Since this column runs on Monday, I may be the last pundit to weigh in on the lessons of last week's Democratic primary for governor in Virginia. The danger in all this punditry is to try and draw lessons about the national political environment; the real lessons are about the unique dynamics of low-turnout primaries.

Read more from Mark Blumenthal on how Creigh Deeds won the Virginia Democratic primary at Pollster.com.

Start with turnout. While last week's total vote of 319,177 was on the high side of most expectations, it still amounted to just 6.3 percent of registered voters. While higher than in previous Democratic primaries, that number falls far short of the 45 percent of registered voters who turned out for the state's 2005 gubernatorial general election, the 53 percent who voted in the 2006 Senate election or the 76 percent who voted for president last year.

If the "horse race" poll results we looked at in Virginia miss "essential context," the most important omission involves our understanding of how low-turnout primary voters differ from those who vote only in general elections. Sadly, I can point to no research from Virginia that contrasts the attitudes and media habits of validated, off-year primary voters to all others. (If we did, we would certainly find far more partisanship and -- I would wager -- more exposure to all forms of media.)

Even so, there are some lessons from last week's election that tell us something about low-turnout primaries:

Primaries Are More Volatile. In general elections, partisan affiliation binds a significant segment of the electorate to their party's nominee, so campaigns fight over the relatively few voters who are uncertain about their choice. In primary elections, voters usually lack a comparable anchor, so the potential for big shifts in candidate preference is greater.

Negative Messages Are Dangerous. In a general election or a two-candidate primary, one candidate can bash the other with the expectation of ultimately picking up any support the opponent loses. Do that in a primary with three or more viable candidates and, more often than not, the negative attack hurts the recipient of the attack but the shift in support goes to a third candidate not involved in the skirmish.

That classic scenario played out once again in Virginia. After Terry McAuliffe had blanketed much of the state with television advertising (outside the Washington market), Brian Moran chose to devote a significant share of a smaller advertising budget to attacking McAuliffe. McAuliffe lost support, but Creigh Deeds was the ultimate beneficiary.

Newspaper Endorsements Matter More. Without the crutch of partisanship or clearly defined issue differences, voters have a harder time deciding, especially in an "open" race where they do not know any of the candidates well, as was the case in last week's primary. So a high-profile editorial endorsement by a universally respected newspaper can have a profound impact.

And, yes, for Democratic primary voters in Northern Virginia, the Washington Post is universally respected. Just ask observers across the Potomac River in Washington, where the Post's endorsement in local primary elections has been rightly accorded "kingmaker" status.

The surprise nature of the Post's endorsement of Deeds also gave it greater impact. The lesson is that "surprise endorsements matter," says Republican consultant Patrick Ruffini, suggesting parallels like the Des Moines Register's endorsement of John Edwards in 2004 or the Manchester Union Leader's endorsement of John McCain in 2008.

Internet Advertising May Persuade. At 3 p.m. on the Monday before the election, the Deeds campaign started running a "Google Blast" of display advertising touting the Post's endorsement. The ads appeared on most of the sites in the Google Ad network and geo-targeted Northern Virginia, reaching what some estimate as more 80 percent of Internet users in the region.

Of course we cannot credit any one form of media for Deeds beating McAuliffe and Moran on their own Northern Virginia turf. The Deeds campaign also featured the endorsement in a television ad that ran in the region, mailed copies of the endorsement to past primary voters and blanketed median strips with "Post endorses Deeds" signs.

The endorsement gave the Deeds campaign a simple, easily communicated message that had special power with a primary electorate. So while both yard signs and the Internet ads may have been effective, their success may have been unique to the special circumstances of this race. Still, in this race and in the special congressional election earlier this year in New York's 20th District, campaigns have used Internet advertising to try persuading uncertain voters.

Message Matters Most. Browse through the television ads still posted at the McAuliffe Web site and you get a sense of why his huge fundraising advantage failed to translate into victory. The ads flail from one vague theme to another -- "new ideas in Richmond," "think big," "my job is to protect your job" -- emphasizing an eclectic mix of issues not usually associated with state government, including renewable energy from offshore windmills and chicken manure.

Watch the Deeds ads and you see three consistent elements: a clear theme ("the Democrat most qualified to be our next governor"), an embrace of continuity with two popular Democratic governors (Mark Warner and Tim Kaine) and an emphasis on issues more commonly associated with Virginia state government (education and transportation).

Even in a low-turnout primary, where conventional wisdom emphasizes field organizing, coherent messaging is most important. Of all of Virginia's lessons, this one is most universal.