How do you arrive at an 8.4 percent drop? And isn't a 4.2 percent chance at success dangerously close to the margin of error for "no chance whatsoever"?

The number has been too high for too long. For anyone willing to parse the numbers, she's been sunk since even before PA, which made the mid-teens Deathwatch inexplicable. Now that it is so obvious that even the horse-race-loving media can't deny it, the 1-in-7 chance fantasy of the Deathwatch is no longer sustainable. Deathwatch has rejoined the reality-based community.

4.2% is probably too high, but it is at least in the ballpark of the chances that Obama somehow self-destructs before the convention.

LMoE

4.2?!?! Whoa, look at that baby go!

How does margin of error even apply? This isn't a poll. The odds are what they are, or at least are claimed to be what they are claimed to be.

FWIW the Iowa electronic markets have had Hillary at closer to .21, or 21% chance to win, lately. But her odds fell off the cliff last night, and are now at about 10%.

I expect it to go even lower as the trickle of superdelegates toward Obama become a steady stream. But as long as she's still officially in the race, I don't expect it to go much below 5%. There's always someone out there willing to take a 20:1 flier on Obama being caught in bed with the proverbial dead girl or live boy in the next few weeks. Hell, "none of the above" is still at 3.7%.