Yeah it is an interesting stat that I don't have complete data for to try to put this into context or historical context. It would be interesting I think to look at all the RB and how often they might gain 0 or less yards on their rushing attempts.
As the article states, 32% was the second worst in the league, (which makes me wonder who was worse than Mack) in 2017. Looking at the last 3-5 years of this would give it more context than this, but I think it is fair to say that 32% is in the bottom 10% or so of RB if we did look at a larger time frame.
I think a lot of coaches will bench a RB if they gain 0 or lose yards this frequently, so not many RB would get the chance to do worse than Mack as far as this goes. They wouldn't get the opportunity to do worse than this.
It is a stat that actually reflects what myself and others saw Mack doing with a high frequency. Trying to reverse field instead of putting his head down and taking 2-3 yards when the defense plays well instead of trying to create a big play by running against the play design.
That creativity and the big plays are something you want the RB to do to a certain degree and you do not want to coach that out of the player, however the losses and zero yard gains are something coaches want to limit. The two things are somewhat related. I am not sure how you coach Mack to get rid of the negative plays, but still keep the positive ones.
I don't really know of a comprehensive way to look at this for all RB.
On PFR you can get a list of the players total plays and then could count them to find the number of times Mixon did this. PFR does at least put the plays in order, the longest yards to the fewest.
Joe Mixon
I kind of knew this anyways, but the most frequent outcome of a rushing play is a RB gaining 2 or 3 yards.
For Joe Mixon I count 39 times he gained 0 or less yards last season. This was 21.9% of Mixons runs. So about 10% less frequently than Marlon Mack.