The men's and women's college basketball tournaments are down to just four teams each, but here in my NASCAR Bracket of Massive Significance, I'm not ready for that yet. Our quarterfinals have left us with some very intriguing matchups:

Greg Biffle versus Martin Truex Jr.: Biffle's numbers at Martinsville are less than stellar. He's never had a top-5 finish there. Truex just has one top-5 but was eighth there last fall, so he's my pick.

Dale Earnhardt Jr. versus Matt Kenseth: These two have a history dating back to the Nationwide Series before it was the Nationwide Series. Junior has been excellent at Martinsville lately; I like him to reach the semis.

Kevin Harvick versus Tony Stewart: Fun matchup between last year's Martinsville winners. Stewart won his last start, but his three before that were all 24th or worse. Harvick has three straight top-5s at the Paper Clip, so he's my man.

Clint Bowyer versus Paul Menard: The only non-1-versus-2 matchup in the bracket. Bowyer's Martinsville numbers aren't great, but Menard has never had a top-10 there.

Three-man show?

In the previous 10 races at Martinsville, three drivers have separated themselves from the pack: Jeff Gordon, Denny Hamlin and Jimmie Johnson.

Those three all have an average finish of 5.1 or better in that time. No other driver has a mark better than 10th.

So, looking inside the numbers, who's the man to beat? Or who's the guy to pick up if you can have only one on your fantasy team? That depends on how you look at it.

In terms of how many times these drivers have been passed the previous 10 races, Johnson has been passed the fewest at 292. And he does the best of staying up front with an average position of 5.4. The other two each have been passed more than 300 times in those races, with average positions between sixth and seventh.

But if you look at pure speed, Gordon has been the fastest driver on the track most often. He's run the fastest lap on 559 circuits over the 10 races. Johnson's at 491, and Hamlin 420.

At the end of the race, a tight finish also favors Gordon. Over the final 10 percent in each of those 10 races, Gordon has a plus-2 pass differential, while Johnson and Hamlin have put up negative numbers.

However, not much separates these drivers, and all three will be heavy threats up front all day Sunday.

Looking for trouble

Every week, our stats and analysis team compiles a breakdown of the wrecks at the next track on the schedule. Here's this week's takeaway:

For everybody who missed the slam-banging action at Bristol two weeks ago, you might be in luck.

Since 2004, when Martinsville was repaved, no track has featured more accidents than Martinsville with 154. The next highest is Charlotte with 108, then Bristol at 103.

And when deciding what end of the track to focus on, pick Turns 3 and 4. Since 1990, there've been 130 accidents in Turns 3 and 4, opposed to 86 in Turns 1 and 2.

The Eliminator: Martinsville

For those of you new to my little blog, every week I use a device called The Eliminator to make my race pick. It's pretty simple: Instead of telling you why one guy will win, I'll point out why everybody else has to lose. The driver remaining, by process of elimination, is the race winner.

1. Eleven of the past 12 Martinsville winners finished in the top 20 in the previous week's race ( 26 eliminated, 19 remaining).

2. The past seven spring Martinsville winners finished ninth or better in the previous Martinsville race (12 eliminated, seven remaining).

3. The past nine Martinsville winners had a top-10 in the most recent Richmond race (three eliminated, four remaining).

4. The past seven Martinsville winners had a top-5 finish in the previous New Hampshire race (three eliminated, one remaining).

Your winner: Tony Stewart