Skyblivion's delay announcement back in December had one line that stuck with me more than the new 2026 target: spellcrafting and underwater combat were cut from launch, with both promised post-launch as part of a "living project" plan. That's official, straight from the team lead, not a rumour. But sit with what it means. A volunteer project whose entire reason to exist is rebuilding Oblivion feature-for-feature inside Skyrim's engine looked at the spellmaking altar and said: not at launch. Skyrim's tech has no native concept of a player-authored spell. Every spell in that game is a hand-placed designer artifact, and fifteen years later we're all still living inside that decision.
So here's my case for TES VI being where it ends.
For two straight games, magic was a craft you practiced, not a menu you shopped from. In Morrowind you found a spellmaker, combined any effects you'd learned, set magnitude, duration and area, and paid for the privilege. The system trusted you, and we repaid that trust with glorious nonsense. Drain Health 100 points for one second on touch: deletes most things on Vvardenfell, costs a pittance to buy and almost nothing to cast. Jump spells big enough to clear Vivec's cantons in one terrified arc (pack Slowfall, and ask Tarhiel how it ends otherwise). Oblivion moved the altars to the Arcane University and gated them behind guild recommendations, which honestly improved things; earning the right to break the game is good design. And break it we did: alternate two custom spells stacking Weakness to Magicka and Weakness to Fire, and a beginner's fireball starts hitting like the wrath of Mehrunes Dagon.
Yes, broken. Hold that thought.
Skyrim dropped the whole system, and the official line at the time was roughly that assembling numbers in a menu didn't make magic feel magical. Spells should be authored things with identity, cast viscerally rather than configured. Let me state the strongest version of that fairly, because it isn't stupid: spellmaking flattens magic into arithmetic. Players converge on one optimal spell, hand-designed spells lose their specialness, and the wizard fantasy quietly becomes an actuary fantasy.
I don't buy it anymore, for two reasons. Curated-only magic gave us the least interesting system in the series anyway: Destruction that stopped scaling, Mysticism deleted outright, a vending machine where the workshop used to be. And balance was never the line Bethesda actually held. Skyrim kept alchemy and enchanting, and the restoration-potion loop breaks that game harder than any custom spell ever broke Oblivion; nobody proposes removing alchemy over it. Balance in a single-player RPG is a social contract with yourself. The fix for degenerate spells is design, not amputation. Morrowind already had a cast-failure chance, and big magnitudes can be gated behind found knowledge. Make me earn Weakness to Magicka 100 from a Telvanni master's research notes. I will gladly do the homework.
And TES VI is unusually well placed for this. Howard has confirmed it runs on Creation Engine 3, with the upgrade focused on world systems and close-camera detail, and there's a credible report (Windows Central, so weigh it as such) of engine work built for Starfield that carries forward to future games. A spell is just a bundle of parameterised effects; the genuinely hard parts are VFX, UI, and the nerve to let players combine things. If this generation's engine pitch is systems, spellcrafting is the system. Howard has also framed TES VI as a return to classic Bethesda exploration RPGs. Fine. Then bring back the classic toy.
Even the books argue for it. Malviser's Response to Bero's Speech defends destruction magic as an expressive art rather than a blunt instrument, and that's the heart of this for me. Spellmaking is the only crafting system in these games that lets you author a verb instead of a number. Smithing makes your sword hit harder. Spellmaking makes a thing that didn't exist before, with your name on it.
There are older threads here on TES VI magic and feature wishlists generally, so let's keep this one narrow. If spellmaking came back with exactly one hard restriction to keep the designers sane, cast-failure chance, magnitude caps, research-gated effects, or no stacking weaknesses, which would you accept, and what's the first custom spell you'd rebuild?