"Building on a Faulty Foundation"
The quality and rework in a project is another key dynamic of our simulation. The total error rate (far left side) determines how many errors are put into the project as work is completed. The more error generation the higher the cumulative errors are in our project. Future work that is done based on this poor work will have a higher error rate from poor pre-work and so the total error rate will rise. This is a self-reinforcing loop. If we do nothing to remove errors in our work, the rate of error production will continually rise.

For example, misidentifying the database protocol in initial systems specifications in an IT project would lead programmers of the data related modules to use the wrong database calls. Each of these errors would have to be fixed individually. In construction, a poor job of framing can lead to crooked wall panels and room layouts out of square. Poor room layout means re-cutting and reordering floor coverings or tearing out the faulty walls.

"Remove Errors"
The most common remedy for errors is to have some error identification and removal process that takes out the accumulated errors. There are various approach and philosophies but the dynamics are similar.

"No Time for Error Work"
In many projects there is little time specifically allocated for these error reduction activities. As the team tackles these errors with their limited resources, the scheduled tasks begin to fall further and further behind. When the schedule adherence is sufficiently low, the team implicitly or explicitly decides to cut the error reduction activities and the quality standard. A typical thought process is " We are so far behind. We must be having too many meetings. Let's do our QA meetings every other Friday instead of every Friday." The net result is that we are QAing twice as much work in a meeting of the same length and we must lower our standards to make this work.

"No Time to Align"
The second big source of errors is the errors that we make every day in the course of our work. These are further subdivided in our simulation between "natural errors", the normal error rate at which people can be expected to make mistakes, and errors from poor alignment, the rate at which errors are added because team members are not on the same page. Errors from poor alignment is primarily driven by time spent on alignment. Required communication rises in proportion to the square of the team size. As the team grows, it must expend more time on communications or they will increase their error rate. Although widely understood in the project community, time spent on alignment is often one of the easiest things to cut since its impact is delayed and largely invisible. When the project gets behind in the game the team has the opportunity to decide the relative importance of time spent on alignment.