The word (or pattern) length is a critical parameter that can have a significant impact on sensitivity and speed. The default settings work well for most applications, but for very large datasets and other resource-intensive tasks, it can be worthwhile to experiment with different values. You're welcome to ask me for advice.