Spellbook.

Phil Walker-Harding has designed some of my all-time favorite games, including Cacao, Silver & Gold, Super Mega Lucky Box, Imhotep: The Duel, Sushi Go!, Gizmos, and more. He’s been on something of a cold streak lately, unfortunately, with a number of games that felt unfinished or insufficiently tested, and it continues with his most recent big release, Spellbook, a game with a decent concept that ends way too quickly.

Players in Spellbook get a set of seven cards, each of which shows three ‘spells’ on it (with rare exceptions) that players can cast to gain additional powers throughout the game. On your turn, you may collect spell tokens in the seven colors, and if you collect at least three of a color, you can cast the associated spell, gaining either an immediate bonus or a new action for the rest of the game, plus victory points, but losing the ability to cast the other spells on that card. You may also store a token on your player board on every turn, with spells that allow you to store two or more if you cast them, which awards you the most points and also triggers the end of the game when someone fills all 14 spaces on their board.

In each turn, you get up to three actions, tied to morning, midday, and evening, with all spells fitting into one of those three times of day. In the morning, the base action is to take one visible token from the market or two random ones from the bag. In midday, the base action is to store one token on your board. In the evening, the base action is to cast one spell. As the game progresses, you’ll have better actions available from spells you’ve cast, such as allowing you to swap some of your tokens with those in the market, or allowing you to discard one token with a specific symbol to draw 4 from the bag, so the game speeds up. And that’s the problem: Spellbook ends before you can get anything interesting going at all.

If a player just muddled along and stored a token on every turn, they wouldn’t win, but the game would end after 14 turns, which might be a reasonable number – but the game should never last that long because of the actions available that let you store multiple tokens at a time. The game ends either when someone fills their board, which is the only way we’ve ever had this game end, or when someone casts a spell of all seven colors. I’m pretty confident that the cast-and-store strategy is the dominant one, both because it offers more points and because it ends the game more quickly, but that consistently left us with the sense that we’d barely played the game. Some spells aren’t that useful anyway, but you might cast only three of them before the game ends, and that just isn’t very fun to play. It wants to be an engine-builder, but that would require more turns, and there’s too much randomness involved in getting the tokens you need for spells (with one way to create a ‘wild’ token that’s too difficult to change the calculus). I have a hypothesis that larger publishers in board gaming are pushing to get more titles out rather than fewer, better-quality ones, and this feels like it supports my belief – at best, it just wasn’t tested enough, because there is no way people played this a bunch without saying the game ended too soon.