Part of the problem is that they don't chain their runs off each other, nor do they run PA out of their most common or successful run plays.
For example, we had some early success with the counter O, but we don't run anything off of the same look. That is, with a guard and TE pulling towards the 'backside' of the play, we don't run anything where those guys pull and the RB goes towards the frontside. Maybe if we ran a trap or a dive where those guys both pull and block towards the middle? It would give, initially at least, the same look, but they would be two different plays. If the defense starts biting hard on one play, run the other and watch your RB pull in 15-20 yards a pop. An A-gap dive or a variation on an inside zone split would work wonderfully. Add in a play action off the same look, and that's another possibility for a big play.
Right now, there's no real continuity or connection between the plays being run. It's about as random as pulling plays from a hat.
It also ties into identity. Right now, we don't have an identity, we don't have a core group of plays we can point to and say 'this is what we like to do'. A good offense starts with a core group of plays. For example, inside zone, outside zone, bubble screen, slant bubbles, levels, corner curls. From there, plays are added that, initially, look identical to the core plays, but go different directions. Like a zone counter split, read option, inverted veer, fake bubble fade, corner post (with an underneath out route) and so on. Finally, a smart OC will add in situational plays, meant to beat specific fronts and coverages based on who they play any particular week. And if you've got a good enough QB or an excellent communication system (signs or signals from the sidelines) you can put in hot routes and audibles to beat particular blitzes, especially ones your opponents like to run.
Anything else on top of that is just style. The core principles remain the same, regardless of what 'system' an offense runs. West Coast, pro style, spread, even being 'multiple', whatever, doesn't matter. Establishing an identity by having and running a core group of plays makes everything an offense does easier. A defense will tool itself to try to stop your core plays, but by running a play that directly exploits a defense expecting something else, and you'll have huge, backbreaking explosive plays. And if they don't adjust to stopping your core plays, since they're the plays your offense is the most familiar with and (hopefully) the best at running, you'll still be able to drive down the field.
Osborne's teams did this. The counter trey was specifically designed to exploit defenses that were over committing to stop the power O. The read option was specifically designed to take advantage of overly aggressive 'contain' defenders.
Right now, our offense doesn't seem to have any 'core' plays, let alone any counters to our core plays. They're not chaining plays together, and many times it seems like each new look only has one play that's ever run out of it. It's a poorly designed offense. While they have some well designed individual plays (at times), there's no continuity between plays, nothing that really links them together. That's why defenses are able to key in and stop up seemingly at will.
At least, that's my opinion anyway.
