Review: Shivers 2: Harvest of Souls

Sierra (creators).

Review written by Andrew Plotkin

Pretty good
Not bad
Very good
Good overall, but with major flaws
Writing and dialogue
Pretty lousy, although the singers can sing, and a couple of the melodies aren't bad
Forgiveness rating
You can't get stuck; you can kill yourself by making some mistakes persistently, or by taking too long in (a few) timed puzzles

I knew what I was letting myself in for. Shivers was one of those unapologetic puzzle-stacks of the early age -- 1995. The sequel was two years later, but did I expect Sierra to have made huge sweeping strides? Heck no.

Stack of puzzles, no apologies in sight. Check. Let us consider details.

The story is more complicated, but I can't say it gains much by it. Your band has come to Cyclone, kneecap of nowhere, to film a rock video. In fact, they've already arrived; you got into town late. But they've disappeared. In fact, so has everyone else.

So you begin looking around -- there's hardly anything else to do -- and find that 1) the town is a little desert paradise of corruption and grumpiness; 2) people were already disappearing when the band arrived; 3) the band planned to investigate the mess and put clues (bear with me, okay?) in their rock videos; 4) another stranger in town was helping them with video; 5) a mysterious figure in a Kachina ceremonial mask was helping them with funds; and 6) two, count them, two mysterious ghosts are trying to give you clues.

I list all this by way of saying, for heavens' sakes, it is possible to have too much storyline in a game.

Yes, okay, a really good writer could have put it all together. Pity about that.

I don't even mind, much; a story was hovering around somewhere. It's certainly no worse than "Here's a map, go find the twelve sticks, or meeeeet your dooooom. Hop to it."

As in Shivers, the game world is goofy; but in a boisterous sort of way which is hard to dislike. I was actually reminded of the circus games I've played (Ballyhoo and Bad Day on the Midway) -- colorful characters, larger-than-life, practically trailing spotlights; all inbred small-town seediness on the inside. The scenes are drawn in garish hues. Somehow it all works. Nothing sophisticated, but it gets the mood across.

The game action, as I implied, is finding the twelve Anasazi prayer-sticks. Each is guarded by, surprise, a puzzle -- arranged by some ghost or other -- your training on the path to become "the warrior". (Er, never mind. Hop to it.)

As in Shivers (again), the designers use hit points as a pacing tool. A few places in the game hurt you if you investigate them without proper precautions. Mild bops on the nose; you won't lose the game unless you open the wasp-filled box ten times in a row. The prayer sticks are where the damage system really kicks in.

I'll pause to go into detail here, because I find the example interesting. In my opinion, the designers did three things right and one thing wrong.

When you find a bahos (prayer stick), it immediately starts to drain the life from you. You have to head to Devil's Mouth Canyon, the spookiest spook-filled place in town, and put the bahos into a hidden altar. But wait: first you must pass through the canyon (filled with fire-throwing carved spooks), find the appropriate carving, and solve a puzzle there to activate your stick.

The life-draining curse of the bahos actually works very well, in plot terms. The damage is very slow. As soon as you touch your first stick, you get plenty of cluing -- outright instructions, really -- about what to do with it. Walking across town takes only a few seconds the first time; after that, you can use the in-game map and click straight to the canyon.

So the draining is not a serious threat to your game session. Instead, it keeps the storyline moving in order. You find a stick, you go to the canyon and stash the stick, you go find another stick. (Various plot elements accrue as you make progress.) You can't throw the timing off by finding five or eight sticks and never doing anything with them. There's no time to explore when you're holding a bahos. Stash it first. See?

On the other hand: the spooks in the canyon do heavy damage. Heading for the canyon isn't an urgent matter, but searching the canyon for that activation puzzle is. You have a minute or two, at best. And that doesn't work. Every player I know of simply saves the game, searches, and then restores and heads straight for the carving. The time limit is no longer a plot reminder; it's just a way to force the player to save and restore the game. How does that add to the game? (Answer: it doesn't.)

Back on the up side... while you're actually solving the activation puzzle, you take no damage. No time limit there.

And finally, when you stash the stick, you get a healthy boost of, um, health. More than enough to recoup all the nose-bops and life-draining you took in finding that stick. (But not enough to recoup the fire-blasts in the canyon.) So, aside from the one save-and-restore, your health isn't a long-term problem. You aren't likely to be stuck near the end because you failed to husband resources near the beginning.

I was hot-and-cold about the rest of the game design, too. Puzzles are intertwined with each other, and intermixed with clues, in an agreeable way. You can tackle sections of the game in almost any order, while still finding clues and information that build on each other. The result feels like a story -- well, a rising plot-line, at least -- even though it's different for each player.

However, some of the tricks used to pull this off don't work so well. Some puzzles aren't solvable until you find the explicit clues you need to solve them. The designers may have thought nobody would guess their clever ideas; but, of course, they were wrong.

Yes, it's not good to stumble across a solution when you have no idea what you're doing. (I did that just once, in this game.) And it's definitely bad to be able to solve a puzzle by brute force, before you find the clues which allow a clever solution.

But it's also bad to prevent a solution from working, without rhyme or explanation, when the player has glimpsed half the clues and intuited the rest. And that happened to me a lot. Several times, I went to a walkthrough, only to find I had to re-try ideas that I had tested already -- too early in the game.

One such case, so egregious that it achieves its own category: You find a chessboard in one building, with a clue saying to find four hidden things. Fine. Four chessmen. And yes, there are four chessmen hidden around the town.

But they don't appear until you've seen the chessboard. If you explore the town in the wrong order, you will have looked in all those spots already. I was stuck on that puzzle until I noticed something new in an explored location. And then I had to spend half an hour systematically re-exploring the entire town. Was this fun? (Answer: it wasn't.)

To make the scenario even goofier, I actually had found the chessboard very early in the game. But I hadn't read the note attached nearby. And that note is the trigger. If I had just found four chessmen, I would have brought them to the board immediately; but the game didn't judge me worthy of that.

Not least, the annoyance is because they sometimes handle it right. One combination lock can't be brute-forced, because it has a crucial piece missing. The piece is found in the same location as the combination. No way to get ahead of the plotline -- simple. When you find the combination, you still have to figure out what lock it goes to; the missing piece looks pretty generic. So it's still a combination puzzle, not a lock-and-key. Remember that design pattern, designers.

I'm spending a lot of time on the puzzle-design, and you already know the reason for that. But I should add one more note. The music videos.

They are, as promised, full of clues. As it happens, I didn't figure out how to watch them until the game was just about over. (As it happens, I didn't figure it out at all -- I had to get a hint -- but that was just my particular blindness, not a major design problem.)

Anyway, I solved most of the game without the videos, so most of the clues must not have been that vital. Some, however, were crucial. And I don't know how I feel about that. Normally, I like the puzzle-book aesthetic -- a surreal experience with clues fair game anywhere and everywhere. Shivers 2 isn't a particularly surreal world, but the world of music videos is the Surreal Poster Child Universe, right?

It didn't work for me. Maybe if I'd spent the whole game staring at them, instead of flipping through them once at the end. Maybe if the video window had freeze-frame, rewind, and slow-forward buttons. (You can linger over a puzzle-book.) I think I could have liked those parts of the game. But, um, it didn't happen.

And now, I get to talk about the interface. I've been waiting the whole review to talk about the interface.

Not the game-playing interface, which is a bog-standard panning game with inventory. No, I mean the rest of the game. The startup screens, the load and save menus.

Have you seen the announcements of Apple's "Aqua" UI, which will be released with MacOS X? The feature that keeps striking my attention are the window title buttons (close, maximize, minimize), which are unmarked colored spots. Until you move the mouse over them, at which time they acquire icons.

When you start up Shivers 2, you see a five unmarked spots. That is, they have faux-Indian decoration, but nothing that indicates function.

When you move your mouse over them, labels appear. And disappear, when you move your mouse off that particular spot. One label at a time is all you get. To look over your list of choices, you have to scan past the whole row, and then remember what you saw.

Does it get better? Of course. Your initial options are "New Game", "Old Game", "Web Page", "Credits", and "Quit". No problem there. You start playing. You play. You want to save your game. The four game-menu options are "Internet", "Flashback", "Options", and "Map", so you guess it's under "Options".

You are now again faced with five unmarked spots. These turn out to be "Play", "Old Game", "Configuration", "Save Game", and "Exit". This also turns out to be the point where I nearly put my fist through the screen. Note that the opposite of "Save Game" is "Old Game"; "Exit" exits the program, not the menu; and "Play" means "return to the game", not "play a saved game" or "play a movie you've seen" or any other possible meaning.

Yes, of course these labels make sense, after you think about it. They're still completely stupid. Did I mention that "Exit" quits the program without a confirm/deny dialog? Bet somebody's lost progress to that.

It continues getting better. When you choose "Save", you get a one-line dialog box in which you can type a game name. But if you type a name which you've already saved, the earlier game is silently overwritten. And there is no way I can tell to cancel out of this dialog -- if you click anywhere outside it, the game silently overwrites the last game you saved.

Oh, yes, you're limited to 24 saved games. (I'm conservative and hit that limit easily.) If you try to save a twenty-fifth, the one-line dialog box appears as usual, but after you type your game name, it laughs in your face and tells you to delete some old ones. How do you do this? Why, go into the "Old Game" screen, which actually lists all your saves; go up to the three unmarked spots at the top of the screen, and -- whoops, two of them don't have labels even when you roll the mouse over them. The third is "Options", which returns to the unmarked options screen.

I eventually figured out that the "Play" and "Delete Games" buttons don't work until you've selected a game.

Please don't ask my why "Delete Games" is plural, since you can only delete one game at a time.

Suffice it to say that unmarked, tooltip-only buttons are not only bad, but real bad. Pig-biting bad. Laughably bad. Zombies mock them at midnight. Hibernating toads invent better UI in their sleep. Unnamed spawn of Azathoth and the Darkness stop by Earth to say, "Ooh, what a bad idea that is." Glaze-eyed graphic designers, flying past Neptune on the worst Canadian ganja ever scraped out from between the rocks, invent unmarked button interfaces and immediately take holy orders under vows of silence, lest a vengeful populace drag them to Hawaii in search of active volcanos.

Also, my final score was some 850000 out of 800000. Go figure.

On the whole, I really did enjoy Shivers 2. I played most of the way through without using hints (or the "skip this puzzle" button inexplicably hidden under "Configuration"). Then again, the very flexible nature of the game meant that I had just pushed the puzzles I couldn't solve to the end. Keep a walkthrough ready, be sure to save before entering the canyon, and you should be well-occupied for a day or so.

Availability: I saw it in Best Buy for ten bucks. I have no idea where you'll see it.

System Requirements: It's an oldie; wants a 486DX/2 and twelve megs of RAM. Virtual PC had no trouble with it. My biggest problem was that the CD, molded before 24x speed drives, vibrates like heck in my Powerbook.

(I didn't test the Internet linked-play features.)

Game Reviews