In my template-matching code I need the user to pass a floating-point parameter, which specifies whether the algorithm should concentrate only on the best matches (thus work faster) or analyse even low-probability areas (making it slower).
The parameter is linear and normalized such that possible values are in range <0, 1>. It doesn't matter whether the number of resulting matches increases or decreases as the parameter grows, as this can be easily changed. The default value is around 0.5; when the value is on one end of the range, the algorithm should possibly return only one match (and work fast); whereas the other end should mean lots of possible matches and long processing time.
What name should I choose for this parameter such that it makes sense to the end-user? I've been thinking about MatchingQuality or MatchingDepth but neither seems appropriate and self-explanatory.