In my first article, I am going to perform the worthwhile service of proposing a way to save something that probably can't be saved, and that many think doesn't need to be saved in the first place.

That's right, we're talking about the all-star game.

As I see it, the all-star game has two obvious problems right now: 

1. The number of players reminds you, more than anything, of a spring training game.

2. The best players aren't always selected.

As it stands, we have a world where Michael Young can be the Most Valuable Player of the all-star game one year and then spend the entire game on the bench in the next.

Or how about Dmitri Young, J.J. Hardy, Derrek Lee, Orlando Hudson and Aaron Rowand? All five of those players batted in the bottom of the ninth inning of the all-star game last year -- with home-field advantage in the World Series on the line -- and none of them were selected for the all-star game this year.

What we need is a system where we can identify the real all-stars, and have at least some of those people actually playing at the end of the game. 

So who are the real all-stars? The players who are actually all-stars, not the players who had the best first half of the season. To find these people, we need to select them at the end of the previous season.

One way to do that is simply take the top players from each position in the Most Valuable Player balloting. (Whatever you say about the MVP balloting, the writers do at least pick good players.) 

So for the American League, for example, that would give us a starting lineup this year of Magglio Ordonez, Ichiro, and Vlad Guerrero in the outfield, David Ortiz at designated hitter, Jorge Posada at catcher and an infield of Carlos Pena, Placido Polanco, Derek Jeter, and Alex Rodriguez.

Now, Carlos Pena is a problem. Carlos Pena is hitting .235 with 14 homers this season, and those are about in line with his career averages if you exclude last season.

I can hear some of you now: "What a brilliant system, Matt. You give us Carlos Pena instead of Kevin Youkilis, who is...well, I don't remember his numbers offhand, but at least he's hitting like guy who is not in a slump." 

Youkilis is hitting .319 with 14 home runs and 57 RBI. That's where the other part comes in.

The reserves for the all-star game are the players who just missed out on making the game, and the players who are having the best seasons. Those players are put on a ballot, and the fans can vote on those players online, starting in mid-June. 

Of course, this won't change that much if managers feel the need to remove all the starters and give almost everybody a chance to play.

In the last 10 years, the only all-star starter to play the entire game is Carlos Beltran in 2006. Before that, you have to go back to 1997 when a starter went the distance. 

Now, I know players can get hurt in the all-star game. Dizzy Dean, Ted Williams -- no one wants to see that. But still, if home-field advantage for the World Series is at stake, shouldn't the very best players be in there when the game is on the line? I don't think the chance of injury is high enough to be that much of a factor. 

Plus, all the shuttling around of players -- it's like those Little League games where everybody has to get one at bat or an inning in the field or the team automatically loses. You do it in Little League because you're trying to create the mindset that winning isn't the end-all and have kids feel good at the same time. When you have that mindset in a professional game, you're asking for a farce, and that's what we get.

Maybe I'm a hopeless traditionalist. Maybe I'm an angry radical. Or maybe I'm just a guy with a really stupid idea. Whatever you call it, the all-star game needs a change. I know the idea would be resisted on a lot of fronts, and there may be a problem I'm not taking into consideration. 

But before you dismiss this idea as completely stupid, keep this in mind.

According to the 2007 Sporting News Record Book, the major league record for most players used by two teams in one nine-inning game was set by the Yankees and Red Sox on Oct. 2, 2005. 

That record is 46.

Or nine fewer players than were used in the 2007 all-star game.