5E Rogue Archetype - The Gambler

[Reevh]

So I decided to take a stab at a homebrew rogue archetype, and I was wondering what folks think:

Rogue Archetype - The Gambler

Poker Face
Starting at third level you become exceptionally good at hiding your feelings and intentions. Insight checks against you are always at disadvantage.

All In
Starting at third level, when you make a weapon attack, you may choose to go All In. The attack must be eligible for sneak attack and you must choose to do so before making the attack roll. If the attack succeeds, you may add a number of sneak attack dice equal to your charisma modifier. If the attack fails, you overextend yourself and cannot use reactions or bonus actions until the end of your next turn. You may use this ability a number of times per day equal to your proficiency modifier.

Know When to Fold ‘Em
Starting at 9th level, you learn to back away from a bad bet. When an enemy reduces you to zero hit points you may instead use your reaction to drop to one hit point and move up to your movement speed without provoking opportunity attacks. Once you have used this ability you must take a long rest before using it again.

Lucky Coin
Starting at 13th level, your lucky coin takes on the exceptional ability to bend luck in your favor right when you need it most. Whenever you finish a long rest you may flip you lucky coin, rolling 1d4. Even results are Heads and odd results are Tails.
If the result is Heads, you may use your reaction give yourself or any creature within 30 feet advantage on up to two attack rolls, saves, or ability checks.
If the result is Tails, you may use your reaction to impose disadvantage on up to two attack rolls, saves, or ability checks for a creature within 30 feet. This feature may be used after the roll but before the outcome of the roll is known.

Royal Flush
At 17th level, sometimes luck just works out for you. When you roll sneak attack damage, if at least 5 of the dice have the same result, you may roll the dice again and add the damage to your result. If Royal Flush is triggered by a critical hit, do not double the Royal Flush dice.

[Evaar]

I'm not a math guy, so the numbers here may need tweaking, but I think you've got a really strong premise here supported by cool and flavorful mechanics. I would definitely consider playing it.

Just two notes:
*I might change Poker Face to advantage on Deception checks. It's just a little more active on the player's part - although Insight and Deception are often contested, you don't always know when the DM is rolling Insight against you, but you always know when you're rolling Deception.
*I understand the premise of All In, but I wonder if there's a slightly more elegant mechanic for it. I worry that as written it incentivizes the Rogue to remember to, say, Cunning Action Disengage first, THEN attack, since that way they've already used the bonus action before it gets locked out. Not that that's a balance concern really, it's just a little awkward. Honestly I think you could just lock out reactions only, still let the Rogue have bonus actions, and that would still be fine - you have several valuable reactions available as a Rogue, both from the base class chassis and from this subclass, not to mention the value of off-turn attacks for Rogues.

Really good stuff, though. You should playtest it and see if maybe Ludic 's DPR calculator can confirm whether the damage is in line.

[Psyren]

All In is too strong, with that you can reliably get 5d6 (2d6+3d6) sneak attack at level 3. (by comparison, that much SA isn't usually available until 6 levels later.) I would consider half-Cha here or just adding Cha mod to the damage.

Lucky Coin meanwhile is too weak at 2 uses per long rest. I'd make this = proficiency bonus. It also needs a "that you can see" restriction.

Royal Flush is a fun idea but be wary as it combos with All In - both just in terms of more damage (if both of them apply to a given hit you now have 20d6 Sneak Attack as currently written, and that's without a crit), as well as the fact that All In adds dice so it makes it more likely that you'll get a flush in the first place.

The "only non-crit dice count" restriction is trouble too. There are tables that roll all the dice together for a crit (Virtual Tabletops like roll20 in particular do this), so determining which dice came from the crit and which were "base" can become a problem.

[Peelee]

The Mod on the Silver Mountain: Moved to Homebew.

[Breccia]

*I might change Poker Face to advantage on Deception checks.

For the record, I disagree...but not by much.

Generally speaking, it is easier to give yourself Advantage to a Skill check than to force Disadvantage on the opposition. If nothing else, Inspiration. The Gambler therefore has the ability to bluff really well, if they can grant themselves Advantage while forcing Disadvantage on the opposition.

There's also the question of Natural 20 vs. Natural 1, if you're using those for skill checks.

And of course, if your opponent has Advantage, countering that sounds pretty handy.

That's about it. Numerically, I'm not sure there's a ton of difference between "I have Advantage" vs "You have Disadvantage" when it comes to the dice actually being rolled. For a while I thought imposing Disadvantage was numerically stronger than having Advantage, but I ran some numbers and no, it's not.

As for the rest of the subclass:

All-In's limit to its uses per day keep it basically in line. It's strong, but let's see how strong.

The example mentioned was a 3rd-level Gambler with 16 Cha, doing +3d6 extra damage. That's one first-level spell's damage. And they can attempt it twice, and these attempts could miss. Still, if it hits, the attack does on average +10.5 damage.

I can think of two comparisons, neither of which come from a subclass:

1) A 3rd-level Sorcerer could, if built this way, Quicken their spell and also throw a cantrip, such as fire bolt. The Sorcerer's chance of hitting with this bonus cantrip are probably the same as the Gambler hitting with All-In...I mean, technically higher, the Gambler's Cha is 16 I doubt their Dex is 16 also but whatever...and the average of 1d10 is 5.5 barring a Feat or something. The 3rd-level Sorcerer can do this once/day because it takes 2 SP to do it and they only have 3.
2) Or, a 3rd-level Fighter could Action Surge and attack again. They, too, probably have the same chance of hitting as the Gambler. Giving the Fighter Str 16 and a maul, that's 10 damage. They can also only do this once per day.

Incidentally at higher levels, the Fighter and Sorcerer pull ahead. At 17th level that extra fire bolt does 22 damage, so unless the Gambler has a Cha of 24 they've fallen behind. Also the Sorcerer can do it eight times/day not six, plus they can burn spell levels to...you get the idea. The Fighter's Action Surge can only be used twice, so the Gambler can All-In more than the Fighter, but they can attack three times with each. Even if they're still using a nonmagical maul and 16 Str, which for the record no they're not, but even if they were that's 30 damage. More realistically but not OP, a +2 weapon and 18 Str raises that to 39 damage. Both of these classes easily add more damage in a single round than the Gambler can.

Sorcerer's total bonus per day: 8 x 22 = 176
Fighter's total bonus per day: 2 x 39 = 78
Gambler's total bonus per day with 16 Cha: 6 x 10.5 = 63

Since the Sorcerer's Cha and the Fighter's Str are likely higher than the Gambler's Cha (being their secondary stat), the damage boost the Gambler gets isn't keeping up. Which to be fair, it shouldn't, it's a subclass ability.

The issue, then, is the apparent unfairly-high damage at low levels. There's an easy fix: All-In cannot add more SA dice than what you normally get. This also takes the "heat" off "forcing" your Gambler to have a 16 Cha at level 3. At 3rd level, their Cha could be 14, 16, or 24, and they'd still add +2d6 averaging 7 maximum.

Another option? Delay the use of All-In. Watch any gambling movie, the hero never ever bets all their chips on their very first hand without looking at their cards. The showdown is later in the game. Yes, the bonus damage would still be high, but if the Gambler can't use All-In till round 3 or 4 or something, they won't ruin the encounter by one-shotting the End Boss before anyone else gets to even move. "For each attack/hit/sneak attack [the subclass should specify one] on the target, you get one rank of Better Hand, up to your Cha modifier. Before you declare your attack, you may spend as many ranks of Better Hand as you want, and if the attack is a successful Sneak Attack, add +1d6 per rank of Better Hand you spent". Something like that.

Royal Flush is...not very well worded. I'm going with the assumption that you meant "roll the dice that all matched again" and not all the dice total. Anyhow, this attack is facing some long odds.

Baseline, if you roll 9d6, the chance of any one number showing up 5 times is about five percent. Rolling those five dice again would add +5d6, duh, which is +17.5 on average. Baseline, Royal Flush is adding 1 damage per Sneak Attack, on average. Unlike All-In, you can't make this happen just because you want to -- it's 100% luck. Well, barring special modifiers like a magic dagger that turns all 1's into 2's or something.

But just to see the far end, let's go 10d6 plus an All-In from a 20 Charisma. Because why not. When rolling 15d6, the probability of any five (or more) dice being the same number is much higher -- 54% give or take. You're also seeing decent odds of six or even seven dice matching. The bonus damage doesn't increase much, stopping just shy of +19. Therefore, for a maxxed-out high-Cha Gambler using All-In, Royal Flush adds...a total of 9 extra damage (the total bonus is 10.1 but they were getting +1 already). Okay, yes, +9 damage on average doesn't suck at all. But even with about as favorable a situation as you can get, it's a coin flip whether you get it or not.

By contrast, the Sorcerer is looking at Empowering spells. Speaking of coin flip, a level 20 Sorcerer rolling 10d8 and counting the dice less than 4 also has a roughly coin-flip chance to see five (or more) of them, reroll those five, and end up adding about the same. The average of one through four is 2.5, the average of a reroll is 4.5, so if the Sorcerer sees five of those, they're on average going to add +10 to the end result by Empowering and rerolling those five dice. If they don't see five low rolls, oh no, their damage roll was good, they're not exactly complaining.

The difference, of course, is that the Sorcerer must spend a SP to do this...but will do this when the rolls are poor, not blind luck. (They're also probably hitting mulitple targets with a 10d8 spell at 20th level, the Gambler is not). In an interesting coincidence, by level 20, the Sorcerer can throw that bonus fire bolt while also Empowering a 10d8 spell six times, just like the Gambler. Except the Sorcerer can then use a Bonus Action, burn a third-level spell slot, and do it again next round, too.

Incidentally, if I had to "fix" Royal Flush, it would be to rip off the Sorcerer. The Gambler can re-roll the matching dice, if they're all 1's, all 2's, or all 3's. If they're higher, the Sneak Attack damage is already high enough and the Gambler doesn't need the help. This can help a Gambler with a poor Sneak Attack result, but cuts down the massive overkill of a Sneak Attack All-In adding to an already high damage output.

What if I read it wrong, and all the Sneak Attack dice are re-rolled?

Well then, the damage is way too high and the ability is broken beyond repair. Baseline it's not bad, you have a 5% chance to "Sneak Attack Crit", as listed before, and the overall average damage gain per Sneak Attack is +2. But as also shown before, All-In on a 10d6 Sneak Attack has a 50+% chance of doubling, and that boost is both too high and also too common.

One last suggestion: Know When To Fold Them should not work if the attack outright kills the Gambler. There's "lucky" and then there's "shot in the skull and didn't die".

[Olffandad]

I like the concept too, very cool!

How about for Poker Face -

As a reaction, any Insight checks against you automatically fail. The person rolling the Insight check is not aware of this effect. This effect lasts for one minute. You can do it a number of times per day equal to your Proficiency bonus.

All In needs a huge risk to balance a huge reward, like the target inflicts a critical hit on you if you miss with All In.

Royal Flush is brilliant thematically - I would have it be 1-2-3-4-5 on the dice but that's way too easy to do if you get enough dice.

You could also add a few Gaming Set proficiencies as bonus skills - the Gambler should be a skilled player of different games, even without the gambling.

[Reevh]

I'm not a math guy, so the numbers here may need tweaking, but I think you've got a really strong premise here supported by cool and flavorful mechanics. I would definitely consider playing it.

Just two notes:
*I might change Poker Face to advantage on Deception checks. It's just a little more active on the player's part - although Insight and Deception are often contested, you don't always know when the DM is rolling Insight against you, but you always know when you're rolling Deception.
*I understand the premise of All In, but I wonder if there's a slightly more elegant mechanic for it. I worry that as written it incentivizes the Rogue to remember to, say, Cunning Action Disengage first, THEN attack, since that way they've already used the bonus action before it gets locked out. Not that that's a balance concern really, it's just a little awkward. Honestly I think you could just lock out reactions only, still let the Rogue have bonus actions, and that would still be fine - you have several valuable reactions available as a Rogue, both from the base class chassis and from this subclass, not to mention the value of off-turn attacks for Rogues.

Really good stuff, though. You should playtest it and see if maybe Ludic 's DPR calculator can confirm whether the damage is in line.

Thanks for the feedback. My feeling on Poker Face was that there are a lot of things that give advantage on deception checks, including sometimes just good RP if your DM chooses. Disadvantage on the insight checks of others would still apply even if you already have advantage from other sources. Also, my DM sometimes rolls insight checks against us without telling us he's even doing so. This would apply in those situations too.

I agree it's a little inelegant on the turn in which you use it. The intent is that you lose your bonus action on the next turn, and your reaction until the end of your next turn. You have a suggestion for better wording? Maybe All In could require your bonus action to use? I feel like that's a bit punitive though, as it precludes using your cunning action that round, which isn't my intent.

[Reevh]

All In is too strong, with that you can reliably get 5d6 (2d6+3d6) sneak attack at level 3. (by comparison, that much SA isn't usually available until 6 levels later.) I would consider half-Cha here or just adding Cha mod to the damage.

Lucky Coin meanwhile is too weak at 2 uses per long rest. I'd make this = proficiency bonus. It also needs a "that you can see" restriction.

Royal Flush is a fun idea but be wary as it combos with All In - both just in terms of more damage (if both of them apply to a given hit you now have 20d6 Sneak Attack as currently written, and that's without a crit), as well as the fact that All In adds dice so it makes it more likely that you'll get a flush in the first place.

The "only non-crit dice count" restriction is trouble too. There are tables that roll all the dice together for a crit (Virtual Tabletops like roll20 in particular do this), so determining which dice came from the crit and which were "base" can become a problem.

Thanks for the feedback.

My view on All In is that even if you have 16 CHA at level 3, which is fairly high if you're prioritizing DEX, you only get 3d6 (average 10.5 damage) more damage twice per day, and that's if they even hit. Yeah, it's a big boost to that one round if you do hit, but it's not so crazy out of line with things like Action Surge, which might do 1d8+4 (average 8.5 damage), and has no penalty for failure. Plus, the Action Surge is once per short rest, which means you may be able to use it more than twice in a day. This tilts farther in the fighter's favor when you add on Battlemaster dice.

I also compared it against the Phantom Rogue subclass, whose Wails from the Grave feature lets you do half your sneak attack dice, rounded up, up to your proficiency times per day. Same number of uses, 2d6 dice at level 3, no penalty, and scales better in the mid-levels. This leads me to think this is not particularly overpowered at low levels, and may in fact be a bit underpowered until you get your hands on Royal Flush.

Lucky Coin is my least favorite thematically, but I'm OK with the mechanics. Twice per day makes this a little bit similar to a Divination Wizard's ability to replace dice rolls twice per day. It's weaker, but it's also not the core identity of the class, so I'm OK with that. Someone else suggested that I could change it to "Luck of the Draw," and have the player draw a card from a deck and have a different outcome depending on the suit of the card drawn (or roll a d4 if you don't have cards handy). I like that flavor better, but I'm not yet sure how to implement it.

The synergy between All In and Royal Flush is intentional.

I agree the "only non-crit dice count" restriction for Royal Flush isn't especially great. I was worried that if you could attack a paralyzed enemy and use All In at 17th level with 20 CHA, you'd get 56d6 sneak attack dice, which seems crazy. But then I again compared it against what other classes can do at that level and it seems a bit more reasonable. At level 17 if you get to sneak attack a paralyzed enemy, you SHOULD blow them up.

[Reevh]

For the record, I disagree...but not by much.

Generally speaking, it is easier to give yourself Advantage to a Skill check than to force Disadvantage on the opposition. If nothing else, Inspiration. The Gambler therefore has the ability to bluff really well, if they can grant themselves Advantage while forcing Disadvantage on the opposition.

There's also the question of Natural 20 vs. Natural 1, if you're using those for skill checks.

And of course, if your opponent has Advantage, countering that sounds pretty handy.

That's about it. Numerically, I'm not sure there's a ton of difference between "I have Advantage" vs "You have Disadvantage" when it comes to the dice actually being rolled. For a while I thought imposing Disadvantage was numerically stronger than having Advantage, but I ran some numbers and no, it's not.

Thanks for your feedback, I agree.

As for the rest of the subclass:

All-In's limit to its uses per day keep it basically in line. It's strong, but let's see how strong.

The example mentioned was a 3rd-level Gambler with 16 Cha, doing +3d6 extra damage. That's one first-level spell's damage. And they can attempt it twice, and these attempts could miss. Still, if it hits, the attack does on average +10.5 damage.

I can think of two comparisons, neither of which come from a subclass:

1) A 3rd-level Sorcerer could, if built this way, Quicken their spell and also throw a cantrip, such as fire bolt. The Sorcerer's chance of hitting with this bonus cantrip are probably the same as the Gambler hitting with All-In...I mean, technically higher, the Gambler's Cha is 16 I doubt their Dex is 16 also but whatever...and the average of 1d10 is 5.5 barring a Feat or something. The 3rd-level Sorcerer can do this once/day because it takes 2 SP to do it and they only have 3.
2) Or, a 3rd-level Fighter could Action Surge and attack again. They, too, probably have the same chance of hitting as the Gambler. Giving the Fighter Str 16 and a maul, that's 10 damage. They can also only do this once per day.

Action Surge is once per short rest at that level, not once per day.

Incidentally at higher levels, the Fighter and Sorcerer pull ahead. At 17th level that extra fire bolt does 22 damage, so unless the Gambler has a Cha of 24 they've fallen behind. Also the Sorcerer can do it eight times/day not six, plus they can burn spell levels to...you get the idea. The Fighter's Action Surge can only be used twice, so the Gambler can All-In more than the Fighter, but they can attack three times with each. Even if they're still using a nonmagical maul and 16 Str, which for the record no they're not, but even if they were that's 30 damage. More realistically but not OP, a +2 weapon and 18 Str raises that to 39 damage. Both of these classes easily add more damage in a single round than the Gambler can.

Sorcerer's total bonus per day: 8 x 22 = 176
Fighter's total bonus per day: 2 x 39 = 78
Gambler's total bonus per day with 16 Cha: 6 x 10.5 = 63

Since the Sorcerer's Cha and the Fighter's Str are likely higher than the Gambler's Cha (being their secondary stat), the damage boost the Gambler gets isn't keeping up. Which to be fair, it shouldn't, it's a subclass ability.

The issue, then, is the apparent unfairly-high damage at low levels. There's an easy fix: All-In cannot add more SA dice than what you normally get. This also takes the "heat" off "forcing" your Gambler to have a 16 Cha at level 3. At 3rd level, their Cha could be 14, 16, or 24, and they'd still add +2d6 averaging 7 maximum.

I think it's not SO much more damage than other class abilities that it needs nerfing, and unlike those other abilities, it has a penalty if you fail, which I feel is enough balance. The whole idea here is risk vs. reward to match the gambler theme.

Another option? Delay the use of All-In. Watch any gambling movie, the hero never ever bets all their chips on their very first hand without looking at their cards. The showdown is later in the game. Yes, the bonus damage would still be high, but if the Gambler can't use All-In till round 3 or 4 or something, they won't ruin the encounter by one-shotting the End Boss before anyone else gets to even move. "For each attack/hit/sneak attack [the subclass should specify one] on the target, you get one rank of Better Hand, up to your Cha modifier. Before you declare your attack, you may spend as many ranks of Better Hand as you want, and if the attack is a successful Sneak Attack, add +1d6 per rank of Better Hand you spent". Something like that.

I don't mind the idea of building up a "pot" of sneak attack dice that you can later bet on, but I'm not sure this is my favorite implementation. Also, I just already think it's not crazy overpowered as it is.

Royal Flush is...not very well worded. I'm going with the assumption that you meant "roll the dice that all matched again" and not all the dice total. Anyhow, this attack is facing some long odds.

Baseline, if you roll 9d6, the chance of any one number showing up 5 times is about five percent. Rolling those five dice again would add +5d6, duh, which is +17.5 on average. Baseline, Royal Flush is adding 1 damage per Sneak Attack, on average. Unlike All-In, you can't make this happen just because you want to -- it's 100% luck. Well, barring special modifiers like a magic dagger that turns all 1's into 2's or something.

But just to see the far end, let's go 10d6 plus an All-In from a 20 Charisma. Because why not. When rolling 15d6, the probability of any five (or more) dice being the same number is much higher -- 54% give or take. You're also seeing decent odds of six or even seven dice matching. The bonus damage doesn't increase much, stopping just shy of +19. Therefore, for a maxxed-out high-Cha Gambler using All-In, Royal Flush adds...a total of 9 extra damage (the total bonus is 10.1 but they were getting +1 already). Okay, yes, +9 damage on average doesn't suck at all. But even with about as favorable a situation as you can get, it's a coin flip whether you get it or not.

By contrast, the Sorcerer is looking at Empowering spells. Speaking of coin flip, a level 20 Sorcerer rolling 10d8 and counting the dice less than 4 also has a roughly coin-flip chance to see five (or more) of them, reroll those five, and end up adding about the same. The average of one through four is 2.5, the average of a reroll is 4.5, so if the Sorcerer sees five of those, they're on average going to add +10 to the end result by Empowering and rerolling those five dice. If they don't see five low rolls, oh no, their damage roll was good, they're not exactly complaining.

The difference, of course, is that the Sorcerer must spend a SP to do this...but will do this when the rolls are poor, not blind luck. (They're also probably hitting mulitple targets with a 10d8 spell at 20th level, the Gambler is not). In an interesting coincidence, by level 20, the Sorcerer can throw that bonus fire bolt while also Empowering a 10d8 spell six times, just like the Gambler. Except the Sorcerer can then use a Bonus Action, burn a third-level spell slot, and do it again next round, too.

Incidentally, if I had to "fix" Royal Flush, it would be to rip off the Sorcerer. The Gambler can re-roll the matching dice, if they're all 1's, all 2's, or all 3's. If they're higher, the Sneak Attack damage is already high enough and the Gambler doesn't need the help. This can help a Gambler with a poor Sneak Attack result, but cuts down the massive overkill of a Sneak Attack All-In adding to an already high damage output.

What if I read it wrong, and all the Sneak Attack dice are re-rolled?

Well then, the damage is way too high and the ability is broken beyond repair. Baseline it's not bad, you have a 5% chance to "Sneak Attack Crit", as listed before, and the overall average damage gain per Sneak Attack is +2. But as also shown before, All-In on a 10d6 Sneak Attack has a 50+% chance of doubling, and that boost is both too high and also too common.

It is intended to be all the sneak attack dice. I did the math as well and came to the same rough probabilities you did. 5% chance to get an extra 9d6 damage on a normal sneak attack isn't all that crazy for a level 17 ability. Pairing it with All In gives you a much better chance of getting the bonus sneak dice, but it's not guaranteed and still comes with a significant penalty if you fail. In the craziest scenario, you go All In on a paralyzed target at level 20 with 20 CHA, letting you get 60d6 extra sneak attack dice. That seems insane, but again compare that to what a level 17 Action Surging Fighter with Great Weapon Master is doing to that same paralyzed target, and it becomes a lot more reasonable. If a level 20 rogue is sneak attacking a paralyzed target, it SHOULD explode.

As for just regular usage, 5% chance for Royal Flush is actually worse than the increased crit range from the Champion fighter subclass, since it it's effectively doubling the crit range, but only doubles sneak attack dice and not all dice. But it's fun to see when things line up.

One last suggestion: Know When To Fold Them should not work if the attack outright kills the Gambler. There's "lucky" and then there's "shot in the skull and didn't die".

Yeah, this is a fair point.

[Reevh]

I like the concept too, very cool!

How about for Poker Face -

As a reaction, any Insight checks against you automatically fail. The person rolling the Insight check is not aware of this effect. This effect lasts for one minute. You can do it a number of times per day equal to your Proficiency bonus.

All In needs a huge risk to balance a huge reward, like the target inflicts a critical hit on you if you miss with All In.

Royal Flush is brilliant thematically - I would have it be 1-2-3-4-5 on the dice but that's way too easy to do if you get enough dice.

You could also add a few Gaming Set proficiencies as bonus skills - the Gambler should be a skilled player of different games, even without the gambling.

I think losing your reaction and bonus action for a round is a fairly significant risk. You'd lose uncanny dodge, effectively doubling the damage an enemy does to you once that round, and you'd lose Cunning Action, which can get you out of danger, or an offhand attack which could help you still get sneak attack damage if you missed with the mainhand. I feel like it's a risk that's in line with the reward.

Adding gaming set proficiencies is a very good idea.

[Breccia]

Pairing it with All In gives you a much better chance of getting the bonus sneak dice, but it's not guaranteed and still comes with a significant penalty if you fail.

It's true that it's not guaranteed, but it's also true that it happens the literal majority of the time. 54% or so. Keep that part in mind.

[Reevh]

It's true that it's not guaranteed, but it's also true that it happens the literal majority of the time. 54% or so. Keep that part in mind.

Yeah, but for a level 17 ability that requires you to use one of 6 uses per day at that level to do, I'm kind of OK with that.