D&D 5e Multiclass Tier Rankings, Clustered!
The Dungeon Dudes on Youtube recently finished up a series on Multiclass Tier Rankings in D&D (playlist, summary video). The results can be collated to show which combinations are best for multiclassing.
These results reflect Monty’s and Kelly’s opinions - yours [and mine!] might be different
Here are the results as a table (if your character has more levels in class A than B, then you are multiclassing out of A into B, and A is your ‘primary’ class):
| Primary\Secondary | Art | Barb | Bard | Cle | Dru | Figh | Monk | Pal | Rang | Rog | Sorc | Warl | Wiz |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Artificer | 1 | 2 | 3 | 1 | 4 | 1 | 3 | 2 | 4 | 2 | 1 | 4 | |
| Barbarian | 1 | 2 | 3 | 3 | 5 | 3 | 4 | 3 | 4 | 1 | 1 | 1 | |
| Bard | 2 | 1 | 4 | 1 | 3 | 1 | 4 | 2 | 3 | 4 | 5 | 1 | |
| Cleric | 2 | 1 | 1 | 2 | 4 | 3 | 2 | 2 | 3 | 1 | 2 | 2 | |
| Druid | 1 | 4 | 1 | 4 | 3 | 3 | 4 | 4 | 2 | 2 | 1 | 1 | |
| Fighter | 3 | 5 | 3 | 5 | 1 | 2 | 4 | 4 | 4 | 2 | 1 | 4 | |
| Monk | 1 | 3 | 1 | 5 | 4 | 4 | 1 | 5 | 5 | 2 | 2 | 2 | |
| Paladin | 2 | 4 | 5 | 3 | 2 | 4 | 1 | 1 | 3 | 5 | 4 | 1 | |
| Ranger | 2 | 3 | 1 | 3 | 3 | 5 | 4 | 1 | 5 | 2 | 1 | 1 | |
| Rogue | 4 | 3 | 4 | 4 | 5 | 5 | 4 | 3 | 5 | 3 | 2 | 4 | |
| Sorcerer | 2 | 1 | 3 | 3 | 1 | 3 | 1 | 5 | 1 | 3 | 5 | 2 | |
| Warlock | 2 | 3 | 3 | 2 | 1 | 4 | 2 | 5 | 3 | 3 | 5 | 2 | |
| Wizard | 4 | 1 | 2 | 4 | 1 | 4 | 1 | 2 | 1 | 3 | 3 | 1 |
… but that’s not very interpretable.
So, I looked around for libraries which can do 2D clustering visualizations and I came up with d3-force. With it, I can setup some forces between nodes in a graph so that classes which are easy to multiclass together will appear close to each other in space. Of course, with this method we lose information about primary vs. secondary - maybe I’ll try that some other day.
- Classes are color-coded by their primary ability (for ambiguous cases I just picked one)
- Hover over each dot to see the class name
- You can click+drag a dot to rotate the visualization
The results are not too surprising - the clustering agrees with the general opinions of the Dudes:
- The clustering of Sorcerer and Warlock and Paladin is very tight (they are easy to mix)
- Wizard is way out there on his own with only Artificer for a friend (don’t multiclass in or out of Wizard)
- Fighter, Rogue, and Cleric are in the middle (they are easy to multiclass with just about anything)
FYI, here are the forces I used to create the simulation:
const simulation = d3.forceSimulation(nodes)
// keep things centered
.force("center", d3.forceCenter(width / 2, height / 2))
// prevent overlaps
.force("radius", d3.forceCollide(5))
// set distance based on the quality of the multi-class link
.force("link", d3.forceLink(links)
.id(d => d.id)
.strength(d => (1 + Math.abs(d.value - 2)))
.distance(d => (200 * (1 - d.value/4) )));