dWAR isn't the figure that you want. It's designed to compare cross-positional value, and so first basemen always do poorly in it.
There are two components to dWAR, one based on runs saved with the glove, and the other is a positional adjustment. You need the positional adjustment because it's easier to find someone who can play first base than it is to find someone who can play shortstop (for example), so a competent shortstop is more valuable to his team than a similarly competent first baseman. Guys who play hard defensive positions get a positive adjustment, guys who play easy ones get a negative one.
If you want to see how many games a player won/lost with his glove, look at Rfield and then divide by (approximately) ten. Keith Hernandez won his teams about 12 games with his glove. His dWAR is so low because of the penalty that all first basemen get. It doesn't mean that WAR says he was bad defensively - quite the opposite, having positive dWAR is very good for a first baseman. It means that a shortstop with a similar Rfield is more valuable to his team because he's harder to replace. (Of course his bat is likely to be worse, but that will be reflected in the offensive component of WAR.)
|