Played some more research mysteries
After finishing The Roottrees Are Dead, I played a couple more of these. Here are my thoughts on two!
If you want to go into these games completely fresh, though, I recommend that you do not read this post. While I don't expose the answer to any mystery, I do discuss my strategies for solving these games in ways which may reveal a strategy you'd be happier to discover yourself.
No Case Should Remain Unsolved
One of the better mystery stories I checked out - it relies on some themes which provide good resonance between its story and its mechanics. Out of all these games, I probably admired the mystery itself in this one the most. I wish I could say more, but it's best for players to go in blind.
On the other hand, I wasn't a huge fan of the UI. The game displays conversation fragments as movable "cards". The UI divides the story into timeline columns, each of which is assigned to a different character your police officer protagonist once interviewed. Your failing memory mis-assigned many of these conversation cards to the wrong column, and you earn progress by re-assigning them to the correct speaker. You also gain progress from ordering the conversation fragments in the right order. When you're reordering cards within a speaker's column, the timeline UI kind of forcibly enacts a "lawnmowering" playstyle on you, no matter what.
Lawnmowering is when a player makes random or poorly-supported, vague guesses over and over until they happen upon the solution randomly. It can make a mystery game both less rewarding and more confusing, since the player can make progress without really understanding the mystery. The reason Obra Dinn and The Roottrees Are Dead require guesses to be submitted in large batches (in the case of Roottrees, containing as many as 15 individual bits of guessed data at a time!) is to avoid the player lawnmowering their way through the mystery.
But NCSRU doesn't have batching. When you swap a card into another character's column in NCSRU, it's always placed at the bottom of the stack. Clicking an up or down arrow on the card will swap its place with the very next card above or below it - so you're never placing cards decisively on the timeline, only shifting them between columns and then up and down within each column. When two cards are in the right order, they "glue" together. This means that the act of moving the card up and down in the stack enacts "lawnmowering" without the player really meaning it to - you can glue two cards together without really choosing to have placed them side-by-side, without understanding the relationship between them, without having recently read either of them, or without even having moved either of them - maybe you were moving the card in the middle out of the way instead, and glued the two cards you didn't interact with at all!
It almost feels like the game wants and accepts gameplay where the player just slowly shifts each card up the stack, from the bottom to the top, one by one, until they glue with their above/below neighbors. I definitely did this to a few characters. It's so easy to do that it almost feels intended.
I'd prefer if it was possible to remove the cards from the timeline UI completely, perhaps spawning many of them in a separate "inventory" of unplaced cards. I'd also probably remove the shifting UI verb entirely, so that the player is required to place each one down in what they believe is the correct spot. The corpus is so small that it would still be possible to do a lot of lawnmowering this way... but at least it wouldn't be required of the player.
I have some other quibbles with how the keyword locks work in the game - "blue" locked cards can open right away when you click the correct keyword elsewhere in the corpus. However, each instance of that keyword is usable only once. By the end of the game, the player may have a few blue cards which they need to open... but they have to scan the entire text of the game to find an instance of the keyword which hasn't already been "spent" opening another card. I wish it had been possible to do a reverse keyword search, and find the location of a keyword during the late game. Maybe some unlocked ability for the later half of the game only?
Otherwise, I really enjoyed this one. The story itself was a wonderful mystery and it felt extremely deftly and confidently constructed.
Rivals
I played this one even though I am not particularly knowledgeable or a fan of indie rock or alt country or music journalism, and this seems like a game for people who know and care about one or more of those things. However, the story does have hooks for players like me. The personal conflict between the two titular rivals is relatively well-drawn in the space available for it.
The space available is very small. This is very short game - probably a single evening for most players. I love that! I always say I want shorter games that fit into my life, and this is exactly that. However, there's not a lot of meat to sink your teeth into here, and it's not really a mystery. It's pretty frank about that - the story does not contain a mystery, and the stakes of the game are not the stakes of a mystery. If you were looking for that, it's just not here.
I'm OK with that... an FPS might just be a mechanical excuse, sometimes, to explore a physical space. In the same way, research mechanics might just be an excuse to explore a story.
The game asks you to order the chapters in a book onto a chronological timeline. Every year from 1997 to 2009 has one or more slots for book chapter titles, and each title represents something in the story of interpersonal conflict between two famous musicians. When you get five in the right spot, the game will confirm your correct choices - and leave your incorrect ones untouched.
You can also place a chapter title in more than one spot on the timeline at the same time.
These two facts - that you can over-apply a guess to many slots, and that the incorrect placements won't count against you - means that the most efficient (and as far as I can tell, intended) strategy in this game is to apply a chapter title to as many places as it could possibly fit at once, then start futzing around until you have at least 5 different opportunities to be correct. When I saw that someone died in 2003, I applied that chapter title to every one of the month slots in 2003. When I saw that an event occured in March, I put that event on every available March in the game.
I don't mind that the puzzle invited me to do this kind of... limited? tactical? lawnmowering. However, it was possible to reuse this strategy 3-4 times during the short course of the game... and every time I used it, it meant that I was not engaging very closely with a lot of the clues. I knew that it wasn't worth trying to really deeply understand some of the events in the story if I found myself feeling stuck... I could just finagle my slate of guesses in such a way that it was possible to reuse this more shallow strategy.
I enjoyed listening to the audio interviews where characters are talking about their experiences to music journalists. On the other hand, however, I don't think I ever listened to one of the 8 singles they produced for this game all the way through even once. The interviews were really, really dense with clues, but the main clue associated with each music track is simply what genre it is in. I was never required to listen to lyrics or even read lyrics to solve a mystery in the main plot.
I wonder if this decision was made to make the game more accessible to people like me - people who don't have a special interest in indie rock or alt country. On the other hand, I kind of feel like if you're going to make your interview audio dense with clues, it might have been possible to do similar things with at least some of the music.
Finally: I was VERY impressed with the dense, utilitarian UI. At the end of the game, every piece of evidence in your inventory fits on screen all at once. When you make guesses, it's on the same screen. Everything is in front of you at all times. It really helps a game like this when you can scan your eye over a lot of the evidence at the same time! I was so impressed by how usefully presented and juxtaposed all the evidence in the game remained, no matter how much of it I unlocked.
I think there's a lot of room in the world for 2-3 hour mysteries that all fit on the same screen. I've picked at this game a lot in this post, but I think that it's precisely the size in needs to be to fulfill the role it wants to fill.