shooker wroteCOLONNick wroteCOLONGuys, there is a fundamental problem with any TOI measure. First and foremost it's a goalie GP stat, secondly it's a skater GP, and finally it's a skater TOI measure. That needs to be remembered and understood.
we understand it, goalies get shtoi already. it is an amount you can overcome when it is say 8 minutes a night rather then 60 per goalie gp. hate to break it to you but G/A/Pts/StPTS/Hits/BS/FOW/Shtoi etc are already biased towards teams with more gp, that is a reality we already live in.
Its a joke that you are 'breaking' anything to me.
Fundamentally you're making a statistical error regarding impact. There is no unique lack of impact from finesse dmen, it's a shared lack of impact with all finesse skater positions. I just showed (if you care to read or understand how stats actually work) that Getzlaf is a better piece in this league (or a net impact measure) than Sidney Crosby (even in a relatively low scoring year by Getz), is that reflective of the NHL? But it is a necessary value-adjustment because we wanted to have a league where building in the 'right' sort of bottom players, matters. We could remove hits/bs/shtoi and have a scoring league like every basis fanasty experience there is out there.
With all non-toi measure there is more variation contributed to the individual player than there is for a GP. For toi measures that is not true, GP at the very low example (shtoi/pptoi) is still so little variation between players that GP is the first control.
Last year the NHL median was 11.27 PIMS/gp/team, compared to even the best shtoi per GP, GP is still the biggest factor.
a few more interesting z-scores (as shiv noted, no measure of cap - I'm sure I could standardize cap and make divisor wherein we can see impact per cap used):
Ovechkin with a MASSIVE impact of 2.273354753 stdDev AVERAGE above the mean - he's the top 1.1502% - also the highest paid player - makes a lot of sense IMO and he's a rather important player in the NHL
.
Oshie, with an approproate 0.685901249 stddev above the mean, meanig he's in the top 25% - good player, but not individual week changing impact.
MSL a nice clear example of how we undervalue finesse players, with an impact score of 1.138834471 std dev above the mean, meaning he's in the top 13%. Not quiet reflective of his actual NHL impact.
Jamie Benn a 0.885989312 putting him in the top 18%. Interestingly, Patty Kane is in the top 7.1472%, despite being UNDER average on 5 categories, he's so far above in the other 7 that his total impact is rather large. This is reflective of the nature of the individual stats, because ther average player doesn't have any FOW, the spread on the category means not contributing to that category isn't a big impact in distance from the mean.
I would note, that MSL's impact is not overly different (especially not year to year) from Shea Webers. and is JUST above Phaneuf (1.12092771).
For net impact, it should be no shock to anyone, that well-rounded players make a bigger impact. Player who do not contribute in various categories, are ultimately below the average in many categories, which minimizes the importance of their dominace in a few categories.
I'm going to be honest here, I did not want to bring it up, but this is actually a tied-in issue with our scoring system issue. Wherein on the season, winning more
categories is more important than winning more
weeks. IE our first place team is not necessirlary the team that won the most weeks, it is the team that won the most categories. So in the regular season you're better served by trying to win more categories, and not just trying to win the game. Anyways, this is a topic already covered this summer, but adding another category, specifcally onethat is single position relevent, makes this issue bigger.
The deeper I get into the numbers (and likely lose the non-advanced stats inclined) the more evidence I find that we value, as per design, well rounded players more than any specializing players. Adding any value to scoring players, although it would justify their cap more easily, also increases the value of these well rounded players, and hence elimating it's very purpose.
In order to get Ryan Suter worth more than Dion Phaenuf, we'd need to remove PIMS, HITS and add DefPoints. Even in that case they would be practically equal value. Guys that don't show up on the stats sheet, are hard to give value to, adding more catgeories that they do show up on, may increase their value, but it also increases those who are all already appropriatley valued ( a confounding effect).