Quote:
Originally Posted by NegativeSpace
Yes. The more I read about xGF the less I think it’s a significant measure of success. It is useful for noticing longer term trends but it has some limitations of usefulness.
|
I also think we tend to use it wrong in some scenarios. Analytics supporters (myself included) have always gravitated towards using 5v5 numbers for analytics.
And that 100% is the right decision when comparing players - because for an individual player their PP/PK/5v5 ice time mix and usage can greatly skew their xGF. You can't compare all situations for a forward that plays lots of PP and no PK, to a player that plays lots of PK and no PP.
But when looking at team metrics and team results we really should be looking at All Situation metrics. Special teams are too big a part of the game to just ignore them when assessing team effectiveness and you're excluding over 20% of the ice time data we have if you look only at 5v5.