I played Dungeons & Dragons all weekend long, two 6-hour sessions of a new campaign in the Witcher universe. I'm playing an unexpectedly effective bard, specialized in melee combat, less in magic. But the bard remains a hybrid character, between spellcaster and fighter, and that makes the difference between the two concepts very evident.

In 5th edition D&D, fighters and other non-spellcasters have a rather constant output of damage. It is modulated by chance, sometimes you hit, sometimes you don't hit, sometimes you roll high damage, sometimes you roll low damage. But mostly you do the same attacks every round, with a decent average damage. Spellcasters work with a different model: The have a number of spells they can cast per day, and on average those spells are more powerful than the attacks of the non-spellcasters. But once they run out of spells, they are basically reduced to cantrips and less efficient weapon attacks, which deal less damage than the non-spellcasters.

At low levels this is reasonably balanced: The spellcasters have limited amounts of spell slots, so they either preserve them, or run out. That gives the non-spellcasters opportunity to shine, especially in long fights, or when there are lots of fights in an adventuring day with little opportunity to rest. The so-called "linear fighters, quadratic wizards" problem appears a mid-level, and gets worse with high-level characters. Basically the non-spellcasters keep getting a bit better with every level, improving their combat stats, improving their proficiency, or getting additional attacks. But the spellcasters get much better over time: They get both *more* spells, and *better* spells. Instead of running out of spells and having to cast cantrips that are less good than non-spellcaster attacks, they only run out of ultra-powerful high-level spell slots, and still cast lower level spells which are better than anything a non-spellcaster can do. The concentration rule makes spellcasters in 5th edition a bit less overpowering than in editions 1 to 3.5, but they still rule at higher levels.

The Dungeon Master can modify this a bit with the kind of encounters he presents, and how many encounters he puts between two opportunities to rest. Many damage spells are area effect spells, and thus gain in power when there are lots of enemies around, preferably close together. Fights against a single big monster is more of a gamble for spellcasters; they might have spells that effectively neutralize the big boss monster, but usually that involves the monster failing a saving throw, so it is not a sure deal. A big fireball is less efficient on a single big monster than on a group of smaller ones. The number of encounters per adventuring day is a big tricky, because lots of fights one after another fit well into a dungeon setting, but less well into some other settings.

One of the consequences of this linear fighter, quadratic wizards problem is that campaigns rarely reach high level. The disparity in power becomes so annoying that the group falls apart, and then starts another campaign at lower level again. Even many of the official WotC published campaigns end somewhere between level 11 and 15, and there aren't any campaigns getting mid-level players to the level 20 cap. I also like the solution of the warlock class, which has few spells (only 4 spell slots at level 20, compared to 22 spell slots for a wizard), but recovers them more easily, and casts them at high power. But seeing that people disliked the much more balanced 4th edition, it seems that the linear fighters, quadratic wizards problems is here to stay.