Sunday, October 01, 2006

Overestimating the little guys (and girl)

Most bloggers who made predictions about the number of delegates each candidate would receive got it, if the numbers continue in their present course, pretty much right. Kennedy was a little higher, Rae a little lower, Brison a little higher, but the overall pattern was correct.

There is, however, one major difference between the actual results and the forecast results. The divide between the top 4 and the bottom four is much larger than expected. Dryden was expected to be around 7-8% instead of 5.2%, Volpe was expected to hit 6% instead of 4.4%, Brison had high hopes of 8-9% instead of 3.8%, and even Martha Hall Findlay could realistically hope for 2-3% instead of 1%.

Meanwhile, Ignatieff was expected to be around 27% instead of 230.6%, while Rae, Dion, and Kennedy were expected to occupy the 12-16% range, not the 16.5-19% range.

What this means is that the smaller players are all but irrelevant. Sure, there are situations where the combined power of Dryden, Volpe, Brison, and MHF could exert some change, and if one even Rae, Dion, or Kennedy turn to Ignatieff, one of Dryden, Volpe, or Brison will be required to be a final kingmaker (assuming, of course, that delegates follow their leaders).

Because of these results, I wouldn't be surprised to see a few dropouts soon. Dryden and Brison, for example, might be embarassed by their poor showings and hit the road. Volpe will probably be stubborn as ever. And MHF has nothing to lose by sticking it out, so I imagine she'll hang in there. But in any event, I figure we'll lose at least one of the four over the next few weeks.

So why did the lower four get fewer delegates than expected? I don't think it's because they had less support than we expected. I'm pretty sure those blogger predictions were correct, not in terms of the number of delegates each candidate would receive, but in the number of votes. The problem is that, because of the fact that numbers of votes need to bedivided by 14 and rounded to integers, candidates will only start making their mark once they pass a certain threshold. In my riding, there were a number of candidates who received a significant number of votes that did not get a delegate spot. It's not even that uncommon for one candidate to get one fewer votes than another candidate, but the latter to get a delegate spot and the former not.

In any event, this system of rounding cuts out the marginal support that candidates have within ridings, thus shifting the totals by a few points.

In the end, it doesn't make too much diffference, but it does mean that smaller candidates' hopes of playing kingmaker may come to nothing.

3 Comments:

Blogger Mark said...

Fair point - but here's a counter argument. If Rae/Dion/Kennedy finish in a dead heat on the first ballot, those "little" people will br critical in determining which of them actually moves on.

12:43 PM  
Blogger Penny said...

Any idea how many mail-in votes there might be, when they would be counted and how they could affect things, if at all?

1:58 PM  
Blogger Anonymous Liberal said...

My understanding is that mail-in votes are most freqently found in Quebec, and there may be a dozen or so ridings there that have mail-in votes. I also know of one in BC. I have no real answer though.

If it is the case that they're largely centered on Quebec, it'll mean that Ignatieff, Dion, and Rae gain relative to the other candidates, with Kennedy taking a particular beating relative to the rest of his results.

As of 1:12 AM ET on Monday, there are 20 DSMs in Ontario that have yet to report, 15 in Quebec, 16 in BC, 7 from Manitoba, and a few more from elsewhere.

My guess is that those numbers will work more to Dion's favour than to Kennedy's, and that the final count will have Dion just marginally ahead (by maybe 10-15 delegates) of Kennedy.

Other than that, shouldn't change much.

10:21 PM  

Post a Comment

<< Home