D&D 5E Can an elf rogue be a decent archer in (Basic) D&D 5th edition?

[Dausuul]

Well, I was wrong. The rogue does just about the same with archery--worse in the first couple of levels, slightly better from level 3 onward. Here's what it looks like with archery:

Level 1: +34%
Level 2: +34%
Level 3: +25%
Level 4: +22%
Level 5: -22%
Level 6: -34%
Level 7: -29%
Level 8-10: -15%
Level 11-13: -34%
Level 14-16: -24%
Level 17-19: -14%
Level 20: -28%

10-level average: 89%
20-level average: 80%

A little more about how I figured the numbers: I assume a 70% base chance to hit the monster at every level (but see below). The archer fighter gets a 5% accuracy boost due to Archery Style. Melee fighters get Great Weapon Style's damage-on-a-miss instead, which turns out to be a heck of a lot better than 5% accuracy.

(Slight adjustment to my previous numbers: I neglected to adjust for the fighter's accuracy advantage at level 6, when fighters get their second ability score boost one level ahead of rogues. The level 6 number for melee should be -34% instead of -32%.)

Now, there are a few things I didn't factor in, and they do favor the fighter. A first-round crit with max sneak attack damage is excellent, but a full round of extra attacks seems to be slightly better; Action Surge beats Assassinate. As you point out, magic weapons also tend to increase the fighter's lead. I just ran the numbers assuming +1 weapons from level 6, +2 weapons from level 11, and +3 weapons from level 16. The 10-level averages become 82% melee/86% archer, and the 20-level averages become 66% melee/69% archer. However, magic weapons are extremely campaign-specific; I'm not at all sure these are good assumptions, so I'm leaving them out of the baseline. Finally, fighters spread their damage out over a series of attacks while rogues do it all in one big burst. That means rogues will on average lose more damage to "blow-through."

Google spreadsheet link: https://docs.google.com/spreadsheets/d/18VvHy8Q_lUr5e-mZqvqfsr4_STocWnv_l3scgXd_mlM/edit?usp=sharing

 

[Ruin Explorer]

Wow you really cannot accept the possibility that theoretical averages don't always give those results in play. Amazing. Why do we even roll dice, if the d20 always comes up a 10? Now hear this - it's not mis-remembering, it's not that the fighters (two of them, and then a third that is a ranger) are played poorly, it's that THE ROGUE REALLY HAS DONE MORE DAMAGE. Now if you want to call me a liar, go ahead and call me a liar. Otherwise, accept the possibility that a game with a massive random element to it might not play out based on the averages, even over a year of time.

I think you're missing my point. It's not that I don't accept it - it's that I am saying it is irrelevant to a rules/design discussion of this nature.

Given we're only at 4th level, that may well be true, and you should have considered that before accusing me of lying or misremembering or my players being incompetent. Particularly since every poll we've ever conducted shows people spend most of their time a lower levels.

I have not suggested you are lying or your players incompetent. One player being more skilled or lucky does not render the rest "incompetent", imo (do you disagree?). It does render the results of the session largely irrelevant, though.

I have suggested people's memories are not a great source of data here, and unhelpful to meaningful discussion of the actual rules and mechanics. I stand by that.

The lower level argument is, on the other hand, an interesting one in light of Dasuul's figures (I will produce my own in due course), and one could argue weighting accordingly, but I think you may concur that going from +18% to -25% in a single level is quite severe change, especially as the Rogue never breaks positive again.

 

[Dausuul]

This type of math is applicable to theoretical averages, but that doesn't dictate results. If it did, then every sports team with the theoretical averages that were higher than everyone else would win every game and always go on to win the championship and there would be no purpose to playing. But that's not how life works, and it's also not how this game works. Averages might say you will roll a 1 and a 20 roughly equal number of times, but if the 1 is spent on a listen door check and the 20 on an attack roll, damage will be higher, for instance. Because a huge amount of luck comes into this game.

When it comes to game design, the average is what matters. For every group where a character gets lucky and outperforms the average, there will be another where the character gets unlucky and falls short. However:

Given we're only at 4th level, that may well be true...

Per my numbers above, what you're seeing is in fact average performance. At 4th level, rogues still outdamage fighters by a substantial margin. Things will change when the fighter's first Extra Attack kicks in.

 

[Ruin Explorer]

Well, I was wrong. The rogue does just about the same with archery--worse in the first couple of levels, slightly better from level 3 onward. Here's what it looks like with archery:

Level 1: +34%
Level 2: +34%
Level 3: +25%
Level 4: +22%
Level 5: -22%
Level 6: -34%
Level 7: -29%
Level 8-10: -15%
Level 11-13: -34%
Level 14-16: -24%
Level 17-19: -14%
Level 20: -28%

10-level average: 89%
20-level average: 80%

A little more about how I figured the numbers: I assume a 70% base chance to hit the monster at every level (but see below). The archer fighter gets a 5% accuracy boost due to Archery Style. Melee fighters get Great Weapon Style's damage-on-a-miss instead, which turns out to be a heck of a lot better than 5% accuracy.

(Slight adjustment to my previous numbers: I neglected to adjust for the fighter's accuracy advantage at level 6, when fighters get their second ability score boost one level ahead of rogues. The level 6 number for melee should be -34% instead of -32%.)

Now, there are a few things I didn't factor in, and they do favor the fighter. A first-round crit with max sneak attack damage is excellent, but a full round of extra attacks seems to be slightly better; Action Surge beats Assassinate. As you point out, magic weapons also tend to increase the fighter's lead. I just ran the numbers assuming +1 weapons from level 6, +2 weapons from level 11, and +3 weapons from level 16. The 10-level averages become 82% melee/86% archer, and the 20-level averages become 66% melee/69% archer. However, magic weapons are extremely campaign-specific; I'm not at all sure these are good assumptions, so I'm leaving them out of the baseline. Finally, fighters spread their damage out over a series of attacks while rogues do it all in one big burst. That means rogues will on average lose more damage to "blow-through."

Google spreadsheet link: https://docs.google.com/spreadsheets/d/18VvHy8Q_lUr5e-mZqvqfsr4_STocWnv_l3scgXd_mlM/edit?usp=sharing

Great post, thank you, I'll digest this for a bit before any further comments.

 

[Mistwell]

I think you're missing my point. It's not that I don't accept it - it's that I am saying it is irrelevant to a rules/design discussion of this nature.

And I think if that was your point, you should have said it. Instead, you repeatedly told people their experience was wrong, and then said we were mis-remembering, or the fighters were not built well or played well. So yeah I am pretty sure I was interpreting your responses correctly.

I have not suggested you are lying or your players incompetent. One player being more skilled or lucky does not render the rest "incompetent", imo (do you disagree?). It does render the results of the session largely irrelevant, though.

First, you did say (repeatedly) people were "wrong" when they told you their experience. Second, it's looking like you might have been wrong on your averages.

I have suggested people's memories are not a great source of data here, and unhelpful to meaningful discussion of the actual rules and mechanics. I stand by that.

You didn't suggest "peoples" memories are wrong, you said MY memory was wrong for disagreeing with you. I keep a story hour for the game I am talking about, you are welcome to check it for accuracy. I am right on this and you are wrong - which is not amazing as I am speaking from direct knowledge and you're speaking from pure speculation.

And now that the numbers back up my experience, where it was just proved out at levels 1-4, which is all we've played, the rogue SHOULD be doing more damage, I hope now you'll accept you were wrong to automatically assume my experiences were incorrect and that it was a sign of mis-remembering or the fighters and ranger not being built as well. Instead, it was a flaw in your assumptions, and you never asked what levels we were playing and should have.

None of which would have been an issue at all if you hadn't been so damn aggressive in your tone this whole time.

 

[Dausuul]

One other thing: It's not clear from the rules whether you max out sneak attack dice on a normal (non-Assassinate) crit. A naive reading of the rules suggests that you do. However, Assassinate goes out of its way to say the dice are maxed, which would suggest they normally aren't. I assumed that sneak attack crits do not max sneak attack dice. It doesn't make a big difference either way, though, since rogues don't get the fighter's expanded crit range.

 

[Mistwell]

When it comes to game design, the average is what matters. For every group where a character gets lucky and outperforms the average, there will be another where the character gets unlucky and falls short.

Sure, so he should be talking in averages, not telling people their experience is wrong, they must be mis-remembering things, and/or it's a sign the fighters and ranger were not built very well. It's the tone I objected to - and for message boards that's what matters.

However:

Per my numbers above, what you're seeing is in fact average performance. At 4th level, rogues still outdamage fighters by a substantial margin. Things will change when the fighter's first Extra Attack kicks in.

Makes sense. Given only two of our games in the past 20 years have gone past 10th level, I think it's a fair trade-off point

 

[Ruin Explorer]

None of which would have been an issue at all if you hadn't been so damn aggressive in your tone this whole time.

This seems really ironic to me.

EDIT - My averages were not significantly wrong - but they looked at 20th level - because I was addressing that example (I did not introduce it!), not low levels. I did make an error in not investigating low levels, though, which Dasuul has corrected. I do apologise for saying you were "wrong", but I stand by my general distrust of "experiences".

EDIT EDIT - That said I should have picked up on the correlation of experiences and checked the low-level data as a result, so experiences have some value - not probative, but in flagging things up.

 

[Ruzak]

Thanks to both Rune Explorer & Dasuul for putting together this math! We all have our individual anecdotal experiences, but is great to see where the true average values actually are.

 

[Cybit]

So many things I want to say but can't, darn you NDA. >_> <_<