The de la Hoya Candidate: Remember the first time Oscar de la Hoya lost? It was to Felix Trinidad in 1999, when both men were undefeated; it was boxing's last "Battle of the Century" of the 20th Century. Most observers, including myself, felt that the Golden Boy was an easy winner, and by almost any objective criteria, such as punches landed, he should have won. But the ringside judges didn't agree, and gave a split decision to Trinidad, who was the aggressor from start to finish. It was a disputed outcome, but not, truth be told, a surprising one: judges have always preferred the boxer who carries the fight to his opponent over the one who dances and piles up the points. And it certainly didn't hurt that Trinidad was managed by Don King.

I thought of that today in the context of the Wednesday post-mortems of the Ohio and Texas primaries yesterday. The Golden Boy of the Democratic Party has built up a comfortable lead through a series of decisive victories last month, but now discovers that he still has a few rounds to fight. His campaigning in the week before was sluggish, and he got clocked by his opponent on some nondescript punches. So his first instinct is to rest on his laurels, dance out the final rounds of the fight, and win on points.

Bad idea. For one thing, like de la Hoya after the ninth round of the Trinidad fight, he's not up by as much as he thinks he is. Thanks to the ridiculous distribution of delegates in states like Texas, the meme that the nominating process is not particularly democratic is starting to catch on. Just being ahead slightly in what have come to be known as "elected delegates" isn't going to win the nomination, particularly if Obama closes on a losing streak that includes Pennsylvania, Indiana, North Carolina (and, Kobe willing, any revotes in Michigan and Florida).

SuperDelegates are not an amorphous mass that can be bullied or cajoled into supporting a nominee in August just because he did well in February. The entire reason they exist is that the rank-and-file Democratic primary and caucus participant does not have an astute track record in choosing nominees. They will look to what their constituents want, as reflected in the actual voter preference of their states and districts, and that math is not unfavorable to Ms. Clinton.

Obama needs to close strong. A win in Pennsylvania would help, but bagging North Carolina and either Michigan or Florida would be just as good. There is no reason to believe that party insiders and pols overwhelmingly favor Clinton; in fact, the opposite is probably true.

But most of all, he has to start acting like he has a pair. Only a string of defeats to close out the primaries will convince the Supers that he isn't a viable nominee, so he should stop coasting, and finish off his opponent, the way de la Hoya didn't do against Trinidad.