HFBoards

Go Back   HFBoards > General Hockey Discussion > By The Numbers
Mobile Hockey's Future Become a Sponsor Site Rules Support Forum vBookie Page 2
By The Numbers Hockey Analytics... the Final Frontier. Explore strange new worlds, to seek out new algorithms, to boldly go where no one has gone before.

0.560 is the new 0.500

Reply
 
Thread Tools
Old
02-04-2013, 05:48 PM
  #51
UnrefinedCrude
Registered User
 
UnrefinedCrude's Avatar
 
Join Date: Jun 2011
Country: Canada
Posts: 1,975
vCash: 500
Quote:
Originally Posted by DL44 View Post
So teams don't get a point each as soon as regulation ends for achieving a tie after 60 mins?
But the system was introduced with the purpose of eliminating ties.

There is no column for ties. It is the point a team gets for "losing in overtime" It is a loser point.

if it was a point for tying then the game would end in a tie.

The flaw in the system is because the NHL never implemented a standard for what a game is worth. The league used to value every game at 2 points. When they changed the system they made some games worth more than others.

They did this on purpose to create parity. But I don't like how it affect competitive balance.

If you are going to add extra value to a game in the form of an extra point it should be awarded for winning.

UnrefinedCrude is offline   Reply With Quote
Old
02-04-2013, 06:12 PM
  #52
Doctor No
Retired
 
Doctor No's Avatar
 
Join Date: Sep 2005
Posts: 23,368
vCash: 500
Quote:
Originally Posted by UnrefinedCrude View Post
But the system was introduced with the purpose of eliminating ties.

There is no column for ties. It is the point a team gets for "losing in overtime" It is a loser point.
Actually, the column, as originally stated, was for "regulation ties".

I'm not sure why the NHL site now lists "OL" instead of "RT", but I remember the distinction being important at the time.

Doctor No is offline   Reply With Quote
Old
02-04-2013, 06:23 PM
  #53
MountainHawk
Registered User
 
MountainHawk's Avatar
 
Join Date: Sep 2005
Location: Salem, MA
Country: Vanuatu
Posts: 12,771
vCash: 450
Quote:
Originally Posted by UnrefinedCrude View Post
But the system was introduced with the purpose of eliminating ties.

There is no column for ties. It is the point a team gets for "losing in overtime" It is a loser point.

if it was a point for tying then the game would end in a tie.

The flaw in the system is because the NHL never implemented a standard for what a game is worth. The league used to value every game at 2 points. When they changed the system they made some games worth more than others.

They did this on purpose to create parity. But I don't like how it affect competitive balance.

If you are going to add extra value to a game in the form of an extra point it should be awarded for winning.
It is. You get a point for WINNING the gimmick. You get a point for TYING the hockey game.

MountainHawk is offline   Reply With Quote
Old
02-04-2013, 07:02 PM
  #54
DL44
Registered User
 
DL44's Avatar
 
Join Date: Sep 2006
Location: Left Coast
Posts: 5,477
vCash: 133
Quote:
Originally Posted by UnrefinedCrude View Post
But the system was introduced with the purpose of eliminating ties.

There is no column for ties. It is the point a team gets for "losing in overtime" It is a loser point.

if it was a point for tying then the game would end in a tie.

The flaw in the system is because the NHL never implemented a standard for what a game is worth. The league used to value every game at 2 points. When they changed the system they made some games worth more than others.

They did this on purpose to create parity. But I don't like how it affect competitive balance.

If you are going to add extra value to a game in the form of an extra point it should be awarded for winning.
Consider the fact they change the rules of the game as soon as 60 minutes ends. They add the gimmick of OT by dropping a player and then the gimmick of the SO.


- The teams have proven after 60 minutes of play that they are equals on that particular night... congrats! point for each!
- Now let's change the rules... Now go fight for the extra point!
= bonus point to me.



Can you please explain how the system messes with competitive balance? How so?

The argument people sometimes use - Team with less wins sometimes get in over a team with more wins...
Well if that's the case it's because that team that made it also lost a lot less in regulation that the team that failed.... So that's non-starter as an argument for me..

The system is in place and everyone knows it... No one gets screwed by being surprised by it at the end of the season.
The failing team, at the end of the day, failed to 'not lose' enough.. Or not win enough to make for all the losing they were doing in regulation relative to the team good enough to get it over them.

DL44 is offline   Reply With Quote
Old
02-04-2013, 08:22 PM
  #55
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,556
vCash: 50
Quote:
Originally Posted by DL44 View Post
or was it increased parity?

And the point of the system change was to stimulate more decisions which it did... It wasn't the ties that was issue, it was the actual w/l ratio vs ties. Take a look at the rest of the records and compare pre and post OT structure change. It did it's job.
I also thought parity may be a significant factor, but it does not appear that it was. However, the scoring level is (IOW, higher scoring seasons tend to produce lower % of ties after regulation than lower scoring seasons).

Yes, every game is decided now, but games have different values, and they are frequently decided by a skills contest.

Czech Your Math is offline   Reply With Quote
Old
02-07-2013, 11:47 PM
  #56
DL44
Registered User
 
DL44's Avatar
 
Join Date: Sep 2006
Location: Left Coast
Posts: 5,477
vCash: 133
Update on current season...

East
154gms - 13 extra time. 0.522 pt % is average.. (50 pt pace)

West
150gms - 23 extra time. 0.538 pt % is average (52 pt pace)

League wide:
304 gms, 644 pts. 0.530 pt % is average

Overtime rates are down relative to past seasons... VERY down in the East.

DL44 is offline   Reply With Quote
Old
02-25-2013, 12:47 PM
  #57
DL44
Registered User
 
DL44's Avatar
 
Join Date: Sep 2006
Location: Left Coast
Posts: 5,477
vCash: 133
Quote:
Originally Posted by DL44 View Post
Update on current season...

East
154gms - 13 extra time. 0.522 pt % is average.. (50 pt pace)

West
150gms - 23 extra time. 0.538 pt % is average (52 pt pace)

League wide:
304 gms, 644 pts. 0.530 pt % is average

Overtime rates are down relative to past seasons... VERY down in the East.
Did these numbers wrong.. surprised no one caught it... I forgot to cut the total games in half. So at the time, it should of been:

East
77 gms - 13 extra time. 167 pts handed out. 0.542 pt% - 52 pt pace for the ~mid point (approx playoff spot)

West
75 gms - 23 extra time. 173 pts handed out. 0.577 pt% - 55 pt pace.

So that was back on Feb 7th.



Present update....

East
136 gms - 23 extra pts. 295 pts out. 0.557 pt % - 53 pt pace.

West
132 gms - 39 extra pts. 303 pts out. 0.574 pt % - 55 pt pace.


So a slow regression to the mean.. if we are to consider 560 as the mean.

Overall -
268 gms - 62 xtra pts. 598 pts out = 0.558 as the present average.

(preseason prediction for playoff threshold: 54 pts.. )

DL44 is offline   Reply With Quote
Old
02-25-2013, 11:32 PM
  #58
danishh
Dat Stache
 
danishh's Avatar
 
Join Date: Dec 2006
Location: mtl/ott/somewhere
Country: Canada
Posts: 29,654
vCash: 50
i'm interested to see if this works at accurately predicting the playoff threshold. Would be very interesting information down the stretch in future years if you could guess it at the trade deadline.

have you thought about going back looking at past years and coming up with a corrective constant to do this?

__________________
RIP Kev.
danishh is offline   Reply With Quote
Old
02-26-2013, 01:09 AM
  #59
DL44
Registered User
 
DL44's Avatar
 
Join Date: Sep 2006
Location: Left Coast
Posts: 5,477
vCash: 133
Quote:
Originally Posted by danishh View Post
i'm interested to see if this works at accurately predicting the playoff threshold. Would be very interesting information down the stretch in future years if you could guess it at the trade deadline.

have you thought about going back looking at past years and coming up with a corrective constant to do this?
I believe i looked into that a yr or 2 ago and found that within 1 place of the median (15-16 being the median) was the 92 pt team (0.560)... i'm fuzzy on it tho.

But historically there are more absolutely horrible teams ina given yr than absolutely dominating ones which ends up boosting the median to be slightly higher than the mean... but not by much.

generally 0.560 is a good prediction as the minimum to be in the playoff hunt in any given yr.

But absolutely speaking.. i don;t recall for sure, and it's late.. maybe i will check tomorrow.

DL44 is offline   Reply With Quote
Old
03-15-2013, 10:04 AM
  #60
SealsFan
Registered User
 
SealsFan's Avatar
 
Join Date: May 2009
Location: Long Island, NY
Posts: 819
vCash: 500
Quote:
Originally Posted by Ogopogo* View Post
Loser points are garbage. Point percentage is stupid.

I consider all OTL as losses, not these stupid loser point tie things that the NHL considers them.

You win or you lose. You cannot have one team win while the other team gets a tie - that is stupidity.

Winning % is the same it always was - just use L and OTL as losses and you have it.
One of the reasons I don't follow modern hockey. The standings are maddening to try to figure out.

SealsFan is offline   Reply With Quote
Old
03-15-2013, 12:05 PM
  #61
glovesave42
Registered User
 
Join Date: May 2010
Posts: 53
vCash: 500
Quote:
Originally Posted by DL44 View Post
I put this on the NHL Main board when it should of gone in here...

Due to the bonus pt that's been handed out with a lot more regularity since the lockout... 0.500 is no longer the pure league "average" which it has traditionally been.

Everyone knows this... but what exactly IS the present era equivalent for 'AVERAGE'?


Here are the numbers.

05-06 - 2741 pts - 91.37 pts/team - 0.557
06-07 - 2741 pts - 91.37 pts/team - 0.557
07-08 - 2732 pts - 91.07 pts/team - 0.555
08-09 - 2742 pts - 91.40 pts/team - 0.557
09-10 - 2761 pts - 92.03 pts/team - 0.561
10-11 - 2757 pts - 91.90 pts/team - 0.560
11-12 - 2760 pts - 92.00 pts/team - 0.561

Overall in the post lockout era = 91.56 pts/team = 0.558
Over the last 3 yrs = 91.98 pts/team = 0.561

So basically a 92 pt pace

A 0.500 peforming team is now a below par performing team in this era...

So for simplicity's sake... consider a 92 pt pace as average.. or 0.560 as average (basically the whole era with a little more weight to the recent 3 yrs yrs of the league..)


What does this mean for the 48 game schedule...

If OT/SO rates remain consistent - which you can see they have been AMAZINGLY consistent over the last 7 seasons - the league average in pts this yr will be around -54 pts.

That should also be the approximate target pt total to be the 8th/9th seed in either conference.


Presently, in the league..
West - 43 gms with 12 gms going into extra time
East - 44 gms with 8 gms going into extra time
Total : 87 games, 20 gms in extra time = 194 pts handed out = 2.23 pts/gm = 0.557 as the present 'average'.
The 'average' has traditionally increased during the year teams and game tighten up, so should end up near that 0.560 pt as per usual.

So there ya go... Consider 0.560 as league average for reference purposes...

Them the facts... use with enthusiasm and care.

EDIT:
We're talking Point Percentage here
I used a different approach a couple years back and came up with similar numbers. I looked at goal differential and how it affected points per game and found that .500 is roughly 1.1 points per game, which equates to 2.2 total points per game. So, for a 82 game season, a .500 record would be 90 points. For a 48 game season using my numbers, 53 points would be .500.

It's interesting that the rate of overtime games is so much lower this year. That would definitely skew the numbers down (as there are fewer points out there). I'm wondering if that is an effect of the lockout or just a reflection of a relatively small sample size.

glovesave42 is offline   Reply With Quote
Old
03-15-2013, 03:36 PM
  #62
trobby
Registered User
 
trobby's Avatar
 
Join Date: Aug 2007
Location: Ottawa
Country: Canada
Posts: 2,248
vCash: 500
Regulation Win: 3 pts
OT Win: 2 pts
OT Loss: 1 pt
Tie: 1 pt each
No Shootout.

Having 3 pts awarded for a Regulation win would put more emphasis on winning before getting to OT. You'll see more teams trying near the end instead of playing for a point.

If they get to OT, you'll try harder to win because you've already lost a point by not winning in regulation.

The only reason I would award 1 point for getting to OT is because the game dynamic changes when you go from 5-on-5 to 4-on-4.

Tie-breakers in the standings would be Regulation Wins 1st, then head-to-head..

Let's make this official NHL.


Last edited by trobby: 03-15-2013 at 03:43 PM.
trobby is offline   Reply With Quote
Old
03-15-2013, 10:39 PM
  #63
KJF
Registered User
 
Join Date: Nov 2012
Posts: 29
vCash: 500
I think a just as effective, but perhaps simpler to understand calculation is to use points/game in ranking the teams. I think this should be used by NHL and media in ranking the teams during the season, as it cancels out the differences in numbers of games played. For example, as of today in the Eastern Conference, the Bruins would be in 1st (39 points in 25 games is 1.560 points/game), and the Jets would be ahead of the Leafs in 7th place at 1.111 points/game vs 1.107. (As a Jets fan, I am quite pleased about that.)

Oh yes, and the playoff cut-off would typically be ~1.12 points/game.

KJF is offline   Reply With Quote
Old
03-16-2013, 01:07 PM
  #64
Duffman955
Registered User
 
Duffman955's Avatar
 
Join Date: Mar 2010
Country: Canada
Posts: 8,724
vCash: 500
Take out 3 point games, it is stupid. IF you lose in OT, you get nothing. If you are still tied after OT, both teams get point

Duffman955 is offline   Reply With Quote
Old
03-16-2013, 02:05 PM
  #65
DL44
Registered User
 
DL44's Avatar
 
Join Date: Sep 2006
Location: Left Coast
Posts: 5,477
vCash: 133
Quote:
Originally Posted by Duffman955 View Post
Take out 3 point games, it is stupid. IF you lose in OT, you get nothing. If you are still tied after OT, both teams get point
Stupid and dumb aaand a smelly face too.


I swear, they should change the standings to the following 4 columns just to confuse and shut up the 'loser point' crowd..

Reg Wins (2pts) - Tie after regulation (1pt) - OT/SO wins (1pt) - All Losses (0pts)

Van at 13-7-6, 32 pts would become:
9-10-4-7, 32 pts

You eliminate the perception that teams are rewarded with pts for losing.
The 'Loser point' crowd would be so confused without an OTL column to moan about.
The fact they would have to do math to figure out the number of 'loser points' should eliminate 99% of them from posting about it.


Last edited by DL44: 03-16-2013 at 02:13 PM.
DL44 is offline   Reply With Quote
Old
03-17-2013, 06:27 PM
  #66
SealsFan
Registered User
 
SealsFan's Avatar
 
Join Date: May 2009
Location: Long Island, NY
Posts: 819
vCash: 500
How about we add half a point for an arm wrestle or a tiddly-wink flip after the shootout?

SealsFan is offline   Reply With Quote
Old
03-18-2013, 11:39 AM
  #67
Beerfish
Registered User
 
Beerfish's Avatar
 
Join Date: Apr 2007
Posts: 10,854
vCash: 500
Don't tell my fellow oiler fans this. A number of them like to trot out the old 'We are almost a .500 team thus things are going well!' Statement on a regular basis.

Beerfish is online now   Reply With Quote
Old
03-23-2013, 01:15 PM
  #68
Windy River
Registered User
 
Join Date: Jan 2013
Posts: 375
vCash: 500
Quote:
Originally Posted by leeaf83 View Post
well to be fair, there's always been more than 2 results (win, loss, or tie)

but the current system is inherently flawed for 2 main reasons;
-some games having 2 points awarded with others having 3
-the possibility of a game having a winner and NO loser.
There may have traditionally been more than 2 results, however that has changed. We have a very effective system in place so that all games have a clearly defined winner and a loser. The extra 1 point is basically "hangover" from the previous system. It doesnt make any sense to award it in my opinion. Going back to a simple W-L makes sense, and use the OTL as a background stat in the event of a tiebreaker come playoff seeding time. Then winning percent will mean something.

Windy River is offline   Reply With Quote
Old
03-23-2013, 01:29 PM
  #69
Windy River
Registered User
 
Join Date: Jan 2013
Posts: 375
vCash: 500
Quote:
Originally Posted by DL44 View Post
So teams don't get a point each as soon as regulation ends for achieving a tie after 60 mins?
They do but why should they? They are really just a few minutes away from awarding a winner and a loser so why get all point happy and give out points when there are players still on the ice trying to decide the game?

Windy River is offline   Reply With Quote
Old
03-24-2013, 02:02 AM
  #70
DL44
Registered User
 
DL44's Avatar
 
Join Date: Sep 2006
Location: Left Coast
Posts: 5,477
vCash: 133
Quote:
Originally Posted by Windy River View Post
They do but why should they? They are really just a few minutes away from awarding a winner and a loser so why get all point happy and give out points when there are players still on the ice trying to decide the game?
Because they change the rules of the game after 60 mins are played.

DL44 is offline   Reply With Quote
Old
03-24-2013, 03:17 AM
  #71
CpatainCanuck
Registered User
 
CpatainCanuck's Avatar
 
Join Date: Sep 2008
Posts: 3,704
vCash: 500
"New" as in only 7 years old.

CpatainCanuck is offline   Reply With Quote
Old
03-24-2013, 11:11 AM
  #72
Windy River
Registered User
 
Join Date: Jan 2013
Posts: 375
vCash: 500
Quote:
Originally Posted by DL44 View Post
Because they change the rules of the game after 60 mins are played.
Well I wouldn't call it changing the rules, they are the rules that are already in place. When a team takes a couple quick penalties and we have 5 on 3 play, it's not "changing the rules" even though the on ice play has changed dramatically.

Regardless, what does that have to do with awarding a point? I mean the game is still tied, but there are still a few additional minutes of play specifically for this situation, that WILL result in a winner and a loser.

I just don't understand the sort of 'box' thinking where you must award a point after 60 minutes. I think if we were adopt a simple W-L system, it would be very short time before people stopped mourning the loss of the extra point.

Windy River is offline   Reply With Quote
Old
03-26-2013, 09:18 AM
  #73
MNNumbers
Registered User
 
Join Date: Nov 2011
Posts: 1,218
vCash: 500
Can I get in this discussion, too? I have been intrigued b the 3/2/1/0 system as well, because as a math guy, I hate it that some games give 2 pts in the standings and some 3. I have found myself this year hoping that "ANYBODY" scores near the end of the game so that both teams don't get points, because that is an advantage to the Wild. Both teams (Detroit and Anaheim, for example) aren't gaining points.

As far as the current system, I look at the standings and think Wins-Losses. So, right now, the Wild are +9, Vancouver is +8. In a full season, something like +8 to +12 has normally been the playoff cutoff. This is a 60% season, so probably +5 - +8 will be the cutoff (this matches well with the OP - he says 54 - which is +6).

I always do a 3/2/1/0 standings for my own interest, as well. For these, I count 3 for a Reg Win. I don't have the data for the East right now, but for the West, there are differences between this and the actual standings:

NHL
In Order: CHI, ANA, MINN, VAN, DET, LA, STL, SJ where Minn has earned the 3rd spot by points, as well as leading the NW Division.

3/2/1/0 system (by pts/gm):
In Order: CHI, ANA, MINN, LA, VAN, DET, STL, DAL, where again Minn has earned the 3rd spot by points, too.

So, the differences are:
In 3/2/1/0, LA jumps from 6th to 4th
DAL takes SJ's place in the playoffs. It should be noted that the change for Dallas amounts to 3 pts in the standings right now, or 8 pts in a full year. That's alot!!!

So, my take basically is that the 3/2/1/0 is more mathematically pure, but doesn't affect the standings much (A little can mean a lot, though. LA was an 8th seed last year, remember.)

My real argument is that the shootout is a random event. So, because random events tend to fall to a bell curve, then some 'good' team is going to have bad shootout results, and miss the playoffs over the 'luck' of the shootout.


Last edited by MNNumbers: 03-26-2013 at 10:22 AM.
MNNumbers is online now   Reply With Quote
Old
03-26-2013, 05:53 PM
  #74
DL44
Registered User
 
DL44's Avatar
 
Join Date: Sep 2006
Location: Left Coast
Posts: 5,477
vCash: 133
Quote:
Originally Posted by MNNumbers View Post
Can I get in this discussion, too? I have been intrigued b the 3/2/1/0 system as well, because as a math guy, I hate it that some games give 2 pts in the standings and some 3. I have found myself this year hoping that "ANYBODY" scores near the end of the game so that both teams don't get points, because that is an advantage to the Wild. Both teams (Detroit and Anaheim, for example) aren't gaining points.

As far as the current system, I look at the standings and think Wins-Losses. So, right now, the Wild are +9, Vancouver is +8. In a full season, something like +8 to +12 has normally been the playoff cutoff. This is a 60% season, so probably +5 - +8 will be the cutoff (this matches well with the OP - he says 54 - which is +6).


You've nailed it right on the head with that bolded part.
It's about how many wins you can accumulate vs regulation losses... That's the only thing that matters at the end of the day...

i.e. losing in the gimmick parts of the game (OT & SO) is negated, while winning matters.

If a team OTL's, it's basically burning a game that does nothing for them in the standings... well.. they avoided a reg loss - which is the good thing.

What the standings are all about is Minimizing your Regulation Losses, and maximizing Wins.

It's the difference between Wins and Reg Losses that's important.

I was gonna make a thread about it, but it would of just spurred a semantics conversation.

When i run through the standings, i too look at the wins vs reg losses... basically pay less attention to pts.

(as mentioned by MN above..)
As it pertains to this thread over the course of an 82 gm season, +10 is league average = 92 pts. (0.560) (also approximate playoff bar).

This season... if it was on track as 0.560 being average over 48 games, that would be an average of/playoff bar of around +6. (playoff bar seems actually seems a little lower right now due to the the number of dominate teams vs the lack of pure scrub teams)


(Wins - reg Losses = pts your team is over total games played (or 0.500))

Edit:
Standings in the West as i look at them..
1. Chi +20
2. Ana +16
3. Min +9
4. Van +8
5. Det +6
6. LAK +6
7. StL +5
8. SJS +3
--
9. Nas +2
10. Dal +1
11. CBJ 0
12. Pho -2
13. Edm -2
14. Cal -2
15. Col -5


Last edited by DL44: 03-26-2013 at 06:20 PM.
DL44 is offline   Reply With Quote
Old
03-29-2013, 11:01 PM
  #75
Ari91
Registered User
 
Join Date: Nov 2010
Location: Toronto
Country: Canada
Posts: 9,451
vCash: 500
While it's a flawed system to have regulation games worth 2 points and overtime games worth 3, I'm not cool with eliminating the loser point unless they also eliminate the shootout. I don't think any team should lose out all the points available based on a skills competition.

Ari91 is offline   Reply With Quote
Reply

Forum Jump


Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 09:18 AM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"
Contact Us - HFBoards - Archive - Privacy Statement - Terms of Use - Advertise - Top - AdChoices

vBulletin Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. 2014 All Rights Reserved.