The Friday Night Effect follows three young-ish women – flatmates living somewhere in Ireland – as they prepare for and then head out to a night on the town. By the end of the play, we're told, one of them might die. It's only a "might", because the plot isn't fixed yet; at crucial points the action stops, and we the audience are invited to vote on what a particular character should do.

It's appropriate that we exercise this degree of control, for the question of how much you can guide someone else's life is very much at the heart of the play. The key character is Collette, played with feisty unpredictability by the ever-impressive Eva O'Connor; Collette has bipolar disorder, and today seems to be on the threshold of a manic phase. She's also in an evidently destructive relationship with the domineering, dislikeable Brian, leaving her two flatmates – Sive and Jamie – with the unenviable task of trying to give advice at a time she least wants to be advised.

All three women are convincingly portrayed, offering a real sense that you're peeking in on a relationship that's been ebbing and flowing for years. The script finds opportunities to explore Collette's condition through natural, unforced dialogue, at times managing the difficult trick of being both funny and respectful. The episodic structure successfully keeps the plot cracking along, and although it deals with some weighty issues, it always stays entertaining.

The male parts, in contrast, are performed by just one actor – and largely played as grotesques. I'm not sure that decision was the right one; for Collette's story to make any sense, we need to understand what she sees in Brian, and the influence he has over her would be more disturbing if his character seemed more real. A side plot involving Sive's work as an escort is interesting, but it's really the makings of a whole separate play; I could obviously be wrong about this, but it felt a little like it was shoehorned-in as a favourite cause of one of the creative team.

But of course, it's the audience voting – the opportunity to influence the course of events – which most defines The Friday Night Effect. These techniques are still relatively novel in theatre, but they're much better-explored in written fiction and computer games, and they often rely on the illusion of agency; the audience controls some fine details, but the plot-line always converges back onto the story the playwright wants to tell. Done subtly, there's nothing wrong with this, but the key is to make it seem like the decisions being made affect the course of characters' lives – even if, in reality, they don't.

In this case, I feel the workings are a little too clearly on display. At one point, we made a decision to ignore a lecherous bouncer and avoid getting thrown out of a nightclub; a few lines later, the women had an argument and decided to leave anyway. A couple of major choices, including one involving cocaine, are made without asking us – which, given the expectations that had been set, felt a little like cheating. And the big final dilemma comes out of nowhere, when ideally it would have seemed like the inevitable conclusion of the decisions we'd collectively made.

All that said, the audience around me loved what – for many – will be their first taste of this style of theatre. And it was interesting, too, to discover just how unpredictable the group's reactions were: time and again, I was surprised that the vote didn't split along obvious gender lines. So as a social experiment, The Friday Night Effect works, and as a story it's well worth hearing. There's still a lot of room left for the concept to grow, and I'll look forward to seeing where it goes next.