The problem with looking at their VORP or WAR is that those are counting stats. Typically when a player is performing at a rate near or below replacement level he won't be given the opportunity to compile a lot of negative value. Therefor, in my opinion, simply looking at players VORPs or WARs when they are near replacement level is extremely flawed. In addition, you are looking at a relatively small sample size, especially for Bush. A rate stat such as xFIP (while still flawed) will probably provide more insight into how sucky those pitchers in question were last year. Player: xFIP Para: 4.64 Bush: 4.79 Looper: 4.87 Soup: 5.26 Granted, someone like Ruz could probably provide a lot better statistical analysis, but I think some simple conclusions can be drawn here. I think a replacement level xFIP for the NL is around 5.25. So last year, even with Bush's injury, and Parra's absurdly unlucky BABIP and mental issues caused by that, both were still significantly better than Soup. In addition, going forward I would expect Soup to be the only starter on this team to perform below replacement level.