I'm wondering what the impact of a house rule that would allow players to choose which subclass to get features from whenever they would get a subclass feature. For example, a wizard could get Divination Savant and Portent (from School of Divination) at second level, and then Potent Cantrips (from School of Evocation). However, features that increase when you gain other subclass features (for example, the Battlemaster Fighter gets more maneuvers, and more and bigger Superiority Dice) only improve when you take a feature from that specific subclass.

I think this would allow min-maxers to break the game, but this is only a guess. I'd like an answer that gives at least one example of a combination that is much more powerful then the current options.

Note: this will not permit players to get subclass features from other classes.

  • 3
    \$\begingroup\$ Is there a reason you want to do something like this? \$\endgroup\$
    – AnnaAG
    Feb 20 at 9:29
  • 6
    \$\begingroup\$ I've voted to reopen, since you made it more focused, but what are you trying to accomplish here? See What types of questions should I avoid asking?. This feels a lot like "What if _____ happened?" It's hard to tell if you're a DM looking for more options for your players (I'm guessing not), or if you're a player looking to build to justify a build to your DM (maybe?), or if you're just poking the rules and saying "what if _____ happened?" I think your question could be improved if you said what you're digging at. \$\endgroup\$
    – Jack
    Feb 20 at 18:01
  • 2
    \$\begingroup\$ I think after the update this is easily answerable now. Asking for at least one example is a definite criteria and one that is easily met. Gloomstalker level 3 + any number of combat relevant later bonuses from other subclasses easily surpasses the strongest current Ranger in many campaigns (which is typically just a base Gloomstalker). Rogue Assassin which can take Scout level 13 (Adv on initiative rolls) instead of a ribbon pushes the limits on what Rogue ambush builds can do, etc. \$\endgroup\$ Feb 20 at 18:01
  • \$\begingroup\$ Also, do you intend to allow multiclassing at your table? If so, abuse cases could expand further, and players would be left with a dizzying number of choices at level up. Or is this intended to replace multiclassing while giving characters more build options? If so, asking for a direct comparison between multiclassing and this option might be a more direct way to understand the impact of this option. \$\endgroup\$ Feb 20 at 18:06
  • 1
    \$\begingroup\$ @Jack one example might be useful to show players why you, as DM, refuse to implement such things. It's not the first time I heard this idea. \$\endgroup\$
    – Mołot
    Feb 20 at 19:39

1 Answer 1


This Will Lead to Power Creep

Whether or not it "breaks your game" depends on what you already allow at your table (unlimited multiclassing vs "story justified" multiclassing vs. no multiclassing, feats, etc.), but it definitely allows single class characters to reach heights they otherwise couldn't if they are specializing in one thing (i.e., min-maxing).

There are many different things a PC can optimize their character for. One of these is "Ambush" damage, or damage in the first round of combat (preferably against surprised opponents). I will focus on this as one relatively easy to understand example and a couple of subclasses, but the problem absolutely exists for almost any optimization objective you can imagine (for any class that has subclass features related to that objective), for many classes.

Both Rogue and Ranger's can be built using these rules to be far more effective at round 1 damage (especially surprise rounds) + combat generally than any individual published subclass of those respective classes. The features they give up (while thematic) are largely worthless for this optimization target and are thus precisely the types of trades a min-maxer would look for.

Example 1

Class: Rogue

Race: Harengon (+ Proficiency to initiative)

Subclass level 3: Assassin's Assassinate (automatic crits against surprised enemies; advantage against anyone who hasn't taken a turn yet).

Subclass level 9: Phantom's Tokens of the Departed (even without the ability to use level 3 feature Wails of the Grave, this gives advantage to 2 types of saving throws + out of combat utility)

Subclass level 13: Scout's Ambush Master (Adv. on Initiative + you and all allies have adv. against first creature you hit during first round until start of next turn)

Subclass level 17: Thief's Thief's Reflexes (2 turns in first round of combat, 1 at normal intiative + 1 at initiative - 10)

Compared to Assassin, this build has an overwhelming combat advantage. It replaces 2 ribbon/flavor abilities at levels 9 and 13 with combat utility: It more reliably goes first (thanks to Ambush Master) allowing it to trigger both parts of the strong Assassinate level 3 feature more regularly, it shores up some weaknesses by giving the Rogue the ability to have persistent advantage on Constitution and Death saving throws, and it even has some supporting qualities buffing party damage via Ambush Master

Thief's Reflexes further improves round 1 damage in multiple ways:

  1. If your normal surprise attack crit damage is enough to one shot an enemy, you may take out a second enemy with your second turn. This doesn't improve theoretical DPS numbers, but practically is invaluable vs. only being able to double damage vs a single enemy.
  2. If an enemy is particularly hard to hit (either naturally due to AC or else has a reaction or spell that helps them avoid your first attack), having a second chance to deal damage before they take their first turn in the event of a miss salvages your ability to assassinate them.
  3. No constitution saving throws required. You always get a second turn in round 1.

You could also substitute in Scout's Sudden Strike to be able to use your auto-crit sneak attack twice at your normal initiative (instead of waiting) and allow your Ambush Master debuff to last longer, but you give up the ability to sneak attack the same target twice if that's important.

Compared to Scout (who also has features that help with surprise round attacks) you are also up multiple combat relevant features, and in particular Assassinate is a massive damage boost beyond what a normal scout could achieve without it.

Example 2

Class: Ranger

Race: Bugbear

Subclass Level 3: Gloom Stalker's Dread Ambusher + Umbral Sight (Permanent invisibility vs darkvision, extra attack w/extra damage round 1 of combat, bonus to initiative)

Subclass Level 7: Horizon Walker's Ethereal Step (Ability to sneak/slip past even enemies who can see invisibility or have blind sight to set up ambushes)

Subclass Level 11: Hunter's Multiattack (unlimited targets within 10 ft of a point) or Horizon Walker's Distant Strikes (Extra attack as long as you attack--not hit-- multiple targets w/attack action, no radius limit)

Subclass Level 15: Horizon Walker's Spectral Defense (Defensive Reaction you can choose to use only if you're actually going to be hit by attack, halve damage taken)

Compared to Gloom Stalker (who gets a weaker version of extra attack at level 11 that salvages one miss, which hopefully doesn't happen much when attacking with advantage), you replace it with a much stronger version of extra attack (either unlimited within radius, or a third attack with conditions + also 10 ft teleport per attack). You also replace a purely defensive Subclass level 7 ability with one that can set up ambushes and help you get the surprise round you rely on. Level 15 is the least impactful choice, but I'd still prefer a reaction that I can choose to use after knowing whether I'll be hit or not vs. one I use early without knowing (especially if I have other uses for reaction).

Compared to Horizon Walker (who we've stolen multiple features from), we stack a bonus to initiative + extra attack in turn 1 + extra source of advantage, all while keeping the same amount of extra damage the Horizon Walker subclass feature would provide (1d8) AND dropping the requirement to use our bonus action for the damage.

Compared to Hunter the level 3 feature is on another level: Dread Ambusher + Umbral Sight provides another additional attack, source of advantage, match the damage bonus (for round 1), and also gives an initiative advantage.

The ability to both add absurd numbers of attacks (Via Hunter's Multiattack) and also an easy source of advantage AND initiative (Gloomstalker's level 3) combines insanely well with Bugbear's surprise attack feature to weaken whole armies (usually focusing down 1 enemy at a time is ideal, but if you have a wizard following up with an AOE or similar, this can be an effective party tactic).

Multiclass Note

If you do allow multiclassing, these options don't disappear. You can take the upgraded Ambush Rogue to level 17 and stack Gloomstalker on top.

Or you can take Gloomstalker to 15 and add Assassin in as well.

In both cases, you've got the the full power of the "flexible subclass" house rule adding to the shenanigans multiclassing already permits, not just as an alternative to it.

But does it Break the Game?

If you have a specific thing you're worried about, asking a more targeted follow up question including what you consider "game-breaking" could give a specific answer to that question. It would require details like the level range you anticipate your parties playing, whether it's a published adventure with known enemies, whether you are allowing common optional rules like feats or multiclassing, etc.

If you have a very clear target like "could a party of level 7 trivialize the final boss of campaign X with this houserule", "could a single level appropriate character solo the final boss of campaign X with this houserule", or "Could a character achieve higher resourceless DPS with this houserule than under typical multicalssing rules", that type of question would be easier to answer definitively than a general question about "game breaking" in an abstract sense.

  • 2
    \$\begingroup\$ Excellent answer, I do like to see worked out examples that show the truth of a good summary at the top. \$\endgroup\$
    – Phoenices
    Feb 21 at 0:22

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .