[r-t] Blue Line Difficulty

[Mark Davies]

I think a difficulty index, although never an objective metric, is an 
nevertheless and interesting thing. What Graham is trying to achieve 
should probably be described as "a measure of blue line difficulty with 
the assumption that the method and its work is not familiar". I like the 
idea of looking for tricky methods with good music - these may maximise 
interest with a good band, for example.

However, I'm not entirely convinced by the idea of removing the plain 
hunting, as Graham is suggesting. If the length of hunting between bits 
of work is constant, or follows a nice pattern, this will make a method 
easier than if there are unpredictable hunt durations between work. Also 
as Graham has found, it doesn't help with measuring the difficulty of 
individual bits of work.

I think looking at the line as a whole, rather than trying to factor out 
specific aspects such as hunting, is more useful. The basic idea should 
surely be to find the entropy of the information content of the line. 
How many bits do we need to represent it? This must have an analogue in 
the number of brain cells required to remember it.

Measuring the compressed size of the line is therefore the obvious place 
to start, but existing algorithms such as zip are a bit too 
sophisticated to apply well to such a short stream. A basic 
implementation of a sliding-window compression algorithm might give a 
more natural result.

In addition to this, it is I think necessary to give weight to factors 
which ringers find subjectively difficult:

1. Frequent changes of direction
2. Wrong hunting
3. Non-PB leadheads

Separate weighted measures for this kind of thing could be added in to 
form a combined metric.

However you do it, not everyone is going to agree with the ranking of 
methods it produces, but I don't think that matters too much. We're 
trying to measure a subjective concept, so you just have to accept it's 
never going to be perfect. Doesn't mean it's not worth looking at, though.

MBD