Creative Speculation

... which was indeed to "get the results they wanted," but what they wanted was to get the model to agree with known observations (e.g. the WTC7 east penthouse collapsing first). An FEA is not a magic simulation of the real world; it's just a mathematical model where ever input and every interaction between elements needs to specified by the model builder. The problem is, in many cases the input parameters are unknown. So, what is done is to vary the parameters, within the plausible physical limitations, until the model behaves in a way that matches observations. If it does match observations, that doesn't necessarily prove that the other things the model is showing are accurate, but if a model doesn't agree with known observations, there's no reason to think the rest is accurate. It's one thing to question how NIST went about that and how accurate their model is, provided the criticism is objective, but it's quite another to misrepresent what was done and then accuse them of complicity in a mass murder based on that misrepresentation.