Opinion: Detecting Disaster Projects

Computerworld - If you've been in this industry for any length of time, you've probably been caught up in some sort of project disaster. They happen to the best of us, and they cause financial suffering for our companies and personal pain for all involved.
Careers are trashed and personal lives disrupted.
Even by optimistic estimates, about 75% of projects are late, over budget, missing major functionality or canceled outright. So depending on your definition, most of our projects end up somewhere between failure and disaster.
There are several important things to do once you realize that you're facing
a disaster in the making, but you shouldn't do any of them until you are really sure that it's an impending disaster you're up against. So the first key to disaster recovery is disaster detection.
Given that so many projects go astray, you'd think that we'd be better at detecting these sorts of problems. Heck, our default assumption about projects should be that they're in trouble. But that's just not the way we're built.
Why is it so hard to know? Well, I've got a few theories.
No real plan. If there's no baseline to work from, no one really knows that a project is late. Many projects never get to the stage of firming up a detailed plan.
Excessive optimism. In many teams, there's a perpetual optimism that just because the project is behind at the current time doesn't mean that they won't soon catch up.
Fear of admission. When a project team is in trouble, no one wants to go to senior management and admit it. That might bring uncomfortable scrutiny, blame and retaliation. It's easy for team members to delude themselves into thinking, "Maybe no one will notice. Maybe things will get better. Maybe I'll find a new job before someone finds out."
So how do you figure out that you're getting into trouble? How can you monitor projects for those early warning signs that things are going off the rails? Here are a few things to look for:
Poor team morale. This is probably the biggest thing to look for, not because it's the leading cause of project failure, but because it's a great indicator that something else is wrong. Many of the other things listed below may first be visible in the team's morale, since team members will probably be aware of project problems before you are.
Poorly understood team roles. If the people on the team don't seem clear about what their individual roles should be and