You heard me right: Frank Thomas, Barry Bonds, Ken Griffey Jr., Randy Johnson and Greg Maddux are each worth about seven games to each of their teams each year.

I was exposed to this idea a couple of years ago, but it never seemed to make intuitive sense to me. Most explanations of this idea involve protracted examinations of stats: How many runs a pitcher gives up, how much a single, double, home run, or stolen base is worth to the team, how you measure defense, and so forth. Once you put all these factors through some formula, you may get a result, but what does that result really mean? And why should people care if they don't understand - or even want to understand - the formula?

I was thinking about this recently, and hit upon what I think is an intuitive method for demonstrating that in fact the best players are worth only about 7 games per year, and why this is so. It won't tell you which players are worth 7 games, as opposed to only 2, but it provides a team-wide view.

First, we start from the idea that some players are worth zero games. Such a player is sometimes called a "replacement-level player", the sort of player that any club can get for free. They can call some guy up from the minors and stick him at first base, and he will get a few hits, make a few players, but basically do no better than any of a thousand similar other guys. He costs the team nothing. I'm going to call such a player a "scrub".

(There are players who are worth negative games during a season. Such players are usually people a team called up to be a scrub, but who wasn't even good enough to match himself against the scrubs. But if you pick someone at random, he's about as likely to be a bit better than nothing than a bit worse than nothing; across all of baseball, these players will average out to a scrub, who's worth zero games. So let's stick with that.)

The key thing to remember about scrubs is this: Scrubs can win games!

Suppose you're a big-league manager, and - for whatever reason - you're forced to field a team of scrubs, while the other teams are fielding "real" Major League ballplayers. Even so, you're going to win some games. Once in a while an opponent's pitcher will just not have his stuff (or anyone else's stuff, for that matter), or the opponents will boot a ball, or your players will all just happen to hit at the same time, and score a few runs. Over a 162-game season, anything can happen once in a while.

So the question is, how many games will an all-scrub team win in a season?

My answer is: About 40.

Why? Because that's how many games the 1962 New York Mets won, and if there was ever an all-scrub team in baseball history, Casey Stengel's Amazin' Mets was it. It's hard to conceive of a team much worse than the 62 Mets. Sure, there have been teams which won fewer games, but you have to go back to the Dead-Ball Era before 1920 to find them, and the game was quite different back then. Even then, the truly awful teams won around 40 games (e.g., the 1904 Senators, who won 38, but only played 151 games).

So, scrubs can win games. A scrub team will win about 1 out of every 4 of its games in an otherwise standard league, which works out to 40 games in today's 162-game season.

What this means is that the value of a non-scrub player is the number of games he wins for his team over and above those 40 games. No one gets credit for those 40 games, because anyone can win them. You'd actually have to work to win fewer games. Players only get credit when their team wins more games with them than with a scrub.

Having looked at the low end of the scale, let's now look at the high end: Rather than the all-scrub team, we're now looking at the all-star team. The real all-star team, where every position is played by a star player for a full season.

We're talking about a team with a first baseman who hits 50 home runs; a shortstop who turns spectacular players and hits like Cal Ripken Jr.; his double-play partner at second who bats .340 and drools whenever he sees the Green Monster in Fenway Park; the MVP left fielder; the power-hitting catcher who can run the bases and handles the pitching staff like Yogi Berra; and, of course, a rotation filled with the best fireballers and finesse artists and a bullpen of hard-throwers with pinpoint control. Even the bench has the best players for the job.

How many games will such a team win? How many games can any team win, when they're the best possible team? And how much are those players worth?

For simplicity, I'm going to start by making two assumptions: First, I'm only going to consider the regular players and the starting pitchers. Second, I'm going to assume that each of these players is worth six games per year.

So, our regulars: Eight position players, and five starting pitchers, or 13 players in all. (We'll ignore the DH.) Each is worth 6 games. Now it's just a matter of math:

40 + (13 x 6) = 40 + 78 = 118 games

Let's consider: The 1927 New York Yankees, so-called "Murderer's Row", won 110 games in a 154-game season. The 1954 Cleveland Indians won even more: 111 games. Bob Gibson's 1967 St. Louis Cardinals won 101 games - out of 161, this time. Earl Weaver's Baltimore Orioles won 109 and 108 in 1969-70. The "Big Red Machine", the 1975 Cincinnati Reds, won 108. The 1980s topped out with the 1986 Mets (108 games) and the 1988 and 1990 Oakland A's (104 and 103, respectively). In 1993, the Atlanta Braves and San Francisco Giants finished 1-2 in the NL West at 104 and 103 games, respectively.

So, yeah, 118 games is a lot. An awful lot. An actual team winning 118 games in a season would be just as shocking as an actual player hitting 62 home runs or an actual pitcher winning 30 games.

But wait, we're not done. We've ignored the bullpen and the bench. The bullpen is probably worth at least another full player among them, since starters today can't go 9 innings every day. And bench is probably worth less than a full player, since on a real all-star team, the stars wouldn't get hurt as much or need to rest as much, and there'd be little need for platooning. So between them, add two full players, or 12 more games.

So now we're looking at an all-star team winning a whopping one hundred and thirty games in a 162-game season! That's nineteen more games than the 1954 Cleveland Indians won! That's a winning percentage of eighty percent! The 1995 Cleveland Indians won nearly 3 out of every 4 games they played. This team could win more than 4 out of 5! No team has ever been this good over a full season!

Sure, they'd lose a few games - a whole 32 games, actually - but everyone has an off day. Maybe one day Hideo Nomo or Roger Clemens would shut them down. Another day, the Indians or Rockies would figure out their pitchers and club a few homers off of them. But not often. Only once a week, or sometimes twice a week.

And remember: This assumes that our top-flight players are worth six games per year over what a scrub is worth. Just six games.

In reality, some players are worth more than six games, just as a few players are worth less than zero games. A superlative season such as Greg Maddux has had in 1994-95 might be worth 7 or 8 games to the team. Babe Ruth probably had one or two seasons where he was worth 9, 10, or even 11 games to the Yankees. But only one or two, I'd venture (Total Baseball calculates that he peaked in 1923, when he was worth 10.6 games to the Yankees). Ruth was a man ahead of his time, and baseball has advanced so far that it's hard to find such players these days. Today, the best players are worth 6, 7, or maybe sometimes 8 games.

They can't be worth more, because there aren't any more games to win. It's just simply not true that one of Mo Vaughn, Frank Thomas, Shawn Dunston, Chuck Knoblauch, Albert Belle, Mike Piazza and Ron Gant is worth that much more than one of the others. Mo Vaughn may have won the 1995 AL MVP award, but was he worth twice as much as Chuck Knoblauch was? To be worth 12 games, he'd have to be, because Knoblauch was the best second baseman in baseball in 1995.

Reasoning this all out has had one major effect on my thinking about baseball: I no longer believe that one player can "carry his team" for a full season, or even for a large part of it. No matter who that player is, his team is going to win 40 games without him. And most teams in the Majors have several star players, as well as many average ones (who are still much better than a scrub). When you're looking at a 90-game-winning team, 40 games are credited to no one, and the remaining 50 have to be parceled out among several pitchers, and several hitters. Winning teams simply don't have players who are that much better than anyone else on the team, much less everyone else on the team.

Baseball is truly a team sport; you can't win it by yourself. Sure, a dominant pitcher can come close, and be maybe 80% responsible for the games he pitches in. But starters only play one game in five; the other four days it's someone else's job. A great batter can club home runs 'til the cows come home, but without his teammates getting on base, and without the pitcher throwing reasonably well, his team's just not going to win very often.

So when your favorite slugger hits a home run to break up that tie game, cheer him on, and be happy he's on your team. But don't give him all the credit. Remember, he had help.

Tom Glavine was probably pitching.