WAR = Wins above replacement. So assuming if you have an absolutely average player, versus the player that you plug in, plugging in the player with the +WAR means that your team is projected to win X more amount of games. Obviously, that also means that there are players with 0 WAR, meaning they are league average, and those with negative WAR, which means they make their team worse when they are on the floor.
I briefly mentioned about this topic in another thread, but it's worth recapping:
CARMELO stat projections LOVE Marcus Smart!
It is projecting his WAR to be 5+ for the next 5-6 years.
This translate into Smart having a winning value of over 150 million dollars over that time period!
http://projects.fivethirtyeight.com/carmelo/marcus-smart/If Smart lives up to the projections he would be the second most valuable player on the Celtics behind only Al Horford this next season!
Just to give you an idea, CARMELO likes Marcus Smart better than Blake Griffin or Okafor over the next 5 years. Only Westbrook represents an upgrade in terms of winning value according to that metric!
Griffin projected WAR:
http://projects.fivethirtyeight.com/carmelo/blake-griffin/Okafor projected WAR:
http://projects.fivethirtyeight.com/carmelo/jahlil-okafor/Basically, Smart is projected to have higher WAR over the next 5 years than Griffin OR Okafor which gives pause to the idea of dumping Smart for peanuts. He shouldn't be moved unless we are getting a true jigsaw puzzle piece that puts us over the top. For example trading Smart for Okafor is seen as a net loss as far as WAR, and for Griffin would be a wash, but for Westbrook would be an improvement.
Westbrook projected WAR:
http://projects.fivethirtyeight.com/carmelo/russell-westbrook/