Wanna bet Ethan wins it this year?

You don't have to be a gambler to fancy the Oscar odds. This is one area where the punters know more than the bookies, says Neil Young

Even after all these years, I still find Misery very hard to watch. Nothing to do with James Caan getting his ankles smashed, mind – what makes me flinch is the damage Kathy Bates inflicted on my bank account when she picked up her Oscar after I'd staked serious money on one of her rivals – Mr & Mrs Bridge's Joanne Woodward. It was my painful introduction to the alluring, dangerous world of betting on the Academy Awards – a minority pursuit back in 1991.

It all started when I read a frighteningly persuasive article in the Racing Post, which revealed that the veteran Woodward was even money at Caesar's Palace in Las Vegas, but an astronomical 8-1 with the only British firm daring enough to quote odds, William Hill. They made Bates an odds-on favourite, but, as the Post pointed out, "It's hard to imagine voters who in the past have plumped for heavyweight actresses and equally heavyweight films like Gandhi suddenly latching on to anything involving Stephen King." I was down at William Hill before the doors were unlocked.

The Woodward debacle warned me off Oscar betting for a while, despite the fact that all the other high-street bookies gradually joined Hill in tempting punters by pricing up the "money" categories. It wasn't until 1997 that I finally spotted what looked like a serious "rick" – gambling-speak for a major odds error. By then an avid reader of Hollywood trade magazines, I'd picked up a groundswell behind Frances McDormand as Best Actress for Fargo, even though most critics' awards had gone to Brenda Blethyn for Secrets & Lies. Patriotic UK firms made Blethyn a short-priced favourite, with McDormand 6-1 – odds that looked absurdly generous when Nicolas Cage called her up to the podium to receive her prize.

The next year was trickier, with four English Best Actress candidates: Helena Bonham Carter, Julie Christie, Judi Dench and Kate Winslet. But I'd done enough research to know that, when the shortlist features a majority of non-American names, the Yanks are the ones to bet on. The bookies hadn't twigged, and pushed As Good As It Gets' Helen Hunt right out to 8-1 – a costly error for them, a lucrative one for me. A scrap between Gwyneth Paltrow and Cate Blanchett in 1999 was more straightforward. I guessed that Paltrow had the edge and took the best price available – a skimpy 7-4. But this was my third consecutive Best Actress win, something not even Meryl Streep could boast. It was just as well, as I'd put together a trifecta of disasters in the equivalent male races: Tom Cruise, Peter Fonda and Nick Nolte each somehow contriving to lose Best Actor to freakish opposition. By 2000 my Best Actress record led me to pile money on Annette Bening, dismissing Boys Don't Cry's Hilary Swank as an unknown in an obscure film. The Academy did not concur.

Last year, with Julia Roberts unbackable and unopposable as Best Actress, I looked elsewhere for value, and saw the strong likelihood of vote-splitting among the supporting actresses. This eliminated the two Brits (Dench and Julie Walters) and the two Almost Famous co-stars (McDormand and Kate Hudson) and left Pollock's unconsidered Marcia Gay Harden. One foolhardy bookie was quite happy to lay 14-1 – and I was back on track.

Oscar betting has mushroomed over the past decade – the big-name firms issue odds on Picture, Director and the four acting categories, and they've been joined by more enterprising internet-based rivals like Blue Square and Paddy Power (if you think you've got a sure thing for Best Animated Short, Paddy will accommodate you). While the oddsmakers are inevitably more Oscar-savvy than in years gone by, this remains one of the very few areas where they have no real advantage over clued-up punters who do their homework.

This year's Best Supporting Actor is a prime example, with the bookies concentrating on heavy-hitting Brits: Jim Broadbent (6-4 for Iris), Ian McKellen (7-4 for The Lord of the Rings) and Ben Kingsley (5-2 for Sexy Beast). But the history of Oscar vote-splitting suggests we should favour the "makeweight" Americans, Jon Voight (Ali) and Ethan Hawke (Training Day), especially since they're available at a mouth-watering 10-1 and 12-1 respectively.

There's also the "sop factor" to consider: many Hollywood observers rationalised Gay Harden's "shock" success by suggesting her prize was compensation for her co-star Ed Harris not winning Best Actor. Perhaps a similar mechanism lay behind similar "surprise" wins in supporting categories by Cuba Gooding Jr (sop to Jerry Maguire's Tom Cruise) and James Coburn (sop to Affliction's Nick Nolte).

I've had only a smidgen on Voight, on the basis that (a) he acts in a mask (b) he isn't in the movie very much, and (c) Ali's Will Smith isn't as dangerous a Best Actor contender as Training Day's Denzel Washington. Despite his Bafta antics, Russell Crowe remains a sufficiently steady favourite to suggest Washington will again miss out. Meaning, in turn, significant numbers of guilty compensation votes flocking to his co-star Hawke.

The 31-year-old Texan does finally seem to have ditched his goatee-stroking, pseudo-intellectual novelist image – he em-erged from Michael Almereyda's Hamlet update with credit, and his regular work with Richard Linklater (continuing this year with Tape and Waking Life) has done him no harm. He is, strictly speaking, the lead in Training Day, the movie that's provided him with the strongest reviews of a career that stretches back to 1989 Best Picture nominee Dead Poets Society.

Those 12-1 odds (paddypower.com) are massively generous – I'd make Hawke a narrow second favourite behind McKellen, given the growing Hollywood feeling that Rings may well sweep the board. But remember 1978, when Alec Guinness was nominated as Supporting Actor for the "Gandalf" role in Star Wars, alongside a fellow Brit, a Russian, an Austrian, and the lone American Jason Robards? Robards's win proved patriotism was even mightier than the Force – we're about to find out if the wizardry of Middle Earth can put up more potent opposition.