Kos asks for predicitons on Iowa numbers, but he doesn't say what numbers these are. This WaPo article gives a nice summary of how the caucus works. It mentions there will be reports of two sets of numbers. One gives the initial preferences of the caucus attendees. [UPDATE: I should mention this an unofficial count done by the press, not the state party] The article likens this number to a primary result. At the end of the night, though, the state will give the "state delegate equivalent", which is obtained by a computer program calculating the number of delegates a candidate will end up recieving based on phoned in results by the precint leaders. This is more the "official" result. If we are only interested in the number of delegates towards winning the nomination, only the latter number is of interest. On the other hand, there are only 45 pledged delegates up for grabs in Iowa. This is barely over 1% of the total delegates available. So the real significance of Iowa is how candidates use it to raise money and gain support in other states. For this, a candidate might use whatever numbers best suits him. In any case, just for the fun of it, here are my predictions about pledged delegates won in Iowa:
The bigger questions, though, will be how whatever numbers get spinned? Will anyone drop out of the race based on, Iowa? (No) How will they affect New Hampshire? (Not much) I think if four candidates do win delegates in Iowa, that in itself is a bigger story than how many each wins.
UPDATE: Micky Kaus goes into detail on four possible counts. The two I have mentioned would correspond to his counts #2 and #4. Kaus indicates that the most reported result will be an entrance poll (count #1). The WaPo article does refer to this and says it is done at 50 of the 1,993 precints. I'm not so concerned with what numbers are reported though the night, though, as what numbers are played out the following week until NH and how those numbers are played.
UPDATE 2: According to the NY Times the state delegate count will be available by 10pm CST. and thus should be reported on the late local news and the many morning papers. I'm still confused, though, on what this count is. The Times says it is the percentage of state delegates. The WaPo gives this nice primer of how Iowa selects their delegates. Is the computer program giving a prediciton of what percentage of the 3000 state delegates, that select 16 pledge candidates, each candidate will win? It seems most likely the results will be giving the percentage of county delegates each candidate won across the state. These county delegates will then in March select 3000 delegates to the state convention in June and 3000 to the district convention in April. This is done by a process like that of the precinct level where nonviable cacuses can realign. A delegate (at least at the county level) is also simply free to change their mind. This could certainly happen if a candidate has dropped out by the time of the appropriate convention. So when it comes down to it, it seems that the state numbers will be bandied about, but it's difficult to figure out yet how those number translate into delegates won towards the nomination.