Accuracy statistic – new feature

1000minds has released a new accuracy statistic, giving you the power to tune the accuracy of your decision models.

Accuracy is a measure of how well you have ‘trained’ your decision model to reflect your preferences. The more trade-offs you make, the more confidence you can have in your decision model’s outputs.

Accuracy is especially important when scores (rather than just ranks) matter, e.g. for cost-benefit analysis / value-for-money analysis.

You can increase accuracy by making more trade-offs, especially those involving more than two criteria, and by increasing the number of levels on your criteria. But how do you know when to stop? That’s where the new accuracy statistic helps you.

Technically speaking, the accuracy measure is driven by the range of possible preference values that are consistent with your trade-offs. This range (solution space) gets narrower as you make more trade-offs – allowing 1000minds to zero in on your preferences. More formally, the accuracy statistic is about decreasing measurement uncertainty.

The new accuracy statistic is presented, along with more extensive help, when you've finished answering trade-offs involving two criteria. Give it a try!