04/01/2004 12:00AM

Final Four is going to the 'dogs


America loves the underdog. That's what we are told at least. Everyone seems to get excited when there's a huge upset. Maybe it's the fact that we love to see the mighty take a fall.

So it always amazes me that when it comes time to put its money where its mouth is, the general public still tends to bet on favorites. Members of the public see upsets all the time, but then when trying to handicap a game - or a horse race for that matter - they knock the underdog ("They're going to get blown out," "He's got no shot," etc.).

I consider myself a contrarian in my handicapping, and it has served me well during the NCAA tournament with a 17-10 record with my bankroll plays. I have bet on live underdogs that I felt had a shot to win outright, and when the public started backing certain dogs, I went the other way and laid short prices, even winning some games strictly because of the devalued prices (most notably in the Sweet 16 round with St. Joseph's beating Wake Forest 84-80 as a 2 1/2-point favorite and Georgia Tech beating Nevada 72-67 as a four-point choice). There's no way I would have won or even made those bets if the lines were where they should have been.

But as anyone who gambles regularly will tell you, the key to profits is often knowing when to pass. The bets that make, actually save, you the most money are the ones you don't make. That's been true as I have made selections on the rest of the tournament games and went only 14-19-1. It's very good that I didn't get overconfident and make silly plays on those games.

So, that leads up to this Saturday's Final Four and with the above in mind I wasn't sure if I would offer bankroll plays on just both national semifinals, on just one, or none at all. The lines tend to get tighter and tougher to beat in high-profile games.

As the potential matchups were being determined, I came up with my own line on the games. I felt Oklahoma St. should be favored by 1 if Georgia Tech's B.J. Elder was expected to play, and by 3 if Tech would have to go without its leading scorer. I also tabbed Duke, the only remaining No. 1 seed, as a two-point choice over Connecticut.

When the lines came out Oklahoma St. -5 and UConn -2, I immediately felt there was value on both underdogs. It's understandable why the oddsmakers put out those numbers since the public tends to bet based on the last thing they saw, and in this case that's an impressive win by OSU over St. Joe's and UConn's overall dominance and Duke's struggles. But my first thought was that the ACC is considered the best basketball conference in the land and yet we're getting points with both of its representatives.

As for the championship, due to the national semifinals being played on Saturday after press time for Monday's editions, my pick in the national title game will be available Monday morning only at www.drf.com.

6:05 p.m. Eastern at San Antonio

First, let me get a disclosure out of the way. I have mentioned several times that Oklahoma St. is my pick to win it all, and I need the Cowboys in some bracket contests. A cynic might suggest that perhaps I'm hedging with this pick. While I will admit that if you see me in a sports book on Saturday and Oklahoma St. is rolling to victory, I probably will not be too upset. I still maintain that I'm basing this selection purely on the value in relation to the point spread (again, if anyone suggests I'm cheering for a middle with OSU winning by 1-4 points, I will not deny that the situation has crossed my mind).

But here's my reasoning for the play. Both teams have outstanding defenses and points will be at a premium. For starters, that's a good reason to look at the dog as its unlikely that either team will get blown out by double digits.

While I have been cheering for Oklahoma St. from the start of the tournament and like the fact that it has any number of players who can step up and score the big basket, I have really been impressed with Georgia Tech's depth. The Yellow Jackets have overcome the injury to Elder as well as going into St. Louis with a pro-Kansas crowd and beating the Jayhawks, 79-71. They were an overlay as a five-point dog in that game.

The teams match up on the perimeter: OSU's John Lucas versus Georgia Tech's Jarrett Jack will be fun to watch, and both teams have athletic forwards. But the X-factor could be the play of Georgia Tech's 7-foot-1 Luke Schenscher. He neutralized Kansas's big men last week, and if he crashes the boards and alters some shots, that could prove to be the difference.

PLAY: Georgia Tech for 1 unit.

8:35 p.m. Eastern at San Antonio

Duke is a team that always seems to get knocked for not being perfect, but the fact is the Blue Devils do a lot of things right.

If J.J. Redick gets hot from three-point range, he can shoot out the lights. Point guard Chris Duhon, despite his rib injury, has played lock-down defense, and will draw either Ben Gordon or Taliek Brown, limiting their effectiveness. Luol Deng is a rising star, who can score inside or out. Shelden Williams might be a poor man's Emeka Okafor, the star center for UConn, but he's no slouch.

Daniel Ewing is often overlooked, but he has scored in double figures in each of the last 13 games, and Shavlik Randolph can spell Williams in trying to handle Okafor.

UConn is every bit as talented as Duke and has played great down the stretch. . . . too good, really. In the month of March, UConn played really only one tight game - the Big East title game versus Pittsburgh - and hasn't had to face its biggest weakness, free-throw shooting, in a clutch situation. UConn hit only 60.5 percent from the line during the season, and that is very likely to come back to haunt them here.

I feel strange penalizing the Huskies for being so dominant, but the point remains that Duke has the experience that could prove the difference in a tight game.

PLAY: Duke for 1 unit.

Bankroll record: 17-10 for a net profit of 6 units (based on laying 1.1 units to win 1).
Record on non-bankroll plays: 14-19-1. Record on all tournament games: 31-29-1.

::  (Pdf)

::  (Pdf)

:: Go to , and much more.