Nobody ever said the United States election system was simple. And most media coverage doesn’t make it any easier for the viewer — sometimes due to ignorance or political bias, but more often due to the media’s own twisted lens, which prefers show to substance and keeps trying to drum up excitement by insisting that things are closer and more unpredictable than they really are.

So this is an attempt to provide the Australian viewer with what they need to know to understand the results as they come in tomorrow.

Three general points to start with. First, although Americans are electing a president, the mechanics of doing so are more like a parliamentary system. Instead of winning a majority of the overall vote, the goal is to win a majority in a large representative body: the electoral college, with its 538 seats. (Hence the name of Nate Silver’s wonderful blog, FiveThirtyEight.)

Since the electoral college is state-based, and states (with two minor exceptions, Maine and Nebraska) allocate their seats on a winner-takes-all basis, about four-fifths of the states hold no interest for the result, being safe for one party or the other. Just as an Australian election focuses on marginal seats, the attention tomorrow is all on the marginal states.

Second, elections are controlled locally: all the administrative arrangements vary from state to state and often from county to county.

That particularly affects poll closing times — which would be difficult enough anyway, since the marginal states stretch across four time zones and some of them straddle time zone boundaries. You can’t just look at a map, subtract nine or ten hours, and know when to expect results to come in; you need a list.

Third, the American networks don’t like waiting around for votes to actually be counted. They project results on the basis of exit polls, once the polls (or most of them) in a particular state have closed, and there’s a lot of pressure to be first with a projection. Since this practice got them into trouble with Florida in 2000 they’ve become a bit more restrained, but all except the very closest races will still be called with only a small number of votes counted.

With all that in mind, here, chronologically, is what to look for tomorrow. All times are Australian eastern summer time (GMT+11).

10am: Polls close in (most of) Indiana and Kentucky. Neither is a key state; Kentucky is safe Republican, and although Obama won Indiana last time, it’s the one state his campaign has written off this time. But if the networks don’t call Indiana for Romney immediately that could be a sign that the Democrat vote is holding up better than expected. (Or it could just mean they’re being virtuous and waiting for the portion of the state that’s on central time to finish voting.)

11am: Polls close in Virginia and the greater part of Florida (and three others). Florida is going to be tight, so don’t expect anything from there until 12pm, when its western end closes. But Virginia is a key state: Obama doesn’t need it (although he’s ahead there in recent polls), but Romney will find it very hard to put together a majority without it. If Obama holds Virginia, and particularly if it looks comfortable (that is, if the networks call it fairly quickly), it’s probably going to be a bad day for the Republicans.

11.30am: Polls close in North Carolina and Ohio (and West Virginia). North Carolina is the one marginal state where Romney has a clear lead; it’ll be no surprise if he wins there, whereas if he loses it then it’s going to be a very decisive Democrat victory. Ohio is the single most critical state, so it won’t be called quickly, but when it is it could well decide the election. Look at it in conjunction with Virginia: if Obama wins both, he’s home and hosed. If Romney wins both then he’s got an edge, but not necessarily a decisive one.

12pm: Polls close in the rest of Florida and in New Hampshire and Pennsylvania (and 16 others). If Obama wins Florida then it’s all over, even if he’s lost Virginia and Ohio. New Hampshire is a small state, so it only becomes important if Romney is doing better than expected — if he wins Virginia, Ohio and Florida, then Obama would have to hold everything else that’s in contention. Pennsylvania is on the list only in a speculative way; it should be safe for Obama, but Romney has campaigned there recently to try to increase his options. If it’s looking close, that’s good news for the GOP.

1pm: Polls close in Colorado and Wisconsin (and nine others). If Romney loses Ohio, he could conceivably salvage things by picking up Wisconsin, but the polls are very discouraging. Colorado, on the other hand, is another one that Obama doesn’t really need unless he’s doing surprisingly badly elsewhere, in which case it could be a critical backstop.

2pm: Polls close in Iowa and Nevada (and four others). If the polls are right, there’ll be no need to be still watching at this point, since Obama will already have locked in victory by winning either Ohio, or Virginia plus Colorado. But if Romney has Florida, Ohio, and either Virginia or Colorado under his belt, then Obama will need Iowa or Nevada (or both) to eke out a narrow victory. (Fortunately for him, the polls show him well clear in both.)

3pm: Five states, including the largest, California, don’t close until now or later, but none of them are marginal. Counting, however, will be still going on across the country, and if one or more of those large key states is very close, it could be four or even five o’clock before we know the winner.

You can see the whole thing in graphical form at The New York Times or a stripped-down version at Buzzfeed. The gist? Obama has a lot more paths to victory than Romney does; for Romney to pull off a win, just about everything has to go right for him, or the polls have to be dramatically and systematically wrong.

Best case for Obama is that he holds onto his current leads and also manages to reel in Florida, which would give him 332 electoral college seats to Romney’s 206. Best case for Romney is that he pulls off an upset in Ohio and also manages to mop up, say, Virginia, Colorado, New Hampshire and Iowa, which would give him victory by 285 to 253.

The first of those looks a lot more likely to me, but we’ll know soon enough.