2013-14 NHL thread

[ChicosBailBonds]

I'm not talking about the whole game-of-inches thing. Every game is a game of inches. An NFL team makes a huge first down by an inch. A hockey player hits the crossbar with a shot. An NBA player has a shot bounce off the rim 3 times before going in. A line drive barely hits the outside of the chalk, giving the batter a walk-off double. We all get that.

I'm talking about a conference-finals-winning play that deflects off one player's stick, deflects off another player's shoulder and somehow eludes the huge, padded goalie to find the net.

And the crazy thing is, that happens in hockey 2, 5, 10 times -- or more -- in a 7-game series.

There is no game with a bigger luck factor (or bad-luck factor for the Blackhawks last night) than hockey.

I would agree, that luck is a big part of it, but in a 7 game series that tends to even out.  There's a reason the #1 team in that rating system seems to win the Cup or play for it each year.  It's not a metric measuring luck, it's a metric measuring puck possession. shot differential and several other factors. 

I get your point on the winning goal, but it's just one play of hundreds each game.  Crawford doesn't give up a rebound, the tying goal isn't scored.  A chip in the ice doesn't make a shot take an odd angle, maybe Quick stops it.  Lots of that stuff happens, as you point out.  But someone had to take that shot, someone had to tempt the luck and put it in play.  You don't shoot that 50 footer, the luck doesn't happen.  As such, you make some of your luck.

[ChicosBailBonds]

Lenny, I get the math just fine.

Kings, lost their first game.  In the NCAA format, they would be done.  One loss...done.  Instead, they got 3 do overs, you get none in the NCAA.

Biggest crapshoot in all of sports, period.  One bad game, you're done.  In the NHL, NBA, MLB, you can have multiple bad games, mulitple do overs.  The math is not hard.

[Lennys Tap]

OH no, Vegas?  LOL.   For the millionth time, the odds are to move a betting line....PERIOD.

Oh yes, Vegas. The place where guys like you arrive in $60,000 cars and go home on $800,000 buses. Not a great deal of action on hockey but I sincerely doubt if any of the big players are so unsophisticated that they're not familiar with your adjusted Fenwick ratings that you think are so foolproof. If they were foolproof Vegas would be busted by now, but I haven't heard of any casinos closing lately. You must have busted a few yourself. Which ones?

[g0lden3agle]

You don't understand math. If a game is 50/50, a coin flip, flipping that coin 7, or 9 or 11 or any number of times doesn't change the odds at all. Best of 7000 yields the exact same odds as one flip. In games where their is a prohibitive favorite a longer series does help the better team. But when a team that finishes 25th of 30 in the NHL but is supposedly better than your 2nd place Ducks by 3 places the whole league must be a crapshoot/coin flip.

You just said yourself that the longer series helps a superior team's odds of advancing to the next round.  Flipping that around, wouldn't the NCAA one-and-done format be the worst odds for the best teams to advance each round, thus making it more of a crapshoot?

[ChicosBailBonds]

You just said yourself that the longer series helps a superior team's odds of advancing to the next round.  Flipping that around, wouldn't the NCAA one-and-done format be the worst odds for the best teams to advance each round, thus making it more of a crapshoot?

Don't confuse him or call out his own admissions of sinking himself....it's more subtle when he thinks he has it right and nearly everyone knows he doesn't

[ChicosBailBonds]

Oh yes, Vegas. The place where guys like you arrive in $60,000 cars and go home on $800,000 buses. Not a great deal of action on hockey but I sincerely doubt if any of the big players are so unsophisticated that they're not familiar with your adjusted Fenwick ratings that you think are so foolproof. If they were foolproof Vegas would be busted by now, but I haven't heard of any casinos closing lately. You must have busted a few yourself. Which ones?

I've given you the last 6 years, and 5 times the team with the best Fenwick has gone at the very least to the Stanley Cup Finals.  The data is the data.  You didn't even know about Fenwick until a week ago, I suspect there are many others out there don't either.  I'm glad I could help educate you a bit.

There are outliers in anything, that's the nature of sports and statistics.  Nevertheless, to date, SAF has been a fantastic predictor of success in the playoffs the last 6 years. 

Some people swear by the Team Corsi ratings...LA was ranked 3rd this year in Corsi 5 on 5. 

[Lennys Tap]

You just said yourself that the longer series helps a superior team's odds of advancing to the next round.  Flipping that around, wouldn't the NCAA one-and-done format be the worst odds for the best teams to advance each round, thus making it more of a crapshoot?

The longer the odds the more favorites are helped by a series of games. When Kentucky plays Boston U. the Terriers have a much better chance to win one game than they do to win 4 of 7. And when you play Powerball you have a much better chance to win that day than you do winning 4 of 7 days. So, is the NCAA tournament more random as a one and done than IT would be as a best of 7? Undoubtedly. But to determine whether it is more random the other sport's playoffs (best of 7, one game, whatever) it is only ONE of several factors to consider.

The two biggest factors (there are others I won't bore you with) that make the NHL playoffs much more random than college basketball are a) the parity between the teams and b) the amount that pure luck or chance affects outcomes. Consider:16 of 30 (54%) of NHL teams make the playoffs, 68 of 366 (19.6%) make the NCAAs. In the NHL, it's relatively common for a team that earned a #1 over an 82 game schedule to lose in their opening series. In the NCAAs a team that earned a #1 seed over a 35 game season has never lost. This year's Final Four, in what was probably the most wide open tournament ever, consisted of a 1,2,8 and 9, a grand total of 20. This year's NHL last 4 standing were a 7,9,10 and 12, totaling 38. Expected total if pulled randomly from a hat? 34 for both. So maybe the biggest crapshoot in NCAA history was roughly half as big a random event as this year's Stanley Cup. Not surprising that the greater parity produces more random (crapshoot) results, since the closer the games are to coin flips the less meaningful a one and done feature affects randomness (the odds of winning 4 of 7 coin flips are exactly the same as winning one flip - 50%).

On to luck. For Chico, the 35 game regular season in college basketball is the determinate of who is the best team. In hockey, the 82 game regular season means nothing, and the real best teams are determined by something called the Adjusted Fenwick something or another. It rates teams on percentage of time they possess the puck (among other things) and attempts to rank them after the "luck factor" is subtracted. It has been a more accurate predictor than seeding for NHL playoff series, but it's sample size is very small (only been around since 2008). But the most interesting thing about Fenwick isn't that its been good at picking playoff winners over a short period - it's Fenwick's conclusion that luck plays an ENORMOUS role in hockey and that who wins and loses has very little to do with who's better and very much to do with who's luckier. If they're as smart as Chico says, 9 teams who didn't make the playoffs (including Florida, who finished 29th with 66 points) were better than Colorado, who finished 3rd with 112 points. But they were unluckier. And New Jersey and Vancouver, who finished well out of the playoff picture, were better than #2 Anaheim, just unluckier. Imagine Jeff Sagarin asserting, for example, that Penn State was better than Wisconsin last year, just unluckier. We'd all have a good laugh. But this is Chico"s guy, and if he's right, every hockey game and series is infinitely more of a crapshoot than any basketball tournament could ever be.

So, in conclusion: the one and done format DOES make the NCAAs more of a crapshoot than it would be if teams played best of 7 series, but that is more than mitigated by, among other things, the parity and luck factors present in the NHL. Thanks to Chico's guru Mr Fenwick for helping make the case beyond a doubt.  
 

[ChicosBailBonds]

Good try....A for length, C for logic, but I'll throw a B in there for attitude.


Where did I say the regular season for hockey doesn't matter?  I didn't. What I said was, just because you have the best record which is how they seed in the NHL, does not mean you are the best team.  You can pile up wins against a bad division, etc.  Schedules aren't balanced, etc.  Afterall in this very diatribe of yours you are arguing that the better team "UNDOUBTEDLY" will have a better chance of winning in a longer series, and YOU have said the best teams are by records.  Yet, the President's Cup winner has won the Cup how many times in the last 10 years?  Twice.

Why?  Well, now you're saying it is because it is a crapshoot.  LOL.   When the more logical answer is that the President's Cup winner isn't the best team, especially when in 7 game series they "UNDOUBTEDLY" have ample opportunities to prove this.

Your failure in logic on this is assigning best team with best record, when you know full well that the schedules aren't balanced, injuries happen during the season, etc, etc.  This is why a COMMITTEE that seeds teams knowing this information typically does a better job than just taking the best record, because they know who you played, know about key injuries, etc.  There is a reason why more advanced rating systems using deep analytics rate teams differently than just what their record is.  It is also why they prove to be much more accurate.

Finally, I'd recommend a statistical communique to Professor Jay Bergen, mathematician at DePaul.  Why don't you let him educate you a bit on the crapshoot of winning the NCAA tournament vs winning the Stanley Cup.  Would you like an introduction?

[swoopem]

Chicos it's the President's Trophy not cup

[Lennys Tap]

Good try....A for length, C for logic, but I'll throw a B in there for attitude.


Where did I say the regular season for hockey doesn't matter?  I didn't. What I said was, just because you have the best record which is how they seed in the NHL, does not mean you are the best team.  You can pile up wins against a bad division, etc.  Schedules aren't balanced, etc.  Afterall in this very diatribe of yours you are arguing that the better team "UNDOUBTEDLY" will have a better chance of winning in a longer series, and YOU have said the best teams are by records.  Yet, the President's Cup winner has won the Cup how many times in the last 10 years?  Twice.

Why?  Well, now you're saying it is because it is a crapshoot.  LOL.   When the more logical answer is that the President's Cup winner isn't the best team, especially when in 7 game series they "UNDOUBTEDLY" have ample opportunities to prove this.

Your failure in logic on this is assigning best team with best record, when you know full well that the schedules aren't balanced, injuries happen during the season, etc, etc.  This is why a COMMITTEE that seeds teams knowing this information typically does a better job than just taking the best record, because they know who you played, know about key injuries, etc.  There is a reason why more advanced rating systems using deep analytics rate teams differently than just what their record is.  It is also why they prove to be much more accurate.

Finally, I'd recommend a statistical communique to Professor Jay Bergen, mathematician at DePaul.  Why don't you let him educate you a bit on the crapshoot of winning the NCAA tournament vs winning the Stanley Cup.  Would you like an introduction?

I give you an A for Avoidance - the rating system you love concludes that luck plays a vastly greater role in who wins a hockey game than a basketball game. Vastly. The more games are determined by luck, the more random. The more random, the bigger crapshoot.

I also give you an A in your favorite subject - silly, meaningless name dropping.

F in math, Incomplete, trending F in logic.

[ChicosBailBonds]

Chicos it's the President's Trophy not cup

Thank you...you get the idea.  The team with the best record doesn't mean they are the best team in a world of uneven schedules, injuries, etc.  Yet, this is the logic my pal uses to justify the crapshoot because he associates the seeding, which is based on record, to make his argument.  Flawed from the start.

[Lennys Tap]

Thank you...you get the idea.  The team with the best record doesn't mean they are the best team in a world of uneven schedules, injuries, etc.  Yet, this is the logic my pal uses to justify the crapshoot because he associates the seeding, which is based on record, to make his argument.  Flawed from the start.

Let's sum up:

1. If we use traditional tools to judge randomness (regular season, seedings, etc.), you lose.

2. If we use "new metrics" like Fenwick to actually analyze, we find the reason that you lose - hockey is such a random, luck based game that sometimes the better teams lose much more regularly than the worst teams do - to the point where the team with the 29th best record over an 82 game schedule can be better than the team with the 3rd best record over that same period. And there are several other anomalies nearly as outrageous. When you try to explain these literally upside down conclusions of Fenwick by citing injuries and unbalanced schedules it's a pathetic reach of a drowning man.

It's true that the one and done feature adds some randomness to the NCAAs and the best of 7 format takes a little (very little) volatility from the Stanley Cup. I've not denied that. But you're so hung up by that one tree you missed the whole forest.

[MU82]

Isn't the NFL also one-and-done?

Doesn't college basketball also have unbalanced schedules?

Isn't this argument as silly as it is timeless?

[g0lden3agle]

Let's sum up:

1. If we use traditional tools to judge randomness (regular season, seedings, etc.), you lose.

2. If we use "new metrics" like Fenwick to actually analyze, we find the reason that you lose - hockey is such a random, luck based game that sometimes the better teams lose much more regularly than the worst teams do - to the point where the team with the 29th best record over an 82 game schedule can be better than the team with the 3rd best record over that same period. And there are several other anomalies nearly as outrageous. When you try to explain these literally upside down conclusions of Fenwick by citing injuries and unbalanced schedules it's a pathetic reach of a drowning man.

It's true that the one and done feature adds some randomness to the NCAAs and the best of 7 format takes a little (very little) volatility from the Stanley Cup. I've not denied that. But you're so hung up by that one tree you missed the whole forest.

Just so we set the record straight, your most recent use of examples of seeding to prove your point were not very right.  Those 8 and 9 seeds in the final 4 this last year were 29-32 seeds and 33-36 seeds.  You used the conference independent "seed" for the NHL but then the region specific seed of the NCAA teams.

[Lennys Tap]

Just so we set the record straight, your most recent use of examples of seeding to prove your point were not very right.  Those 8 and 9 seeds in the final 4 this last year were 29-32 seeds and 33-36 seeds.  You used the conference independent "seed" for the NHL but then the region specific seed of the NCAA teams.

16 teams in the NHL playoffs, 64 teams (after the play in games) in the NCAAs. If you want to use their total seeds rather than dividing by 4 (and using their regional seeds, fine, but it doesn't change anything statistically. Florida was 1, UW 6, UK 29 and UCONN 35. Expectations for the sum of the FF seeds (if random) would then be 130 (30+31+34+35) versus the actual much lower total of 71 (1+6+29+35). For the NHL the opposite was true. The random expectation (7+8+9+10) of 34 was exceeded by the actual number (7+9+10+12) of 38. So,no contest which was more random any way you slice it.

[ChicosBailBonds]

Lenny, I'll take a picture for you at Saturday's game with a Wish You Were Here sign.

[swoopem]

Lenny, I'll take a picture for you at Saturday's game with a Wish You Were Here sign.

My favorite Pink Floyd song if I had to choose. However, I'd be comfortably numb after the decision

[Lennys Tap]

Lenny, I'll take a picture for you at Saturday's game with a Wish You Were Here sign.

Thanks, but I couldn't care less about going to a Kings/Rangers game. The next time Hoop comes over for beers, though, give me a holler - that's something I'd pay to see.

[ChicosBailBonds]

Isn't the NFL also one-and-done?

Doesn't college basketball also have unbalanced schedules?

Isn't this argument as silly as it is timeless?

Yes, but the NFL plays playoff games in home stadiums, the NCAA does not.

Yes, the NFL has unbalanced schedules, but not nearly as much as other leagues because there are so few games.

You'll find that in all these discussions, the NFL is rarely used because, as you point out it is one and done and you play a bad game and you are gone.  Any one and done scenario is ripe for that surprise.

[ChicosBailBonds]

16 teams in the NHL playoffs, 64 teams (after the play in games) in the NCAAs. If you want to use their total seeds rather than dividing by 4 (and using their regional seeds, fine, but it doesn't change anything statistically. Florida was 1, UW 6, UK 29 and UCONN 35. Expectations for the sum of the FF seeds (if random) would then be 130 (30+31+34+35) versus the actual much lower total of 71 (1+6+29+35). For the NHL the opposite was true. The random expectation (7+8+9+10) of 34 was exceeded by the actual number (7+9+10+12) of 38. So,no contest which was more random any way you slice it.

Have you talked to the good PhD in mathematics at DePaul yet?  I think you would learn something and can change your grading system on logic, etc.

Would you like an introduction online?

[ChicosBailBonds]

Let's sum up:

1. If we use traditional tools to judge randomness (regular season, seedings, etc.), you lose.

2. If we use "new metrics" like Fenwick to actually analyze, we find the reason that you lose - hockey is such a random, luck based game that sometimes the better teams lose much more regularly than the worst teams do - to the point where the team with the 29th best record over an 82 game schedule can be better than the team with the 3rd best record over that same period. And there are several other anomalies nearly as outrageous. When you try to explain these literally upside down conclusions of Fenwick by citing injuries and unbalanced schedules it's a pathetic reach of a drowning man.

It's true that the one and done feature adds some randomness to the NCAAs and the best of 7 format takes a little (very little) volatility from the Stanley Cup. I've not denied that. But you're so hung up by that one tree you missed the whole forest.

No Lenny, because your logic is already flawed in your randomness layout.  Look again what you opined...think about it for a few minutes.  If it doesn't come to you, walk away, come back in an hour and look again.

I meant it about the introduction, he's done some great stuff on sports and game theory.  You might be quite enthralled about what he has to say on your theory.

[Lennys Tap]

Have you talked to the good PhD in mathematics at DePaul yet?  I think you would learn something and can change your grading system on logic, etc.

Would you like an introduction online?

I present facts that you can't refute (saying someone is wrong because you say so is not a refutation) and you name drop. Par for the course.

[g0lden3agle]

Have you talked to the good PhD in mathematics at DePaul yet?  I think you would learn something and can change your grading system on logic, etc.

Would you like an introduction online?

Can you give the PhD in mathematics at DePaul Hoopaloop's login credentials so he can just post his findings here for all to read?  Why have him talk to Lenny alone when he can enlighten us all?

[Lennys Tap]

Can you give the PhD in mathematics at DePaul Hoopaloop's login credentials so he can just post his findings here for all to read?  Why have him talk to Lenny alone when he can enlighten us all?

Exactly, eagle. I've offered facts to back my assertions. Chico namedrops. He doesn't refute my numbers on the seeds or his very own Fenwick metrics. He can't explain why he thinks I'm illogical, he just FEELS that way. That doesn't strike me as very logical or scientific.

[ChicosBailBonds]

Thanks, but I couldn't care less about going to a Kings/Rangers game. The next time Hoop comes over for beers, though, give me a holler - that's something I'd pay to see.

Weird, the other day you seemed to care so much about the Kings to comment on my wife and kids following them.