![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
jrtomI've been really annoyed recently by the abuse of the phrase "mathematically impossible" in the popular press to describe the likelihood of either of the following:
(a) either candidate gaining the support of at least 2025 delegates (pledged or super)...
(b) Senator Clinton gaining the support of more pledged delegates than Senator Obama...
...by the time the primaries are over (but before the convention).
In short: neither of these are "impossible". Highly unlikely, maybe. But calling it "mathematically impossible" is simply ridiculous.
(a) would be "mathematically impossible" if the difference between either candidate's current delegate count and 2025 were greater than the number of remaining unpledged delegates.
Let's look at each candidate's current stats
(courtesy of http://demconwatch.blogspot.com/2008/02/ultimate-delegate-tracker.html as of 21 April 13:22 PDT):
(delegates, superdelegates, total)
Obama
(1416, 231, 1647)
Clinton
(1253, 255, 1508)
Edwards
(18, 0, 18)
Remaining
(566, 308, 874)
Note that there are 566 remaining unpledged delegates (not counting superdelegates). So clearly if either Obama or Clinton got _all_ of them, they'd have > 2025. (Obama would need 66.6% of these to reach 2025, Clinton would need 91.3%.)
In fact, Obama could come pretty close to 2025 if he lost every single pledged superdelegate that he currently has...if he also won all the remaining unpledged non-superdelegates. (He'd end up with 1982, only 43 short.)
Of course, none of these are likely scenarios. But they're all mathematically possible, dammit.
Regarding (b): this article:
http://www.huffingtonpost.com/m.s.-bellows/no-more-spin-clinton-penn_b_97701.html
does a pretty good job of addressing the numbers and their implications. (One factoid of interest: Clinton needs to win 64% of the remaining elected delegates to catch up with Obama...assuming no movement in the relative superdelegate tallies, and that Michigan and Florida aren't counted as-is (if they are, she has to win 56%). Or, to put it another way, if she doesn't win Pennsylvania by at least 28 points, she has to win by even larger margins in future primaries.)
Again, a streak of 28-point blowouts is unlikely for either candidate. (But not impossible.)
As a side note, in possession of the right polling information, including a probability distribution for each remaining state over the set of possible delegate assignments for each candidate, it would be possible to assign probabilities to the various events of interest (candidate X has >= N delegates, difference of delegate counts for candidates X and Y is <= N delegates, etc.). Sadly, no one seems likely to provide this sort of data (and I must admit that I've got enough to do right now that I wouldn't do the required analysis, possibly even if it were handed to me on a silver platter), but I'd sure like to see _someone_ publish the results of this analysis.
(a) either candidate gaining the support of at least 2025 delegates (pledged or super)...
(b) Senator Clinton gaining the support of more pledged delegates than Senator Obama...
...by the time the primaries are over (but before the convention).
In short: neither of these are "impossible". Highly unlikely, maybe. But calling it "mathematically impossible" is simply ridiculous.
(a) would be "mathematically impossible" if the difference between either candidate's current delegate count and 2025 were greater than the number of remaining unpledged delegates.
Let's look at each candidate's current stats
(courtesy of http://demconwatch.blogspot.com/2008/02/ultimate-delegate-tracker.html as of 21 April 13:22 PDT):
(delegates, superdelegates, total)
Obama
(1416, 231, 1647)
Clinton
(1253, 255, 1508)
Edwards
(18, 0, 18)
Remaining
(566, 308, 874)
Note that there are 566 remaining unpledged delegates (not counting superdelegates). So clearly if either Obama or Clinton got _all_ of them, they'd have > 2025. (Obama would need 66.6% of these to reach 2025, Clinton would need 91.3%.)
In fact, Obama could come pretty close to 2025 if he lost every single pledged superdelegate that he currently has...if he also won all the remaining unpledged non-superdelegates. (He'd end up with 1982, only 43 short.)
Of course, none of these are likely scenarios. But they're all mathematically possible, dammit.
Regarding (b): this article:
http://www.huffingtonpost.com/m.s.-bellows/no-more-spin-clinton-penn_b_97701.html
does a pretty good job of addressing the numbers and their implications. (One factoid of interest: Clinton needs to win 64% of the remaining elected delegates to catch up with Obama...assuming no movement in the relative superdelegate tallies, and that Michigan and Florida aren't counted as-is (if they are, she has to win 56%). Or, to put it another way, if she doesn't win Pennsylvania by at least 28 points, she has to win by even larger margins in future primaries.)
Again, a streak of 28-point blowouts is unlikely for either candidate. (But not impossible.)
As a side note, in possession of the right polling information, including a probability distribution for each remaining state over the set of possible delegate assignments for each candidate, it would be possible to assign probabilities to the various events of interest (candidate X has >= N delegates, difference of delegate counts for candidates X and Y is <= N delegates, etc.). Sadly, no one seems likely to provide this sort of data (and I must admit that I've got enough to do right now that I wouldn't do the required analysis, possibly even if it were handed to me on a silver platter), but I'd sure like to see _someone_ publish the results of this analysis.