D&D 5E Current take on GWM/SS

[Psikerlord#]

Ah, so his AC should have been 19, which means GWM would have been even less effective.


I believe you. I really do. But here is the deal... he got lucky. He needed a 13 to hit, and rolled it 5 out 6 times. Even with precision attack, he should have hit 3, maybe 4, times.

So yes, if you get really lucky, you can do a lot of damage. But the next round, without Action Surge and without being really lucky... you are talking less than 30 more hp. Heck if he got as unlucky in round 2 as he got lucky in round 1.... maybe no HP. Luck is a factor in this game.

And that's the problem with dpr calcs. They ignore minimums and maximums. Then when you actually play the game, and a player gets a bit lucky, stuff breaks. On the other hand you can remove the -5/+10 and you don't need to worry.

 

[Ashkelon]

Define 'significantly' and define under what conditions......

Here is an in depth analysis of the DPR gained by GWM/SS:

https://docs.google.com/spreadsheets/d/1cZGd7HwD1dWSzYqVx4PQqw9dl9YYzKxRv7YJtOojkDE/edit?usp=sharing

It also has an added bonus of showing what happens when you combine those feats with Crossbow Expert or Polearm Master (hint, it is scary).

These sheets ignore the accuracy bonuses grantedone of the many of the readily available sources (wolf totem barbarians, hold person/monster, stunning fist, knocking enemies prone, avenger paladins, faerie fire, bless, battlemaster maneuvers, etc). A typical party will have one or more of these available to them.

IMHO a significant increase in DPR is ~10% which is what a +2 to an ability score typically gives. For example going from 20 STR to 22 STR increases a level 17 fighter's DPR from 46 to 51. GWM increases such a fighter's DPR from 46 to 56. So it gives almost twice the benefit of +2 Strength. Furthermore, +2 Strength has diminishing returns with accuracy boosts. GWM actually does better the more accuracy boosts you have, ie it is multiplicative. For example, advantage only adds ~15% more DPR to the 22 Strength fighter's attacks, but adds 25% more DPR to the Great Weapon Master.

Finally, it is my opinion that feats should not even equal stat bumps for pure DPR in terms of DPR boosting. Feats should grant new options and capabilities while stat bumps should increase raw combat power. If you have feats that greatly increase combat power, you get stacking of them that leads to broken builds and trivial combats. The ability cap is there for a reason. A fighter with his 7 feats could only choose feats that increase combat power by 20% and end up with a +140% damage over a fighter who chose more flavorful feats.

 

[KarinsDad]

The EK Crossbowman in my party died since he went archer and made our frontline porous. Not his fault. We choose to not bring a "tank". We had mages, healers, and skills guys. We paid for not bringing a tank.

When encounter 7 came, we were out of buffs and them bugbears got us good.

If the party was out of buffs and such, why wasn't it time to throw in the towel? That could happen after 2 unlucky encounters, or 8 lucky ones.

 

[Minigiant]

If the party was out of buffs and such, why wasn't it time to throw in the towel? That could happen after 2 unlucky encounters, or 8 lucky ones.

We were escaping.
But bugbears did what bugbears do and they got us in a hallway.

 

[Coredump]

It does match. You're just not reading it correctly on purpose because you're looking for holes to pick in it.

Haha..... or maybe the problem is you presented this 'data' without bothering to say "oh, none of this accounts for Bless nor GWM adjustments" Yeah, that could have been helpful.

Bless.

What makes real data actually 'real' is because it actually occurred. So how did you guys roll a 2.5?

Not sure what you're talking about here at all. A crit is always a hit, no matter what your attack bonus is. If something wasn't bolded it was probably because I missed it.

yes, but when a Crit results in a 32, that means there was a +12 attack bonus.

We minus the 5 off each of his attacks manually, and we assume he is doing SS -5/+10 unless stated otherwise, because it's so good. Bless is rolled separately.

So when the 'data' shows a 28, that is really 28+Bless-GWM... okay, got it.

This is only relevant in a white room. In a real game scenario, smart players don't shoot equivalent AC creatures for testing purposes. They take the most optimal path to doing the most damage.

Not at all, in fact it is *more* important in a real game.

Doing 25 points of damage is more 'valuable' against an AC 19 creature, than doing 25 pts against an AC 13 creature. If someone is targetting low AC creatures, it will inflate their damage numbers, because they are taking the easy way out. It is like saying scoring 25 points against a college team is the same as scoring 25 points against a Pro team. Or in this case, it is like scoring 50 points against a college team means you are twice as good as the person who scored 25 points against a pro team.
In a white room, all HP may be the same, but in a real game.... some damage is more meaningful than other.


But this raises other questions..



You should check your macros, or possibly contact Roll20. The Longbowman's dice are over 2 std dev's from the expected value, so you are talking a 1.5% to 2% chance of rolling that well. The Crossbowman, is over 3.5 Std dev's from the expected value. Which puts it somewhere around 0.03% chance of rolling that well. The crossbowman was getting the equivalent of a +3.5 attack bonus from the Roll20 rolling algorithm. So either there is something wrong, or he got really really *really* lucky.

Some of the damage results look....interesting.
But without information regarding which rolls were hits, or even how many were hits in total.... its hard to make any determinations of that data.



But yes, with a base +12 attack bonus, and then rolling +3.5 over average, and aiming at low AC targets.... SS will do a bunch of damage.
If you compare it to a longbow that isn't rolling as well, and is aiming at much tougher targets.... yeah SS will look a lot better.

 

[Coredump]

Nothing is wrong. No magic items were used in these games, but in most games they are used, so it gets even worse.

Nope, in general +X weapons make GWM *less* of a bonus. And things like Flametongue are even worse.

 

[DaveDash]

Haha..... or maybe the problem is you presented this 'data' without bothering to say "oh, none of this accounts for Bless nor GWM adjustments" Yeah, that could have been helpful.

I mentioned that I was including attack rolls in my original post. It's not my problem if you jumped to conclusions in your eagerness to prove me wrong.

What makes real data actually 'real' is because it actually occurred. So how did you guys roll a 2.5?

I never included 2.5's in any data. I included 2.5's in replies to posts that were like 'ZOMG HOW COME YOUR CROSSBOW EXPERT FIGHTER DIDNT MISS ALL THE TIME? HAX!"

Not at all, in fact it is *more* important in a real game.

Doing 25 points of damage is more 'valuable' against an AC 19 creature, than doing 25 pts against an AC 13 creature. If someone is targetting low AC creatures, it will inflate their damage numbers, because they are taking the easy way out. It is like saying scoring 25 points against a college team is the same as scoring 25 points against a Pro team. Or in this case, it is like scoring 50 points against a college team means you are twice as good as the person who scored 25 points against a pro team.
In a white room, all HP may be the same, but in a real game.... some damage is more meaningful than other.

Way too subjective to have any basis for discussion. Fighting important creatures that may or may not have high AC may also call for the party to expend other resources to gain advantage etc. This is why real world data >>>> white room maths.

Also your assumption is totally flawed. Cleaning up a bunch of low CR's that can still do a lot of damage is very important. I've had skeletons (CR1/2) do ~60 damage to a Level 17 Battlemaster Fighter in one combat. Action economy in this game is important, and so is thinning down numbers of enemy creatures, something a SS archer will excel at.

How do they have a +12 attack bonus without Bless or Magic weapons?

You should check your macros, or possibly contact Roll20. The Longbowman's dice are over 2 std dev's from the expected value, so you are talking a 1.5% to 2% chance of rolling that well. The Crossbowman, is over 3.5 Std dev's from the expected value. Which puts it somewhere around 0.03% chance of rolling that well. The crossbowman was getting the equivalent of a +3.5 attack bonus from the Roll20 rolling algorithm. So either there is something wrong, or he got really really *really* lucky.

Some of the damage results look....interesting.
But without information regarding which rolls were hits, or even how many were hits in total.... its hard to make any determinations of that data.

You can't work out how to get +12 on an attack roll? Really?

You're also not factoring in both sets of data - because there were a lot of 1's rolled in the first set.

But yes, with a base +12 attack bonus, and then rolling +3.5 over average, and aiming at low AC targets.... SS will do a bunch of damage.
If you compare it to a longbow that isn't rolling as well, and is aiming at much tougher targets.... yeah SS will look a lot better.

No one is claiming that SS is overpowered in the low parts of the game. It's around level 8 when it starts to become a bit of an issue, and really around level 13 that things start to fall apart, when you have an attack bonus of +12. With magic weapons, this problem becomes apparent earlier.

You're aiming at a mix of low and high AC targets, but by design there are more lower AC targets than higher AC targets. Go look at the monsters manual. This is a critical factor you keep dismissing but it's a general game fact. One of the core design principals of bounded accuracy is the ability to use low CR creatures all through-out the game, and it's built into the encounter building guidelines and design assumptions.

You're also jumping to a wild conclusion that the longbow archer and picking high AC targets and the SS fighter is not. Now you're just trying to realign reality to match your incorrect expectations.

 

[DaveDash]

Nope, in general +X weapons make GWM *less* of a bonus. And things like Flametongue are even worse.

I've never once been referring to GWM.

 

[Hussar]

But, if the SS is picking off the low hanging fruit, and the Longbow fighter is hitting the big guys, isn't that a win at the table? Everyone's contributing, everyone's doing their thing and no one is stepping on anyone's toes. isn't this a good thing?

But, again, DaveDash, you are committing a big, big fallacy. Your "real world" data is anecdotal. It does not prove anything. You'd need to go over a LOT more data than a couple of fights and a gut feeling. We're talking hundreds of combats across numerous DM's in order to gain a decent set of information.

Heck, your using Roll 20. Rerun the encounters. Do everything exactly the same, but only change the die rolls. Do you get the same results? I dunno. But, that's how you would have to test your hypothesis. As it stands, the only thing you've proven so far is that your players got really, really lucky during a fight and rolled incredibly high. If I'm rolling nothing but 18's and 20's all night, anything I do is going to be seriously skewed. As I said, in order for your "real world" experience to be more broadly applicable, you need to test it more.

 

[DaveDash]

But, if the SS is picking off the low hanging fruit, and the Longbow fighter is hitting the big guys, isn't that a win at the table? Everyone's contributing, everyone's doing their thing and no one is stepping on anyone's toes. isn't this a good thing?

But, again, DaveDash, you are committing a big, big fallacy. Your "real world" data is anecdotal. It does not prove anything. You'd need to go over a LOT more data than a couple of fights and a gut feeling. We're talking hundreds of combats across numerous DM's in order to gain a decent set of information.

Heck, your using Roll 20. Rerun the encounters. Do everything exactly the same, but only change the die rolls. Do you get the same results? I dunno. But, that's how you would have to test your hypothesis. As it stands, the only thing you've proven so far is that your players got really, really lucky during a fight and rolled incredibly high. If I'm rolling nothing but 18's and 20's all night, anything I do is going to be seriously skewed. As I said, in order for your "real world" experience to be more broadly applicable, you need to test it more.

I'm happy to accept that the data I have shown is "anecdotal" that's cool, however I have a lot more data than just that and the trend is more or less the same. If you want to dismiss it, that's fine. Really I'm not invested in this enough to put any more effort into it.

I am rather curious though. What do you have to say about the guy who posted probability analysis that matches my data up thread?