I’ve been in New South Wales for a week and a half now, talking with contacts and monitoring the “vibe” of the state election campaign. There’s no doubt about the verdict: all the signs are that there will be little movement and Labor will be comfortably returned.

But relying just on intuition is always dangerous. We love listening to rumours and gossip, but the risk is that we are really just listening to ourselves. I find it useful to get the feel of a location before trying to pick a result, but the process is inherently subjective: someone who talked to different people could end up with different answers.

So it’s worth setting out what sorts of hard evidence we have for this — or any other — election result. As I see it, we can point to five things.

First is the opinion polls. In NSW they have been consistent for the last two months, with nothing showing a swing of more than about 2% in either direction.

Second is the betting market. Labor is at unbackable odds (25-2 on at Centrebet), and has been consistently firming for several weeks. The individual seat markets have Labor as favorites to hold all of their marginal seats, with only a handful even looking close. Centrebet has the opposition poised to gain only two seats, Manly and Pittwater, both from independents.

The third thing is the behavior of politicians themselves. This needs to be interpreted with care, since there is bluff and counter-bluff. But when one side is basing its appeals on the need to avoid giving their opponents too large a majority, and the other side has stopped even pretending that they think it’s going to be close, the message is fairly unequivocal. In a similar vein is this morning’s SMH story that the Liberals have assembled a campaign team to help in Lane Cove, previously thought safe for them.

Fourth is the evidence of the federal polls. This also requires interpretation; some argue that the stellar performance of the federal opposition is bad news for incumbents in general. But the evidence is that far more people think about their vote in terms of Labor vs Coalition than as incumbent vs challenger, so other things being equal we can expect to see echoes of Labor’s strong federal performance at state level.

Finally, we have the precedents of other states. In recent years the electoral cycle has been a good guide, as state results have moved together. That’s not so much use in NSW, since it’s close to the leading edge: Queensland is the only state where a current Labor government has completed three terms. Its election last year produced only a minimal swing against the government, so for what it’s worth the precedent is in line with the rest of our evidence

Intuition and evidence could both be wrong. But don’t count on it.