For my purposes, VBD is a method of comparing the performances of players who play different positions and players who played in different eras. The basic idea is this: a player derives value by being better than the player he'd be replaced with if he dropped off the face of the earth. Suppose you had Kurt Warner on your fantasy team last year. He was obviously quite valuable to you, as he scored a lot of fantasy points. But what if he had torn his ACL on the first play of week 1? Who would you have replaced him with? Probably someone like Mark Brunell or Jake Plummer or Kerry Collins. You probably had someone like this as your backup. If not, you could probably have traded a backup or two for a QB of that quality.
Anyway, Warner's value to you last year was not the number of fantasy points he scored. It was the number of fantasy points he scored that his replacement wouldn't have. Let's assume, just for the sake of definiteness, that the replacement would have been Plummer. Plummer scored 271 fantasy points last year while Warner scored 392. Hence, our estimate of Warner's value to you was 392 - 271, or 121 points.
That's all there is to it. For every player every season, his value is defined to be his fantasy point total minus the fantasy point total of a theoretical replacement player. That last number (the fantasy point total of the theoretical replacement) is often referred to as the baseline.
Fantasy points = PassYd/25 + RushYd/10 + RecYd/10 + PassTD*6 + RushTD*6 + RecTD*6 - INT*3
First, it makes comparisons possible between players at different positions. Qadry Ismail scored more fantasy points than Tony Gonzalez did last season, but that doesn't mean Ismail was more valuable to your fantasy team than Gonzo was. Why not? Because you probably had someone like James Thrash or Willie Jackson or Hines Ward or Chris Chambers on your bench. If Ismail goes down, you don't have to break the bank to find a replacement who's almost as good. If Gonzo goes down, you're looking at Jay Riemersma or Ernie Conwell. In short, Gonzalez is worth more than Ismail because the difference between Gonzo and Conwell is greater than the difference between Ismail and Chambers. That is, Ismail is more replaceable and hence less valuable.
Second, VBD allows for sensible comparison between players of different eras. Especially for the passing game, the 70s were a very different time than the 90s. In 1974, for instance, Cliff Branch had 1092 receiving yards and 13 TDs. That seems like a pretty good year, but nothing too exciting. But now consider that Branch was easily the #1 WR in the NFL that year. The #2 man, Isaac Curtis, managed only 633 receiving yards and 10 TDs. The #30 WR was Jim Lash, who compiled a measly 631 yards and zero(!) TDs. Put in this context, Branch's 1974 season has more value than, say, Marvin Harrison's 2001 season, even though Harrison had a lot more yards and TDs. In short, Harrison was more easily replaceable in 2001 than Branch was in 1974. Similarly, Bert Jones' 3100 yards and 24 TDs in 1976 was a more valuable season than Scott Mitchell's 1995 (4300 yards, 32 TDs).