Ohhh Franco
Registered User
- Jun 28, 2006
- 1,551
- 5
"Advanced stats" is so overrated it's unbelievable.
At least with baseball it's hitter against pitcher and you can kind of justify it but even in baseball...
Oh, I understand statistics very well and I understand their use, limitations.
If your go-to strategy to find players, build a team is basically HockeySabreMetrics - you're in trouble.
Statistical work can complement scouts, interviews (if possible) with the player and other things but the weight put on the stats in particular in recent years is wrong... in my opinion.
The movie Moneyball was fun but you have to remember the Oakland As didn't run Sabremetrics to an extreme like the movie would have you believe.
Edit: And apparently it won't let me edit to Sabremetrics. Lame. Edit2: REALLY? Sabre metrics. Edit3: for ****s sake e and THEN r
I know for a fact that the team tracks advanced statistics internally but doesn't really communicate them for fear of alienating casual fans. No idea WHO tracks them, but they definitely are monitored.
If thats true, why do we always seem to re-sign the guys that have terrible advanced stats?
And let go of Da Costa?
If thats true, why do we always seem to re-sign the guys that have terrible advanced stats?
And let go of Da Costa?
If thats true, why do we always seem to re-sign the guys that have terrible advanced stats?
And let go of Da Costa?
The movie Moneyball was fun but you have to remember the Oakland As didn't run Sabremetrics to an extreme like the movie would have you believe.
Just like Tim in Buffalo. Taking over the division one team at a time, soon all the prospects will be ours.
Arent we implementig a whole bunch of new cameras that track players, etc? Would that not help to track some forms of advanced stats?
As per the sens model they have had a guy doing this for years, Yost! I mean why pay for those things when you can get it for free, cost per point etc...half joking here.
Seriously though i hope the team has been into this for ahwile now as it was failry clear to most critical thinkers when hockey advanced stats first came to light that there was certainly something relevant and useful there. Another tool to add to the arsenal so to speak.
Article on Dallas and Jim Nill building with advanced stats. Wonder where Jason fit into that.
http://www.dallasnews.com/sports/da...why-he-wants-better-technology-in-the-nhl.ece
That tech is being piloted, and will be league wide within a few years. It will definitely revolutionize the types of stats that are collected. E.g. Corsi etc are a proxy for possession, but with full cameras and tracking technology it is theoretically possible to measure actual possession accurately. It will also do puck speed, player speed, and stuff like that.
This would certainly make me start to believe in advanced stats as a useful tool.
I'm not buying what is currently out there, but admittedly I haven't really looked at the data behind it.
Can anyone provide a link (or links) to the best site to justify the advanced stats? Note that I'm not looking for an explanation of what they are -- I think I know that already -- I'm looking for the statistical analysis that proves they are valid predictors of individual and/or team success.
The problem is, when most people on this message board refer to "advanced" stats, they're usually talking about Corsi or Fenwick. You can just google "correlation between Corsi (or Fenwick) and winning percentage" and you'll get a number of analyses outlining how both these metrics are effective (or not effective depending on the opinion of the author). There are other advanced stats that teams track that go beyond those two, and it's known that teams like the Blackhawks have developed their own, so it wouldn't make sense to completely dismiss "advanced" stats or teams that are using them because we generally have no idea what sort of data they have, what sort of metrics they've developed and how well these correlate to winning.
What the new technology they plan on rolling out within the next few years provides, is an opportunity to get even better and more reliable data compared to the tools we have currently.