CF% 5v5 Correlation to SC Winner

[KingDeathMetal]

One thing that always bugged me about CF% 5v5 as an indicator of SC Winners is that, in the relatively short amount of time that we've been able to track CF% as a team stat, two teams (Chicago and LA) account for 5 of the 10 Stanley Cup winners since 2005-06. Those Chicago/LA teams won without changing their coach or their core group of players.

The other 5 SC winners since that time are basically split between strong possession teams (Anaheim, Detroit), very weak ones (Carolina, Pittsburgh), and a middling possession team (Boston).

Is the strong correlation between CF% 5v5 and the eventual SC winner from all seasons since 05-06 heavily influenced by the particular playing style of the two teams that have dominated the league since then? We really don't have a diverse number of winning teams to analyze in terms of possession.

That led me to this article on CF% 5v5, PDO, and Goal Differential as a recipe for playoff success:

http://thehockeywriters.com/stanley-cup-recipe/

Don't know if this had been posted here last year when the article was written, but the numbers are interesting. It doesn't seem as though strong CF% teams do much better round to round in the playoffs than the strong PDO and Goal Differential teams, not to mention the significant portion of hockey that isn't played 5v5.

I can't seem to find enhanced Team Stats for playoffs (War on Ice is blocked at my job, so maybe it's there?), but I have looked this up before and recall seeing that the last five SC winners were all Top 5 in PDO in the postseason during their winning years. So while they carried strong CF% 5v5 into the playoffs, they all benefited from hot goaltending and accurate shooting when it mattered. Yes, some of the better PDO teams did not advance to the Finals, but many of them (think Anaheim and NYR) were able to consistently reach the later rounds and force elite Corsi teams to 7 game series in the Conference Finals that were decided by one or two goals (last season).

(One other thought: Unlike some of the other strong possession teams, Chicago and especially LA are among the few that consistently post great CF% 5v5 when leading, not only when close, tied, and trailing. Is it possession, or a trait that is specific to those two teams that makes them so successful?)

Some of you guys have probably looked into this much deeper than I have, so tell me: has the predictive power of CF% 5v5 been overstated? Have two incredibly successful teams colored our view when the answer to "who is likely to succeed in the playoffs?" is probably more complex? Is PDO ultimately king when we consider that SC winners have ridden PDO hot streaks in the playoffs in addition to their sound possession game?

I'm genuinely interested, and these are my main sticking points that make me a bit skeptical of using CF% 5v5 as a guide for who we are likely to see playing in May/June. Where am I right/wrong?

Thanks!

Click to read more...

 

[Dertell]

The other 5 SC winners since that time are basically split between strong possession teams (Anaheim, Detroit), very weak ones (Carolina, Pittsburgh), and a middling possession team (Boston).

Not really. After the penguins fired Therrien, they were one of the best possession team in the league with a 53.6 SA CF%. So essentially, since that time it's five strong possession teams (Anaheim, Detroit, Kings, Penguins, Blackhawks), a very weak one (Carolina @ 48.5 CF% SA) and a middling one (Boston @ 51.5 CF% SA).

As far as your point about PDO is concerned: yes, going on a hot streak / getting lucky at the right time is important, but the team that wins a game in general - playoffs or not - will always win the PDO battle. There's just not enough goals scored in a small chunk of games to make sure a single goal doesn't influence the SA/GA ratio to a huge extent.

I'm pretty sure the 2014-15 ducks were a pretty good Corsi team after the TD. EDIT: 53.7 SA CF%. To be fair, it's a small sample size (19 games).

 

[bossram]

One thing that always bugged me about CF% 5v5 as an indicator of SC Winners is that, in the relatively short amount of time that we've been able to track CF% as a team stat, two teams (Chicago and LA) account for 5 of the 10 Stanley Cup winners since 2005-06. Those Chicago/LA teams won without changing their coach or their core group of players.

The other 5 SC winners since that time are basically split between strong possession teams (Anaheim, Detroit), very weak ones (Carolina, Pittsburgh), and a middling possession team (Boston).

Is the strong correlation between CF% 5v5 and the eventual SC winner from all seasons since 05-06 heavily influenced by the particular playing style of the two teams that have dominated the league since then? We really don't have a diverse number of winning teams to analyze in terms of possession.

That led me to this article on CF% 5v5, PDO, and Goal Differential as a recipe for playoff success:

http://thehockeywriters.com/stanley-cup-recipe/

Don't know if this had been posted here last year when the article was written, but the numbers are interesting. It doesn't seem as though strong CF% teams do much better round to round in the playoffs than the strong PDO and Goal Differential teams, not to mention the significant portion of hockey that isn't played 5v5.

I can't seem to find enhanced Team Stats for playoffs (War on Ice is blocked at my job, so maybe it's there?), but I have looked this up before and recall seeing that the last five SC winners were all Top 5 in PDO in the postseason during their winning years. So while they carried strong CF% 5v5 into the playoffs, they all benefited from hot goaltending and accurate shooting when it mattered. Yes, some of the better PDO teams did not advance to the Finals, but many of them (think Anaheim and NYR) were able to consistently reach the later rounds and force elite Corsi teams to 7 game series in the Conference Finals that were decided by one or two goals (last season).

(One other thought: Unlike some of the other strong possession teams, Chicago and especially LA are among the few that consistently post great CF% 5v5 when leading, not only when close, tied, and trailing. Is it possession, or a trait that is specific to those two teams that makes them so successful?)

Some of you guys have probably looked into this much deeper than I have, so tell me: has the predictive power of CF% 5v5 been overstated? Have two incredibly successful teams colored our view when the answer to "who is likely to succeed in the playoffs?" is probably more complex? Is PDO ultimately king when we consider that SC winners have ridden PDO hot streaks in the playoffs in addition to their sound possession game?

I'm genuinely interested, and these are my main sticking points that make me a bit skeptical of using CF% 5v5 as a guide for who we are likely to see playing in May/June. Where am I right/wrong?

Thanks!

It's not surprising high PDO teams are more successful. In a small sample, like the playoffs, winning a few games by riding the percentages can take you a long way.

I don't think this matters though (in predicting who will be "good" in the playoffs) because there hasn't been much or any research yet that suggests PDO is a truly repeatable or predictable skill. Possession is a repeatable and predictable skill. We can't know who's going to have a high PDO. But we can reasonably guess who the better CF% teams are and past studies indicate that these teams are more likely to win.

So yah, if you're lucky you're going to win more. But you can't just decide to be lucky.

 

[KingDeathMetal]

It's not surprising high PDO teams are more successful. In a small sample, like the playoffs, winning a few games by riding the percentages can take you a long way.

I don't think this matters though (in predicting who will be "good" in the playoffs) because there hasn't been much or any research yet that suggests PDO is a truly repeatable or predictable skill. Possession is a repeatable and predictable skill. We can't know who's going to have a high PDO. But we can reasonably guess who the better CF% teams are and past studies indicate that these teams are more likely to win.

So yah, if you're lucky you're going to win more. But you can't just decide to be lucky.

PDO is repeatable insofar as having forwards with consistently high S% on your top lines, and a goalie like Lundqvist/Rask/Schneider/Price who is usually good for a SP% over .920. But I get what you're saying - it's not something that most teams can plan for.

 

[KingDeathMetal]

Not really. After the penguins fired Therrien, they were one of the best possession team in the league with a 53.6 SA CF%. So essentially, since that time it's five strong possession teams (Anaheim, Detroit, Kings, Penguins, Blackhawks), a very weak one (Carolina @ 48.5 CF% SA) and a middling one (Boston @ 51.5 CF% SA).

As far as your point about PDO is concerned: yes, going on a hot streak / getting lucky at the right time is important, but the team that wins a game in general - playoffs or not - will always win the PDO battle. There's just not enough goals scored in a small chunk of games to make sure a single goal doesn't influence the SA/GA ratio to a huge extent.

I'm pretty sure the 2014-15 ducks were a pretty good Corsi team after the TD. EDIT: 53.7 SA CF%. To be fair, it's a small sample size (19 games).

Great point about CF% late in the season - I hadn't considered that. Though it would still put a dent in looking at final CF% 5v5 standings to draw conclusions, it certainly is a great argument for CF% on its own merits. Thanks for pointing that out.