effort and focus? If that extra 10 killed the monster and stopped it from doing great harm to the party, sure. But as we level up and monsters have 100+ ... 200+ (more) hitpoints, it seems to me that a steady damage rate is going to take over.Interesting question. Most things will never give big bonusses. +2 on an ability score is only +1 to attack and +1 on damage.

Or, for instance, Savage attacker - once per turn rerolling your damage isn't a "big" boost either
it's useless half of the times, since you roll above average half the times, and thus not use it
the things where it would be very useful (high damage) usually have multiple die, meaning the chance you roll very low (and thus the reroll would be very useful) is very small.) If you use it only when you roll below average (as you mathemetically should), the average damage you use it on

1d12 -> (1+2+3+4+5+6)/6 = 21/6 = 3.5
-> differs 3 points of the average of a 1d12, however,
-> it's only useful 6/12 times (half the time, you roll less then average), this is a 1.5 damage bonus (once per round, supposing you hit)

2d6 -> (1*2 + 2*3 + 3*4 + 4*5 + 5*6)/15 = 70/15
-> differs only 2.33 points of the average of 2d6, however,
-> it's only useful 15/36 times so,this is a 0.97222 damage bonus (once per round, supposing you hit)

less then 2 damage per round (assuming you even hit in the first place), isn't that massively awesome against monsters with 100-200 hp either, is it?

Opposite to gameplay, bonusses always look small
5% hit or miss more isn't going to make a mathematical difference
at the game table, missing by 1 or just hitting it's AC makes the difference between 0 and a lot of damage
(or, heck halfling reroll fumble is mathemtically almost as good a +1 attack)
savage attack isn't going to make a mathematical big difference
at the table, when you roll minimum damage, you sure as heck gonna like rerolling that
GWM perhaps isn't going to make some huge difference
... but when you hit an enemy he sure as heck gonna remember it!

Back to topic, +2 str vs -5/+10 (not counting the free attack on a crit)
(X+1/20) * (B+1) < ((X-5)/20) * (B+10)
(X+1) * (B+1) < ((X-5)) * (B+10)
(X+1)B+ (X+1) < ((X-5)B + (X-5)10)
XB+B+ X+1 < BX-5B + 10X-50
6B < 9X-51
2B < 3X-17
So
AC 11, +5 attack (15/20 hits, so X = 15), 2d6 + 3(B=10) damage.
20 < 28, GWM is better then +1 attack & damage
The 'optimal' cut-off would be where the original attack has between 12 & 13 hit chances, so with a +5 attack & 10 damage
AC 13 GWM deals more damage
AC 14 a double ability score improvement deals more damage

For advantage, as the amount of damage doesn't really seem to matter, I venture that X being 7 or higher (instead of 6 or higher) being the ideal cut-off. (an AC of 20)


----

Tl;DR: these things will perhaps not give massive bonusses, but they do imrpove the fun.

[Chugger]

Thanks much to above posters. All helpful!