With the limited number of games and varying levels of opposition, football stats often don't mean much. Too many aberrations.

But midway through the Hawaii season we can analyze some crazy numbers. Just as some of the stats in 2006 and 2007 were just too eye-popping to ignore, they are this year—but now they're car-wreck terrible instead of obscenely gorgeous.

Yes, the turnovers are overwhelming: 24 lost to 14 gained. In Friday's defeat at Boise, the Warriors were back to decisively losing what every coach in the world agrees is the most important stat in the game.

But here's something equally bothersome. UH has been outscored 109-55 after halftime of its seven games, and 57-16 against four WAC foes.

In the last five games, it's been 81-16 after the marching band yields. The Warriors failed to score AT ALL in the second half of three games, including 17-0 on Friday.

First half numbers? UH has outscored its WAC opponents 64-33 in the first two quarters (overall it's a 99-78 deficit).

I'm on record about it being smart to close up the playbook some and run the clock when you've got a lead, especially against a team that will only get back in the game with your help. That was the case against LaTech.

But when a football team consistently performs significantly better early than toward the end—and has a losing record—two harsh but fair conclusions can be made. One: Opposing coaches make better adjustments. Two: Opposing players react better in crucial situations.

With that being said, I don't believe this team has given up on itself. It would've happened already. No one's come up with a stat yet to quantify heart, and we won't really know how much this group has until it's all said and done.

I'm reminded in some ways of the 2004 Warriors. They were the opposite in that the offense was forced to carry the defense, the same in that it often didn't work out. When they returned from Fresno State 4-5 after a 70-14 whipping, some guys pointed fingers at each other and June Jones' detractors smelled blood.

They rallied to win their last four, beating two Big Ten teams and UAB in the Hawaii Bowl.

Can this edition do something similar? Maybe.

But not without a quarterback change.

I'm not talking about lifting the starter—until halftime. Inoke Funaki seems OK in the early going. But one mistake often leads to another for him, with a compounding effect we prefer in retirement plans, not interception rates. Some of it has to do with a leaky offensive line, and all the hits he takes running the ball.

Start Funaki, and then start Tyler Graunke—who led the comeback win over Weber State—at the beginning of the second half. If Graunke's still hurt or in the doghouse, next man up, Greg Alexander or Brent Rausch.

In this case, it's like pitchers: It's better to get them out of there one too soon than one too late.