Films are, almost by definition, built on three-act structures. This film was interestingly easily divisible into its acts, with severely differing moods and levels of quality.

In the first act, a depressed widower lawyer is sent to search a recently-deceased old woman’s house for her will (A strange premise, but it’s just a MacGuffin anyway.), given the stern warning that this assignment is his opportunity to show that his dedication to the firm (and one wonders how exactly he will show dedication or a lack thereof in searching her papers for a will) is sufficient. He arrives to a series of typical horror movie harbingers warning him to go home and leave the house alone but of course unwilling to explain their reasons for their treatment of him. While it’s certainly not groundbreaking stuff, it’s handled reasonably well, except for a rather hackneyed exchange between the lawyer and his young son that really features some spectacularly bad writing from Watkins and screenwriter Jane Goldman, building up some tension and mystery about the house and what could be so frightening the locals.

The film at this point is very dark and gloomy, filled with mists outside and a mix of dark browns and deep blues inside. Interestingly, the color is actually warmer than most Hollywood films of its type due to the much more full color range. It’s really a fairly traditional look, but it’s a traditional look that has so fallen out of favor that it’s noteworthy. Watkins and cinematographer Tim Maurice-Jones should be commended for reaching further into the past for a look that once served films well but has fallen by the wayside.

It moves along at a very slow pace save for a completely unnecessary flashback sequence, taking its time to build up tension even as it introduces its element of danger.

The only actors who get anything to do in the first act are Daniel Radcliffe, playing the lawyer, and Janet McTeer, playing the town skeptic/rich guy’s (Ciarán Hinds) deeply disturbed wife. Radcliffe is put in a difficult position, given a character who ham-fistedly opens the film with a straight razor to his throat and tears welling up but is also going about his life in a sort of slow-burn depression–a character whose inconsistency is difficult to take no matter the actor. He has difficulty with the flashier scenes but is adequate at simply looking downtrodden the rest of the time. Hinds has a fairly large role but really has nothing to do. Janet McTeer is a major problem, her ghastly bursts of psychosis being painfully over-the-top and her more “normal” moments being so flat and wooden that it was laughable.

In the second act, the lawyer finally visits the house and begins his investigations, discovering that the old woman had essentially gone insane with anger after the death of her son. He then suffers a series of bizarre ghostly encounters at the house. He comes to the conclusion that the home is haunted and has one of those annoying horror movie conversations where the rational person in the cast is told, “Oh, but you’re just a zealot who doesn’t want to admit that ghosts are real!” with Hinds. It’s again fairly by-the-numbers, and while the first few scares are based on ratcheting tension up, the film then unfortunately falls for the easy jump scares so common to horror films. Worse still, the film lacks a real sense of how to make those jumps work.

The scenes in the house are interesting visually in that the type of old-but-still-opulent look of the house in general is another example of a look that has long gone out of style. The house is filled with candles, appropriate for the time setting but also lending a warming light that most horror films would avoid in favor of a completely cold blue look. The film is also unafraid to use a fair amount of greens and reds inside the house, again showing a full range of color that’s uncommon to modern horror films. The low-key lighting remains standard-issue, darkness obscuring everything without any sign of real contrasting light. However, the long shots that had helped to build tension early in the film are replaced by quick cuts for jump scares and scenes that should be built on composition are instead built on close-ups. It’s unfortunate, because the opening really had some potential, but this act turns the film from self-consciously old-fashioned to standard fare with a hint of an old-fashioned look.

In this act, Radcliffe is almost entirely alone, and, while his depression of the early film seems to have disappeared, he handles what little his character has to do well enough. He appropriately registers feeling as he reads the old woman’s disturbing letters but doesn’t go over the top and his sense of growing fear is quite obvious.

Finally, the third act involves the lawyer finally learning the secret: the old woman’s son died in the marshes around the house and now she haunts the town, apparently at random, appearing to people and then taking control of their children in order to lead them to their deaths (and randomly screaming). He and the skeptic then set about finding the body to “reunite” the mother with her child and thus put her to rest.

Here, the film continues its devolution. The depressed, fearful lawyer has a sudden attack of bravery and tries to save a child from a burning building for no apparent reason, and the bigger emotion once again overcomes Radcliffe, as he just looks laughable throughout the sequence. The reliance on jumps continues and even the film’s previously old-fashioned look devolves into more standard-issue horror fare. A number of logical bumps also appear: The child’s body wasn’t found, so Radcliffe climbs into the marsh to look for it and finds the carriage in which the child died instantly. While the film explains away that it’s the modern technology of the car that lets him pull the carriage up and find the body in it, it’s more than a little silly to think that no one in the village could find that carriage before, and if they could get to the carriage, the body was easy to find. The children, once they die, apparently go to work helping the ghostly woman kill more children, which is just bizarre.

Radcliffe, Hinds, and McTeer continue as they were, with Hinds getting a little bit to do finally and proving capable enough while McTeer remains awful and Radcliffe remains fine at smaller emotions and troublesome at larger emotions.

The film does have a surprisingly appropriate ending instead of falling into the usual trap of implying that somehow love conquers the anger of a ghost, but that doesn’t save the film from its post-first-act flaws.

One note for all acts: Marco Beltrami’s score is all kinds of awful. It’s often distracting, it’s over the top, and it’s just a big pile of horror cliche. He has not been noteworthy in the past–really just a standard-issue Hollywood composer–but this was egregiously bad work in this film.

Overall, it’s actually not a bad film. It’s not groundbreaking in any way, but it does what it sets out to do for about 40-45 minutes before it starts to fall apart. However, it’s really just one good act followed by some standard issue horror movie fare that ruins the possibilities opened up by the first act.

Elsewhere on the internet, discussion of this film is mostly just a discussion of its lead actor, with fans insisting that he’s the next Al Pacino while detractors say he’s the next Keanu Reeves. The truth is, unsurprisingly, in between. I’m interested to see what Radcliffe does in the future, really. I found him unbearably awful in the one truly good Harry Potter film, Harry Potter and the Prisoner of Azkaban (Alfonso Cuarón, UK/USA 2004), but he was later one of the strengths of the series return to form in Harry Potter and the Deathly Hallows: Part I (David Yates, UK/USA 2010). He seems to be dedicated to the craft and he has talked often about the efforts he made to learn from some of the excellent actors (most notably Kenneth Branagh) with whom he worked in the Potter series, so I’m sort of rooting for his development as an actor. However, he most definitely has some limitations as a performer at this point, which actually jibe with the Potter differences (Azkaban expected more big emotion out of him, while Deathly Hallows expected some subtlety that’s usually more difficult to play). That makes it difficult to know whether he’s just a limited actor who’s poor at big emotions but fine otherwise or a young actor improving as he ages.