any one have any predictions on teams or seedings now that the seasons are slowly winding down... whos the top A AA AAA teams as of now

Rough computerized rankings, courtesy of Carl Wood. Nothing official, but they lend some insight into possible seeding.

http://www.icshl.org/Rankings/ranking.htm

Click on the Rankings tab at the bottom.

And I just noticed the "at risk of making Steve B's head explode," caption on the site. Ha, come on Carl, we discussed that it was 4:30 AM, didn't we?

how accurate is that? cause i love whoever did that haha

Ok, I have to grype a bit about this, because I need to understand the logic. It's concerning II teams (Malvern II, LaSalle II etc..) and other non-eligible Flyer's Cup teams. Why do these teams factor into the seeding process for the Flyer's Cup if they aren't eligible to participate in it? I'm ok if they are involved in their league playoffs, but obviously I think the whole "II team" thing is ridiculous, but that's not my grype. Why are W's and L's against these II teams allowed to factor into the FC selection process? Sorry to expose the following teams, who according to the rankings, if the FC were being picked today would be #'s 18 & 19 respectively.

Carl, this isn't directed at you, but the whole system and process we're running under and we all need to find a better way (that includes removing the Jersey teams that can have an impact on a Pennsylvania State Championship, what a joke that is in this day and age, but that's another battle for another day with the Flyers).

Here are the two case in points:

#18- Interboro:
Overall Record = 5-8-3
vs. FC eligible teams = 3-7-2
(1 win vs. a .500 or better team)

#19- Sun Valley:
Overall Record = 8-6-1
vs. FC eligible = 1-6-1
(the win = Twin Valley, no where near .500)

Now these teams aren't in this boat alone with their schedules, but they are the only two who are in the top 20 in the 'A' rankings.

To agree and paraphrase something Matt said before:
If there aren't 20 teams good enough, cut it down to 16 teams!!

Sorry, and I don't mean to pick on these teams, but those numbers aren't and shouldn't be Flyer's Cup worthy!!

In terms of "ranking" teams, my preference is to include all the league games that teams play as "grist for the mill". Just because a II team is ineligible, or impure, doesn't mean that the game results against them--serving as a common opponent for a number of other teams--isn't useful. Plus there are some (EHSHL) teams that play a fair percentage of their total schedule against FC-ineligible teams. Once I start picking and choosing games to consider from each team's schedule, it opens all kinds of possibilities to manipulate the results. Hopefully, you've read the write-up that accompanied the ranking scheme. It explains that the points against Twin Valley aren't worth nearly as much (based on T.V.'s record) as points earned against, say Ridley (AA team). And for what its worth LaSalle II is the scariest "varsity 2" team you'd ever want to play. After rolling through the SHSHL AA last year, they've provided a nice test for the ICSHL A and AA teams that have played them.

Should we rank "A" teams only on games played against other "A" teams, etc? Other than the SHSHL, all the leagues play a mixed schedule of A/AA or even A/AA/AAA teams, so what's wrong with including teams at the bottom, as well as the top, of the food chain? The methodology ranks teams (within a class) based on who they play and beat. I guess I could include LaSalle and Malvern's out of state games, too, if I had reliable information to classify and rank those teams in this scheme, but our own pool of teams is work enough for right now.

The broader question--what to do about 2nd varsities--is an interesting one. I think we need to organize a "JV elite", "varsity B", whatever you want to call it, to move these teams out of varsity Flyers Cup status, but still provide a competitive developmental level for them. That will be harder than it sounds to put together. We (ICSHL) have some varsity programs that weren't ready for prime time, and a couple of JV A programs that were better than the competition we could offer them at JV A.

Jersey teams? the Flyers and the Atlantic District both want them in the Flyers Cup. HS sports are traditionally organized on state lines not media markets, but there it is.

My original proposal in terms of number of teams in an expanded Flyers Cup was to follow the NJSIAA practice--any team .500 or better makes the state tournament; teams within 3 games of .500 can petition to be included. Their state tournament then adds pig-tail play-in games to handle the numbers. I'd prefer that to the fixed number (20) we adopted. 20 schools leaves some deserving candidates out of AA this year, lets some weaker A teams in. I agree.

This message has been edited. Last edited by: CarlWood,

part of the fun is watching teams move up and down as they win and lose. Somebody's got to be 1-2-3-4, etc. The ranking isn't an absolute step, or we'd have to have 1.01,2.34, etc.

Rustin beating PV will do many things in small increments:

1. reduce PV's weighted winning percentage, probably drop them below Henderson.
2. increase Rustin's weighted winning percentage--significantly, since points versus PV are worth a good deal more than points versus, say, Great Valley.
3. increase the weighted winning percentage of teams that did beat Rustin--since Rustin's power point rating goes up.
4. decrease the weighted winning percentage of teams that did beat PV, since PV isn't quite the power team it was before the game was played.

Aston hasn't sent in their scoresheets from Friday, but I've updated the site for the known game results through Friday.

A fundamental difference in rating systems is their purpose: prediction or reflection (of the season record). Predictive schemes, like the old Dunkel rating system, tend to be based on point spreads and point differentials and are geared at predicting game results. A handy tool for the betting public, among other audiences. Things like injuries, inside knowledge, etc., all make those kinds of predictions better. Why do you think the NFL requires injury reports?

The ranking system I'm posting--strictly for conversation, mind you--is reflective of how each team's season is going. After playing 15-18 games, teams' seasons are pretty much in place. We can separate good teams from bad teams: how good? how bad? Who's played the tougher schedule? Rustin? PV?

I told Matt he should publish his ranking method. He gave me his best Robin Williams doing Popeye the Sailor Man, "I seez what I seez, I likes what I likes...", which is actually a decent answer since Matt sees more games than most.

This message has been edited. Last edited by: CarlWood,

Carl,

I understand the process, I guess I'm just not agreeing with it 100%. I just really have a problem with giving any legitimacy to the whole II team thing. Like you said call them what they are, the JV. Very good JV's, but JV's!!

I guess the only parallel I can find in another sport was several years back, Strath Haven Soccer had a II or Independent team they called it. They played a Varsity schedule but it had no bearing on the Central or any other league. Basically all their games were non-league, they were PIAA playoff ineligible and had no affect on District seedings come playoff time, if I'm not mistaken.

I agree with your comments that some of these teams are very skilled and could do major damage in playoffs and the Flyers Cup if they were allowed. But rightfully their not.

Also, if you answered this scenario in another post or I just didn't catch it in the rules, I apologize, but I think it's a joke if Malvern could wind up playing Malvern II in the ICSHL playoffs. Like I said if this can't happen, my mistake. And I'm not saying these guys wouldn't play it 100% and I'm sure would compete likes cats and dogs against each other.

But like they say, "Somethin' just ain't right about it!"

Just venting a bit, but your the guy who has the most knowledge and have done a great job with everything (even though this doesn't make it sound so).

Since most want to complain about everything and contribute no solutions, I will post my thoughts / solutions / contributions under the "Realignment" topic and wait to be tarred and feathered as you have probably been before.

Bad Johnny, I do agree with your thoughts about Varsity II teams and have made those oppinions very vocal.

Your commantary on Carl's ranks however, make me believe that you think that this is how the Flyers Cup teams will be picked. It's not! It's Carl having fun with numbers....he passed it along to me and surprisingly, My completely unmathematical rankings and Carl's rankings based on his formula have been pretty darn close.

In re Carl's computerized rankings, I say "Bravo!!". The methodology certainly looks sound and anyone who puts in that much effort to reach as objective as possible a result deserves an atta-boy. For those who were never in the military, an atta-boy is more highly-valued than many a medal or commendation.

I like Carl's rankings very much but they do have a blind spot. By not taking into account goal differential or even head-to-head results there can be a serious ranking ommission.

In the case of Malvern II, they are ranked behind Episcopal Academy, yet in a head-to-head battle MPII 10-0'd EA 3 minutes into the 3rd period. I think that's significant.

I discuss in my write-up why goal differential is not part of the criteria I use. If I did use it, I'd look at something like the "quality wins" (+4 goal margins) the DVHL tries to use. The question then becomes what is the "right" additive for a "quality win"--add 0.5 points, add a full 1 point? I also debated giving more weight to late season games than early season games. One of the research papers I linked finds that point differential DOES improve the accuracy of ranking systems--in terms of predicting future results--but that the NCAA has specifically directed that it can't be used in any BCS ranking for "sportsmanship" reasons. And the NCAA doesn't employ mercy rules like we do that limit scoring differential.

Head-to-head versus overall record is another tough call. While team A may be better than team B head-to-head, especially in a predictive mode--who will win the next game between the teams? Does that mean team A has had a better season than team B, which is what my ranking scheme is focusing on?

And of course, when you introduce tweaks or refinements into the process aimed at fixing one perceived problem, the "law of unintended consequences" means other things you won't like will pop out elsewhere. I have played with this for several years, using Matt among others as my control, and the current version of this is the best "fit" to the real world I can manage. For my use, anyway, its to be sure I give all teams a fair look and don't let my preferences and limited exposure pre-judge teams. Beyond that, its a fun argument to pursue--half the reason I agreed to let Matt publish it to start with.

This message has been edited. Last edited by: CarlWood,

Carl,

I think that your ranking system is interesting. I never read that you claimed that it is perfect. No system will ever be that. I do like it on the simple basis that it attempts to objectively quantitate each team, without regard to personal feelings or affiliations. There certainly is something to be said for watching these teams play as opposed to punching the numbers into the spreadsheet and let the computer spit out your answer. I have seen games where the better team certainly HAS NOT won, where an individual has "stolen" a game, and that is where Matt and his eyes, can help correct any of the oversight that the computer may have. Those rankings that you put out are simply for amusement purposes only I take it? Although I can see it being useful as a starting point for a FC committee to use. Then all the politiking (Russian spelling I think) can start.

Every Flyers Cup league sends a representative to the Selection Committee, plus Matt Sacks and Rob Baer from the Flyers. Rob tends to be there to be sure everyone is polite. My database of game results also ensures that everyone has access to the same record, game results, etc., that are not posted consistently from league to league.

Each Rep brings his own knowledge of HS hockey to the selection process. Matt, knowing the most, tends to have influence. I try to get out and see as many of the ICSHL/EHSHL games as I can, and a few the other leagues' games, too. That, combined with cross-over games give the best qualitative insight into teams' relative strength.

The selection process also has to accord all the leagues some competitive dignity--we can't simply dismiss an entire group of teams because of our perception of their league. Giving more starting power points (1.2) at all AA teams compared to A teams (1.0), regardless of how good I know the teams to be or not be, is an example of that. In one sense, it models that "politiking" thing, where each league is entitled to some representation in the process.

Truthfully, the really good teams at A and AA can all beat each other "any given Friday". Depends on who shows up, who's hurt, how team strengths and weakness match up. In my mind playing a "4 out of 7" format (a la NHL) is the only sure way to determine that team A is better than team B. Overall, I'd like to reward teams with seedings based on their season's body of work, not a single game or two results.

The NHL plays 82(?) games to pay the bills, and 16-28 to determine the champions. We don't have the money or the time, so I'd like the 18-25 game regular season to constitute the body of work to qualify for post season honors.

The difficulty in head-to head is when say the #16 team upsets the #2 team (Unionville vs Perk Vally and Henderson for example) These are great wins for Unionville, but it doesn't and shouldnt leapfrog them from #16 to #1, just because they beat the top two teams.

Where head to head should be used is when PV and Henderson have similar/Identical records, playing in the same league. Henderson won the head-to-head game yet Carl's system has PV ranked higher....that is what has me scratching my head.....but, I still like the system overall.

Having already said I don't like to criticize without an alternative my only critique (which may be an opportunity to tweak the system, Carl) is the automatic assignment of weights given league classification status. Is it really appropriate to assign the same weight to a LaSalle and a bottom Tier AAA team? a Perk Valley and a weaker A team? A strong AA team with and a weak one?

It appears that what this does is equate A to good...AA to better and AAA to best...similar to Club Tier rankings and it gives unfair advantage to a club that perhaps scheduled one of the weaker teams with more letters in their league.