D&D 5E Bladesinger - a criticism of its design

[clearstream]

But, again, the 1 in 2 metric is a poor assumption. Even going to 1:3 days the 18I bladesinger is taking 9 hits, and 1:4 10 hits. That burns out all 3rd slots. AND the fact that double the previous number of hits for 23 AC cannot be shielded at all. If we go for the 1:10 days, the takes 12 hits.

Remember we're not arguing that BS is invulnerable: this is about relative strength. Look at the numbers: what happens to Defense Champion (AC 21) or GWM Battlemaster (AC 18) on the same basis? They're already destroyed by the 50% point. This reflects my playtests: GWM Battlemaster can't reduce the number of attacks taken enough to survive. I'm going to tidy up these PDFs and then post them here so that others can look at exactly the numbers we're seeing. They clearly show that BS survives a greater percentage of days than the martial characters.

No, which is why I insisted that if we're going to compare how a bladesinger can tank, it needs to be on a per hit basis, as all the other things that confound, including scenario and actions taken biases, will result is skewed representations. For instance, it's possible to kill the giants effectively without a tank character at all -- in your scenario you're already kiting with 3 characters, so why not kite with all 4 and really get in some advantaged DPR?

If you are dissatisfied with the party composition: propose one and we can test it. If you are dissatisfied with tactics, describe the tactics you want applied. Same with the scenario. If you can't propose a scenario in which BS doesn't shine, why should your claim that there is one feel compelling?

What offensive battlemaster? An offensive greatsword weilding battlemaster with GWM, defensive style, and plate is AC 18, +8 (+3) to hit, and deals 2d6+15 per hit, with precision to boost hits (which are at 50%). The DPR output over the bladesinger is (assuming round 1 cast by bladesinger and javelins by battlemaster) almost double (110 to 65). The battlemaster accounts for a giant by himself in 4 rounds, even opening with javelins.

Unfortunately that character doesn't stay standing. They get in some good swings, and fall.


To progress the discussion, do you concur that we do not need to show that BS is invulnerable? What we need to do is look across the cases, and see that more of the time the party with the BS does better than the party with the martial (assertion 1) and at least as well as the party with the Diviner (assertion 3). Thus, at whatever point we can see that other martials are destroyed by, if BS survives past that it is stronger.

 

[Ovinomancer]

Remember we're not arguing that BS is invulnerable: this is about relative strength. Look at the numbers: what happens to Defense Champion (AC 21) or GWM Battlemaster (AC 18) on the same basis? They're already destroyed by the 50% point. This reflects my playtests: GWM Battlemaster can't reduce the number of attacks taken enough to survive. I'm going to tidy up these PDFs and then post them here so that others can look at exactly the numbers we're seeing. They clearly show that BS survives a greater percentage of days than the martial characters.

Which goalpost are we driving for? Is it 'bladesinger can live' or is it 'bladesinger gets to have all of their 3rd level slots open so they can do everything without problems?' The argument you just responded to was questioning the latter assumption, not the former. And, recall, your bladesinger is only surviving because the cleric is sole buffing the bladesinger and holding all other resources available for healing.

If you are dissatisfied with the party composition: propose one and we can test it. If you are dissatisfied with tactics, describe the tactics you want applied. Same with the scenario. If you can't propose a scenario in which BS doesn't shine, why should your claim that there is one feel compelling?

I already offered a fix -- don't use the battlemaster archer in comparisons to the champion party -- and you ignored it. And, again, I still view all of the playtesting you're doing as anecdotal. You still haven't provided clear details like the space the encounter takes place in, lighting assumptions, encounter start distance, party formation at encounter start, etc. And, to be clear, I'm not necessarily interested in those things because they just provide more anecdotes, not data.

Unfortunately that character doesn't stay standing. They get in some good swings, and fall.

Weird. I just finished running SKT, where numerous encounters with giants occur, even whole complexes of them, with a paladin sword and board and a raging, advantage taking barbarian up front, with a rogue and warlock behind them, and, strangely, they did great. I run tough encounters, too. There were some near scrapes, and I pulled every resource out of the party before they finished those strongholds, but they lived, in game, using player tactics to minimize the effective frontage of the giants and deal with them.

Heck, at 6th level that group took down 3 stone giants in an open field encounter at 60' starting distance without any kiting at all. And there was no GWM for any character, the barbarian had 20 CON and 14 STR, and the paladin was sword and board - both closed to melee throwing javelins.

To progress the discussion, do you concur that we do not need to show that BS is invulnerable? What we need to do is look across the cases, and see that more of the time the party with the BS does better than the party with the martial (assertion 1) and at least as well as the party with the Diviner (assertion 3). Thus, at whatever point we can see that other martials are destroyed by, if BS survives past that it is stronger.

To progress the discussion, please pick a goalpost: the example bladesinger can tank and retain all of her 3rd level slots with reliability or the example bladesinger doesn't get hit as often as a non-disadvantage causing tank. The latter isn't that exciting*, and the former I'll refer you to my arguments above.

* The bladesinger with 18I has the same AC as the sword and board champion with defensive style, so it's not the AC base that's the issue. No, it's the wizard's ability to provide itself disadvantage that's the big difference. A dodging champion does as well at survival as the bladesinger. So, once again, we're back to the wizard's ability to trivialize encounters through spells rather than the bladesinger tradition ability being that huge. A figher 1, diviner (or any other) 5 has the same abilities as your bladesinger as far as AC and defense and enough slots to power the combo through the day with plate, shield, and style. Heck, a diviner 6 can do it with the appropriate ability penalties.

 

[Tony Vargas]

*My dislike about BS is less to do with its absolute power (which is undoubtedly high) as its capacity to put Wizards into a martial role (the evidence so far showing they overshadow martial tanks).

I'm not sure why that's important. It's not like the martial-role is zealously niche-protected, this ed: Fighter, Paladin, Barbarian, Ranger, Monk, Cleric, & Druid can all melee pretty well; choice of Race can give you heavy armor proficiency; choice of background martial weapon proficiency; everyone's proficiency bonus scales at the same rate...

It's not like 1e where being able to use the best armor (Plate Mail), use a shield, and use the best magic weapon (probably a longsword) you just happen to find, was both a huge benefit, and comparatively restricted by class.

Why should it matter any more that BS gets added to the long list of sub-classes who are melee-capable, at least, why should it matter any more than having EK on the long-list of sub-classes that cast spells?

 

[clearstream]

Heck, at 6th level that group took down 3 stone giants in an open field encounter at 60' starting distance without any kiting at all. And there was no GWM for any character, the barbarian had 20 CON and 14 STR, and the paladin was sword and board - both closed to melee throwing javelins.

Bear totem Barbarian? What Paladin Oath, Rogue Archetype and Warlock Patron and Pact? Any notable feats?

 

[Ancalagon]

Is a dodging champion a viable tank if he isn't threatening an doing damage?

Sent from my SM-G930W8 using EN World mobile app

 

[Rhenny]

Update...in the Phandalin campaign I'm in now, I'm a 4th level Bladesinger. I can get up into melee, but I don't like staying up their for too long. I've just taken mobility so I can dance around a lot, especially when I have bladesong up, and I can pop a shield spell sometimes when I want to avoid some incoming damage. I save 2nd level slots for Rope Trick so that my party can always find a safe place to get a short rest or hide. Sometimes I'm in a position to use burning hands, but results vary based on rolls. I can fire a long bow from distance or use Frostbite vs. opponents without spending spell slots (unfortunately, many of the foes I tend to want to use this against have higher Con scores though). I've got some ritual utility (detect magic, identify, comprehend languages). I use mage armor every day, and I hold on to sleep (although I rarely find a good time to use it).

My party consists of a monk (the one that gains temporary hit points from killing foes), a rogue, a tempest cleric, and a moon druid. The others do far more damage than I do in melee. The moon druid can turn into a dire wolf twice per short rest, basically gaining 37 temporary hit points, attacking at advantage with pack tactics, doing 2d6+3 damage and chance for knocking prone. He can also turn back into his Half-Elven form to cast entangle, fairy fire, moonbeam, goodberry, cure wounds, etc. The monk can use ki to disengage or dodge, flurry of blows, etc. The cleric can maximize Thunderwave and blast foes with a retributive strike when he's hit in melee - he can heal too. The rogue gets to sneak attack for nice damage, use cunning action to disengage or hide or dash. It all feels pretty even to me. If anything, it's the druid that seems to be the most OP.

The reason why my character is 4th and the others are 3rd is because we had a near TPK vs. the dragon. With my last party, my bladesinger created a Rope Trick in a thicket near where we encountered the dragon. She tried to get the others to go in it (pretending that she could teleport them away if they came to her location), but they didn't listen. The warlock taunted the dragon and started the fight, then he had trouble climbing the rope so he was first to die outright. The paladin eventually made it to the rope, but then he swung on the rope and tried to do an acrobatic feat to land on the dragon to attack it...pretty soon he was dead. The group's cleric died in poison gas, and the rogue got knocked down then brought up to 1 hp (earlier in the fight when the Paladin crawled over to him and did 1 point of lay on hands healing), then he ran away from the rope trick never to be seen again. The reason why my character lived was partly because she saved spells for contingencies, and partly because she played using her 16 Intelligence rather than trying to do too many crazy things. It had little to do with her AC or her ability to fight in combat.

Often, perceived party balance is all controlled by the type of encounters or the pace of the game session.

 

[Mort]

Update...in the Phandalin campaign I'm in now, I'm a 4th level Bladesinger. I can get up into melee, but I don't like staying up their for too long. I've just taken mobility so I can dance around a lot, especially when I have bladesong up, and I can pop a shield spell sometimes when I want to avoid some incoming damage. I save 2nd level slots for Rope Trick so that my party can always find a safe place to get a short rest or hide.

This is where campaign considerations collide with theory crafting outside of actual play. The rope trick helps the party a lot, but certainly interferes with combat effectiveness.

I wonder how that will change when you hit 5th and might consider Leomund's Tiny Hut (assuming your DM allows it). This will free up 2nd level slots for stuff like blur.

Sometimes I'm in a position to use burning hands, but results vary based on rolls. I can fire a long bow from distance or use Frostbite vs. opponents without spending spell slots (unfortunately, many of the foes I tend to want to use this against have higher Con scores though).

Out of curiosity, why pick frostbite? It just seems, well, lackluster. You might be better served with cantrips that require a hit but have no save.

I've got some ritual utility (detect magic, identify, comprehend languages). I use mage armor every day, and I hold on to sleep (although I rarely find a good time to use it).

Sleep becomes less useful with level, but all mages get hit with that. Ritual utility is nice because it doesn't consume resources, other than time.

My party consists of a monk (the one that gains temporary hit points from killing foes), a rogue, a tempest cleric, and a moon druid. The others do far more damage than I do in melee. The moon druid can turn into a dire wolf twice per short rest, basically gaining 37 temporary hit points, attacking at advantage with pack tactics, doing 2d6+3 damage and chance for knocking prone. He can also turn back into his Half-Elven form to cast entangle, fairy fire, moonbeam, goodberry, cure wounds, etc. The monk can use ki to disengage or dodge, flurry of blows, etc. The cleric can maximize Thunderwave and blast foes with a retributive strike when he's hit in melee - he can heal too. The rogue gets to sneak attack for nice damage, use cunning action to disengage or hide or dash. It all feels pretty even to me. If anything, it's the druid that seems to be the most OP.

You seem to hold your own, even though you split quite a bit of resources into non - combat stuff (no second level spells in combat, when you're 4th Level and that's your top spot, can be rough).

Often, perceived party balance is all controlled by the type of encounters or the pace of the game session.

That's, of course true, though in theory it should balance out over time unless the DM focuses only on certain things.


Thanks for the report from the trenches, threads like this one need that.

Sent from my SM-G930V using EN World mobile app

 

[Rhenny]

Hi Mort,

I just picked frostbite so that I had a different means of attack and I just liked the flavor of it at first. It's something I can do even when engaged without suffering disadvantage and if it hits it grants disadvantage on the creature's next attack. I just like playing with effects. Unfortunately, most creatures that are combat oriented have pretty good Con saves so it doesn't always work. It has saved a few of my comrades though, a couple of times. If I could trade it out, I'd probably pick up firebolt, but that's just damage.

As you can see, I'm not a true optimizer. I like to be effective, but I also make most of my choices based on just plain old character concept. For example, I picked up Misty Step as a 2nd level spell too. I like the idea of being mobile and having escape options. In our last game, our DM was really kind. When the cleric was webbed by a giant spider and left behind us, my bladesinger ran up to him grabbed him with both arms and Misty Stepped 30' away so I could get the cleric out of danger and help us retreat.

I really like playing the bladesinger, but mostly because it is a higher INT PC with utility and lots of choices.

 

[clearstream]

Here is a capture of my current PDFs, comparing sword and board Battlemaster to Bladesinger tanking @Mort's 2 Hill Giants. Hypothesis A = BS outlasts BM using no more than Blur + Shield and a divine buff (both have this, it's more efficient than reactive heals); leaving BS's 3rd level slots free for other wizardry. Hypothesis B = BM's higher damage contribution kills Hill Giants faster, so BM takes fewer hits. Remember that this is relative and indicative. Actual play must supplement understanding.



This first image shows the values where BM's contribution kills both giants a round earlier. Look at the summation column (E) and the damage column. The latter shows expected daily damage at that probability. To take Shield into account, step up the damage column a number of times = shields/day (i.e. removing that many hits). Each character can sustain an amount damage per day indefinitely = hit dice + heals. (Warding Bond divides the damage with the cleric hence 6 HD.) As you can see, with killing both giants a round earlier BM goes past what it can sustain 98% of the time, while BS does so only about 25% of the time (once Shield is taken into account).



This second image shows the same thing, but with BM's contribution killing both giants two rounds earlier. More than halving the attacks per encounter. As you can see, with killing both giants two rounds earlier about 67% of the time BM goes past what it can sustain. I've checked what happens if we go the other direction (BS takes longer, rather than BM shorter) and it shifts BS to not sustaining about 38% (still much better than BM). Further, I've checked that damage never spikes to a level that kills either character prior to that final unsustainable blow (i.e. I've taken into account the significance of HP total as well as HP sustainable.)

As @Ovinomancer pointed out, the culprit here is being able to cast Blur on top of Bladesong and other buffs. Blur is self-only. As @Rofel Wodring pointed out, an Abjurer that can find the right armour can emulate this... at the cost of a Wizard level. Eldritch Knight can do it part of each day at 7th level... but not all day until 13th level. Actual play suggests that these PDFs are pessimistic, and elsewhere we've discussed 5th edition being on an "easy" difficulty setting. So BS might not be problematic in your setting because your characters are dominating anyway. @Ovinomancer attests to a non-BS party prevailing over 3 Stone Giants at 6th level: an encounter far more deadly than 2 Hill Giants.

Greater possible relative strength is not the same as strength experienced at the table. Some players never pick Blur, as @Mephista attested to. Some parties won't have a Cleric, or the Cleric will refuse to cast proactive buffs. I'm confident that Bladesong is not correctly balanced. That doesn't have to impact you at all. For my campaign, it means the sub-class can't be used without revision. I'm not sure what the revision is: maybe Blur is what really needs a nerf?

 

[James Gasik]

On the other hand, something I see a lot of at actual tables is a lack of teamwork. Players build their characters to be island nations, never asking or (or even considering) assistance from the other players, and vice versa. For the Bladesinger to function in a martial capacity, such teamwork is necessary, and maybe that's a good thing.

As for Blur- it's not the spell's fault. Consider that a Paladin or Cleric can get similar results in battle with Protection from Good and Evil in many fights. Indeed, with more hit dice, armor, self healing, and other goodies, the Paladin and the Cleric can outperform the Bladesinger in the tank role...

Well they could, if they could reliably maintain concentration while being the target of the bulk of enemy attacks. Neither class gets proficiency in Constitution saves, and the Paladin doesn't get their aura until level 6. This, I feel, is the issue with the Bladesinger.

Being able to add a primary attribute bonus to their concentration checks is better than anything any other caster gets. Even the Paladin only adds a secondary attribute (Charisma does a lot, but it doesn't affect your attacks) to their concentration checks. If you removed this aspect of Bladedance, then the Bladesinger wouldn't be able to "tank" effectively.

Of course, by removing that aspect, you make it very hard for the Bladesinger to actually enter melee combat at all while using spells, but that's already a position other classes are in. The Bladesinger needs the ability to protect themselves in melee, but they shouldn't be receiving more than other casters get.

Up their AC and give them +1 hit point per level like the Dragon Sorcerer. Give them proficiency for making Concentration checks (you can argue this will eventually be better than +Int, but at lower levels it won't be as egregious, plus you don't have to worry about getting +Int on top of Resilient/Warcaster).

Now if they still prove to be better tanks than Fighters, the real problem will be revealed in sharper focus.