No narrative? You ignored my post giving outlandish amounts of praise to Ponting, claim I can never admit error when I have done so twice in two posts, and are not addressing what causes high impact averages.
Meant it as what this piece defines as important, which could be clutch.
But it doesn't men important runs count for 1.5 and less important ones for zero.
That's a fair enough point actually. Maybe need a sliding scale of importance.
If you check the original discussion thread, I had initially proposed more quanta, but I didn't like it for various reasons.
What I wanted to do here was to find out how a batsman performed based on the state of the match when he came to the crease, not the impact of that innings. That's what my match impact measure did.
Your goal is to measure volume of potentially impactful runs defined at start of innings rather than actual impact of runs themselves? Needs as solid a definition of potentially impactful runs as possible. Not always easy to do so independently of actual impact. My runs from a losing position creates an issue here. Bad example under the circumstances. You indicated via the Lara example why it's problematic to address. But in catering to the exceptional aren't you ignoring a potentially valuable set of data? Might make sense to exclude less than x% chance of winning at time of arrival at crease and allowing a couple of outliers to fall below this point. Or lower if it makes more sense to you. Below whatever win % has ever been historically managed maybe if you want to be really conservative.