elliott70 wrote:I think pucksense is pointing out something that is obvious.
Prior Lake is better than LS.
Rankings are all about which team is better.
The best indicator is when one team beats another team.
Especially in the recent past.
Your equation needs to be tweaked to be considered a good ranking system. Right not it is just a calculation.
As both Lee and Mitch point out on their systems, the early season calcualtions are not very good. Your system has no District 12 or 16 teams 'ranked' anywhere. Because they have no numbers. But East Grand Forks, Roseau, Bemidji, Warroad, Grand Rapids may very well have a top 20 to 25 team.
I look at rankings to tell me who will win in this weekends match-ups. History helps develop that, but historical numbers do not tell everything in sports.
But none the less, nice job.
I'll disagree with you somewhat in that a ranking system is not necessarily built to predict the future or who will win this weekends games. Further, ONE singular head to head match up is not, atleast not to me, an indicator that one team is definitively better than the other team. I'd say a minimum of 3 games, (one home, one away and one neutral) would be needed before I could say with any certainty one team is better than the other and even then if they were to both win at home and tie on the neutral rink it'd still be up for debate. The beauty of mathematical ranking systems is that they are based purely on what has happened NOT what MIGHT happen in the future AND they also seem to get better with each passing week given more games to enter data with. Much like the RPI in basketball some mathematical ranking systems are practically worthless until almost 75% of the season is over but that does not mean those systems aren't still going to give you their current rankings and data even in the early season. You just need to know what type of system youa re dealing with and then just acknowledge and know that some systems are better indicators or true rank later in the season as opposed to earlier. This system seems liek one that adjusts itself accordingly the further into the season you get and with more data the more accurate it becomes. I'd bet it's spot on by the end of the season but probably ddoes have early season hiccups every year.
Nothing is perfect not even the so called head to head matchup test, or even common opponent test. I mean what about when Team A beats Team B three times by scores of 6-4, 10-6 and 7-4, Team A loses to Team C three times by scores of 2-1, 4-3, and 3-2, and Team B beats Team C by scores of 7-0, 5-1, and 6-3. Tell me, which team is the "Best team" in this scenario??? (FYI, this is a real thing that happened last year down in our area).
JSR drop your "what about" scenario -- the actual facts are PLS defeated LVS 3-2 in the Eden Prairie Tournament (Fact) and when they play a common opponent in the tournament PLS defeated Rochester 8-0 and LVS tied Rochester 1-1 (FACT), and PLS won the tournament (FACT). Also, regarding the two teams record PLS is undefeated with no ties and LSV has 2 losses and 1 ties, hence with these FACTS PLS should obviously be ranked #1 and LVS #2, until one of them falters or LVS defeats PLS. What I am trying to is have a good mathemetical ranking system be improved upon. So if Monsterbuck1 is the brains behind the ranking system -- I challenge him to improve upon it and not become complacent with the current system used by MyHockeyRankings.com.
Does any of this make sense?
elliott70 wrote:I think pucksense is pointing out something that is obvious.
Prior Lake is better than LS.
Rankings are all about which team is better.
The best indicator is when one team beats another team.
Especially in the recent past.
Your equation needs to be tweaked to be considered a good ranking system. Right not it is just a calculation.
As both Lee and Mitch point out on their systems, the early season calcualtions are not very good. Your system has no District 12 or 16 teams 'ranked' anywhere. Because they have no numbers. But East Grand Forks, Roseau, Bemidji, Warroad, Grand Rapids may very well have a top 20 to 25 team.
I look at rankings to tell me who will win in this weekends match-ups. History helps develop that, but historical numbers do not tell everything in sports.
But none the less, nice job.
I'll disagree with you somewhat in that a ranking system is not necessarily built to predict the future or who will win this weekends games. Further, ONE singular head to head match up is not, atleast not to me, an indicator that one team is definitively better than the other team. I'd say a minimum of 3 games, (one home, one away and one neutral) would be needed before I could say with any certainty one team is better than the other and even then if they were to both win at home and tie on the neutral rink it'd still be up for debate. The beauty of mathematical ranking systems is that they are based purely on what has happened NOT what MIGHT happen in the future AND they also seem to get better with each passing week given more games to enter data with. Much like the RPI in basketball some mathematical ranking systems are practically worthless until almost 75% of the season is over but that does not mean those systems aren't still going to give you their current rankings and data even in the early season. You just need to know what type of system youa re dealing with and then just acknowledge and know that some systems are better indicators or true rank later in the season as opposed to earlier. This system seems liek one that adjusts itself accordingly the further into the season you get and with more data the more accurate it becomes. I'd bet it's spot on by the end of the season but probably ddoes have early season hiccups every year.
Nothing is perfect not even the so called head to head matchup test, or even common opponent test. I mean what about when Team A beats Team B three times by scores of 6-4, 10-6 and 7-4, Team A loses to Team C three times by scores of 2-1, 4-3, and 3-2, and Team B beats Team C by scores of 7-0, 5-1, and 6-3. Tell me, which team is the "Best team" in this scenario??? (FYI, this is a real thing that happened last year down in our area).
JSR drop your "what about" scenario -- the actual facts are PLS defeated LVS 3-2 in the Eden Prairie Tournament (Fact) and when they play a common opponent in the tournament PLS defeated Rochester 8-0 and LVS tied Rochester 1-1 (FACT), and PLS won the tournament (FACT). Also, regarding the two teams record PLS is undefeated with no ties and LSV has 2 losses and 1 ties, hence with these FACTS PLS should obviously be ranked #1 and LVS #2, until one of them falters or LVS defeats PLS. What I am trying to is have a good mathemetical ranking system be improved upon. So if Monsterbuck1 is the brains behind the ranking system -- I challenge him to improve upon it and not become complacent with the current system used by MyHockeyRankings.com.
Does any of this make sense?
JSR wrote:
I'll disagree with you somewhat in that a ranking system is not necessarily built to predict the future or who will win this weekends games. Further, ONE singular head to head match up is not, atleast not to me, an indicator that one team is definitively better than the other team. I'd say a minimum of 3 games, (one home, one away and one neutral) would be needed before I could say with any certainty one team is better than the other and even then if they were to both win at home and tie on the neutral rink it'd still be up for debate. The beauty of mathematical ranking systems is that they are based purely on what has happened NOT what MIGHT happen in the future AND they also seem to get better with each passing week given more games to enter data with. Much like the RPI in basketball some mathematical ranking systems are practically worthless until almost 75% of the season is over but that does not mean those systems aren't still going to give you their current rankings and data even in the early season. You just need to know what type of system youa re dealing with and then just acknowledge and know that some systems are better indicators or true rank later in the season as opposed to earlier. This system seems liek one that adjusts itself accordingly the further into the season you get and with more data the more accurate it becomes. I'd bet it's spot on by the end of the season but probably ddoes have early season hiccups every year.
Nothing is perfect not even the so called head to head matchup test, or even common opponent test. I mean what about when Team A beats Team B three times by scores of 6-4, 10-6 and 7-4, Team A loses to Team C three times by scores of 2-1, 4-3, and 3-2, and Team B beats Team C by scores of 7-0, 5-1, and 6-3. Tell me, which team is the "Best team" in this scenario??? (FYI, this is a real thing that happened last year down in our area).
JSR drop your "what about" scenario -- the actual facts are PLS defeated LVS 3-2 in the Eden Prairie Tournament (Fact) and when they play a common opponent in the tournament PLS defeated Rochester 8-0 and LVS tied Rochester 1-1 (FACT), and PLS won the tournament (FACT). Also, regarding the two teams record PLS is undefeated with no ties and LSV has 2 losses and 1 ties, hence with these FACTS PLS should obviously be ranked #1 and LVS #2, until one of them falters or LVS defeats PLS. What I am trying to is have a good mathemetical ranking system be improved upon. So if Monsterbuck1 is the brains behind the ranking system -- I challenge him to improve upon it and not become complacent with the current system used by MyHockeyRankings.com.
Does any of this make sense?
FACT- this makes sense.
I don't want to get in the middle of this argument but let's remember these are youth rankings (for fun). Prior Lake beat Lakeville South 2-1. There is a lot of great hockey ahead for everyone let's sit back and enjoy it.
PuckSense10 wrote:
JSR drop your "what about" scenario -- the actual facts are PLS defeated LVS 3-2 in the Eden Prairie Tournament (Fact) and when they play a common opponent in the tournament PLS defeated Rochester 8-0 and LVS tied Rochester 1-1 (FACT), and PLS won the tournament (FACT). Also, regarding the two teams record PLS is undefeated with no ties and LSV has 2 losses and 1 ties, hence with these FACTS PLS should obviously be ranked #1 and LVS #2, until one of them falters or LVS defeats PLS. What I am trying to is have a good mathemetical ranking system be improved upon. So if Monsterbuck1 is the brains behind the ranking system -- I challenge him to improve upon it and not become complacent with the current system used by MyHockeyRankings.com.
Does any of this make sense?
FACT- this makes sense.
I don't want to get in the middle of this argument but let's remember these are youth rankings (for fun). Prior Lake beat Lakeville South 2-1. There is a lot of great hockey ahead for everyone let's sit back and enjoy it.
You guys are acting crazy! I simply cut and pasted the mathematical rankings from rankings.com PWGUY33. I'm sick and tired of hearing about PL and LVS and who is #1 - Who cares! Both teams are very good right now but so are others. I certainly don't have the time to go and watch teams or sit in my office to churn out mathematical rankings. I'm very glad that some people do however. I simply wanted to show how two different rankings compare. Here is the impact that mathematical rankings had on general rankings:
PL -1
LVS+1
EP -2
ER +1
OMG-1
Herm -1
Tonka -3
Still +4
High - Even
Edina +2
Are the plus teams over-rated by mathematics or are the negative teams over-rated by general rankings. Maybe time will tell. It is interesting to see that ER, Still, and Edina were all in the East pool in Eden Prairie and came out of the weekend over-rated by the people who are watching and ranking the teams. Does bias somehow creep into these rankings or was Einstein and his mathematical mind truly crazy???
elliott70 wrote:I think pucksense is pointing out something that is obvious.
Prior Lake is better than LS.
Rankings are all about which team is better.
The best indicator is when one team beats another team.
Especially in the recent past.
Your equation needs to be tweaked to be considered a good ranking system. Right not it is just a calculation.
As both Lee and Mitch point out on their systems, the early season calcualtions are not very good. Your system has no District 12 or 16 teams 'ranked' anywhere. Because they have no numbers. But East Grand Forks, Roseau, Bemidji, Warroad, Grand Rapids may very well have a top 20 to 25 team.
I look at rankings to tell me who will win in this weekends match-ups. History helps develop that, but historical numbers do not tell everything in sports.
But none the less, nice job.
I'll disagree with you somewhat in that a ranking system is not necessarily built to predict the future or who will win this weekends games.
I look at
Perhaps not everyone, but think about it....
Further, ONE singular head to head match up is not, atleast not to me, an indicator that one team is definitively better than the other team.
Some people said thatafter the state 4A football game with Bemidji and Rocori...
I heard people say, "Bemidji would...."
Not so, the winner is always the better team (at that point in time).
How do you determien the better team, you play them, thus the play-down system.
I'd say a minimum of 3 games, (one home, one away and one neutral) would be needed before I could say with any certainty one team is better than the other and even then if they were to both win at home and tie on the neutral rink it'd still be up for debate. The beauty of mathematical ranking systems is that they are based purely on what has happened NOT what MIGHT happen in the future AND they also seem to get better with each passing week given more games to enter data with. Much like the RPI in basketball some mathematical ranking systems are practically worthless until almost 75% of the season is over but that does not mean those systems aren't still going to give you their current rankings and data even in the early season. You just need to know what type of system youa re dealing with and then just acknowledge and know that some systems are better indicators or true rank later in the season as opposed to earlier. This system seems liek one that adjusts itself accordingly the further into the season you get and with more data the more accurate it becomes. I'd bet it's spot on by the end of the season but probably ddoes have early season hiccups every year.
Nothing is perfect not even the so called head to head matchup test, or even common opponent test. I mean what about when Team A beats Team B three times by scores of 6-4, 10-6 and 7-4, Team A loses to Team C three times by scores of 2-1, 4-3, and 3-2, and Team B beats Team C by scores of 7-0, 5-1, and 6-3. Tell me, which team is the "Best team" in this scenario??? (FYI, this is a real thing that happened last year down in our area).
Yup, things do get complicated, but when #1 plays #2, the winner is #1 (until you get more information)
But it sure makes all this youth hockey stuff more fun.....
Not so, the winner is always the better team (at that point in time).
I respectfully disagree. The winner is just that, the winner but not necessarily the better team. This is especially true of sports that have goalies. I've seen way too many games where one team completely dominates the other team but still loses. The other team won the game fair and square but that doesn't mean they are the better team. So have to disagree with you on that one.
old goalie85 wrote:Soooo you are saying the goalie is not part of the team?????
No, not what I am saying at all. But when a team (true story) gets outshot 57-4 and 95% of the action is in one zone or the neutral zone, and the team with 57 shots hits the post 7 times and has 4 goals waived off by the ref for "in the crease" (skates barely touching the crease and no goalie interference mind you) but the team with 4 shots on goal wins 1-0 and the one goal was a dump into the zone from barely across the redline that took a funny bounce and jumped over the goalies sitck and pad, are you seriously going to tell me the team with 4 shots on goal is the better TEAM, regardless of how good or not good the goalies were or were not
old goalie85 wrote:Soooo you are saying the goalie is not part of the team?????
No, not what I am saying at all. But when a team (true story) gets outshot 57-4 and 95% of the action is in one zone or the neutral zone, and the team with 57 shots hits the post 7 times and has 4 goals waived off by the ref for "in the crease" (skates barely touching the crease and no goalie interference mind you) but the team with 4 shots on goal wins 1-0 and the one goal was a dump into the zone from barely across the redline that took a funny bounce and jumped over the goalies sitck and pad, are you seriously going to tell me the team with 4 shots on goal is the better TEAM, regardless of how good or not good the goalies were or were not
I agree.....
A single mid-season victory does not always mean the winning team is "better". If Columbus beats Pittsburgh, does that mean the Blue Jackets are better than than the Penguins? Of course not.... The "better" team does not always win....
The power rankings do not, and should not, consider head to head match-ups. Head to head match-ups are used as tie breakers when it comes to seeding teams into a playoff format....when you talk about Power Rankings, it's more about the TOTAL body of work (strength of schedule and AGD).
Any ranking only attempts to tell how a team is matched up to another team. Early or during the first part of the season just cannot be accurate no matter what systems you used. But even the ranking at the end of the season, would not get one team into the regional or state.
With that being said, still like to hear from those who said there are 10 teams besides PL, can beat LVS.
It is easy to make a statement, give your details to back it up!
After a busy Thanksgiving Weekend it's been a little slow on the game front but after watching 12 games this past week here is what we have come up with for the current rankings. We will be in Duluth this weekend for the Spirit of Duluth PWA Tournament. We could see the rubber match between Prior Lake and Eden Prairie.
1. Prior Lake
2. Lakeville South
3. Eden Prairie
4. Elk River
5. Hermantown
6. Osseo/Maple Grove
7. Stillwater
8. Edina
9. Minnetonka
10. Highland Central
11. STMA
12. Jefferson
13. Farmington
14. Centennial
15. White Bear Lake
16. St. Cloud
17. Waconia
18. Mounds View
19. Lakeville North
20. Wayzata Blue
New Teams: Mounds View, Wayzata Blue
Teams Out: Blaine, Woodbury
