Quality of Competition

[Pietraneglo222]

the statement is as literally true as you can get. read the results there and here, they're excellent - denying the stat because you think its not parallel with the Number Revolution is pointlessly ignoring a useful tool

I don't agree that it's useful. The point of studying the stats analytically is to correct the coaches and GMs' misconceptions. If we assume they're right then there's really no point to it all.

 

[overpass]

It is, and I put the "proof" forward. It's a solved problem unless you have a valid objection, which you don't because all I've seen from you is conjecture and rambling.

MOD

You looked at two columns of stats and you thought you solved something. [Mod: in your opinion] the stats describe only in part. MOD

 

[overpass]

For anyone who's actually interested in the topic, this article documented a look at the impact of QoC with multiple iterations.

Brian MacDonald's adjusted plus-minus adjusted for QoC and QoT over multiple iterations using regression.

 

[barneyg]

It's not "i believe". I've mathematically and unequivocally shown that QoT varies by an order of magnitude more than QoC. Making QoC "largely irrelevant", at least in relation to QoT. (Please keep that wording and avoid the obvious strawman you just made in your last post.)

I don't know why you're talking about "actual tests" when we're not talking about theory, we're directly observing the profession as it happens with real data. If it shows up on the data, it happened on the ice and is therefore "an actual test".

You are responding to a personal attack that doesn't exist, please read my previous post. I don't disagree that QoT varies more than QoC, my post had nothing to do with that.

You didn't quote "an actual test" of the hypothesis that "a QoC adjustment wouldn't significantly change Corsi". To conclude on whether something is significantly different or not, you need to compare adjusted numbers with the benchmark. That's not what the blog post showed. If some guy falls from #1 in the league on an unadjusted list to #40 in the league on an adjusted list, that's a significant adjustment in my mind. I'm not even talking about iterations.

 

[Micklebot]

More precisely you have presented some numbers that show that single-iteration Corsi QoC had a smaller range than single-iteration Corsi QoT for the 2011-12 regular season.

How does this change when multiple iterations are used to calculate QoC and QoT? Taking into consideration that if good players tend to play against good players, single iteration QoC will underestimate the true range of QoC, and if good players tend to play with good players, single iteration QoT will overestimate the true range of QoT.

How does this change when looking at plus-minus numbers instead of Corsi numbers?

How does this change when looking at playoff numbers instead of regular season?

How does this change when looking at other seasons than 2011-12?

I think QOT will be more stable across iterations, because the variation in line-mates is far less than the variation in opponents. Usually, a player is on a set line, or d pairing. His opponents can be all over the map. Not that the effect your describing wont be there, I just think it wont be as pronounced with QOT.

 

[Buffalo87]

Analyzing CorsiRelQOC and zone starts

Hey guys, I just noticed this board and started looking through. Some pretty awesome stuff here, great idea to start it up.

Anyways, occassionally I throw stuff together and put it up on my blog. Here is the last one I did, a month or so ago (been pretty busy lately), it's on comparing Corsi Relative Quality of Competition and offensive zone starts using CorsiRel, shots for per 60 min, and shots against per 60 min. Essentially to gauge the impact of each on certain shot-based metrics.

Take a look and let me know what you think.

With so many new metrics coming out in the developing area of hockey analytics, it's important to keep both an open mind to these new statistics, but also a critical mind in evaluating the usefulness of them. This isn't to say any of these numbers are useless, but some are taken to have more significance than they really hold. Sometimes it's important to evaluate some of these numbers to keep them in perspective, and see what they really tell us.

For instance, offensive zone start rate and Corsi quality of competition are two primary "new" metrics that are finding their way into a lot of casual hockey conversations, something you likely wouldn't have seen 12 months ago. These two statistics are what comprise the popular player usage charts which essentially evaluate hard minutes vs easy minutes. People often hold CorsiRelQOC and offensive zone starts on an even plane, assuming that each has equal weight in providing context for a player's production. Is this really the case though? If we were to standardize the values, would an offensive zone start rate of, say, 40% provide an equivalent disadvantage as carrying a CorsiRelQOC of, say, 1.0?

http://www.thehosers.com/2012/09/a-critical-eye-toward-statistics.html

 

[overpass]

Hey guys, I just noticed this board and started looking through. Some pretty awesome stuff here, great idea to start it up.

Anyways, occassionally I throw stuff together and put it up on my blog. Here is the last one I did, a month or so ago (been pretty busy lately), it's on comparing Corsi Relative Quality of Competition and offensive zone starts using CorsiRel, shots for per 60 min, and shots against per 60 min. Essentially to gauge the impact of each on certain shot-based metrics.

Take a look and let me know what you think.



http://www.thehosers.com/2012/09/a-critical-eye-toward-statistics.html

Good article. But I think you've gone a bit too far in your conclusion.

the entire reason for this analysis was to see whether offensive zone start rates or CorsiRelQOC are more of a determining factor in a player's performance. Does a change in CorsiRelQOC affect a player more or less than a change in offensive zone starts? It's pretty clear through the comparisons we made that offensive zone start rate has a much larger effect on a player's production than CorsiRelQOC.

I would amend that last sentence to say offensive zone start rate has a much larger effect on a player's CorsiRel than CorsiRelQOC. Your entire analysis was been limited to performance in terms of driving shooting differentials and has omitted the percentages. Which is fine, nothing wrong with a limited analysis. But I don't think one can draw conclusions from a limited analysis about overall terms like "performance" and "production".