3/28/07 Division A Poll is out!

[Sonny]

LINK:
http://www.collegelax.us/news.php?action=fullnews&id=1173

[laxfan25]

Now there are two weeks to debate how this is going to change again, based on very recent events. There definitely doesn't appear to be any clearly dominant team this year - should be a very exciting tourney in Dallas. They won't even need ovens for pizzas at the games - just put them on the field - the games will be HOT!

[lex]

Not to disrespect the pollsters' work, but these rankings are crazy.
Michigan at #1? They haven't played anyone except for FSU, who also hasn't played anyone.
Duluth at #6? They've only beaten Cal Poly and Lindenwood??
Utah at #13? They beat Duluth and UCSB.
Northeastern in the top ten for one win over AZ? Yet ASU knocks off Oregon and only goes up to #11?
I don't understand.

[Sonny]

The point totals are bunched VERY tightly in a few spots.

Teams 6,7,8 are only seperated by 10 points total!

Teams 10,11,12 are only seperated by 8 points total!

Point totals are more important then poll spots until we get to the final poll.

[lex]

Maybe it would help if the pollsters had some guidance to go by. It seems this league is so wrapped up in the number of points a team wins by. A win is a win. Period. I think, if these polls determine the fate of a team in the post-season, pollsters should have to vote based on some kind of criteria. As far as I know, there aren't any.

Here are the NCAA guidelines, I suggest the MCLA adopt something similar so there is some kind of process invovled:

Primary Criteria - For ranking and selection (all contests leading up to NCAA championships).
Win-loss percentage against regional opponents.
Strength-of-schedule index (only contests versus regional competition).
Win on the road versus a team at or above ..667 14 points
Win at home versus a team at or above ..667 13 points
Win on the road versus a team at or above ..500, but below .667 12 points
Win at home versus a team at or above ..500, but below .667 11 points
Win on the road versus a team at or above ..333, but below .500 10 points
Win at home versus a team at or above ..333, but below .500 9 points
Win on the road versus a team below ..333 8 points
Win at home versus a team below ..333 7 points
Loss on the road versus a team at or above ..667 7 points
Loss at home versus a team at or above ..667 6 points
Loss on the road versus a team at or above ..500, but below .667 5 points
Loss at home versus a team at or above ..500, but below .667 4 points
Loss on the road versus a team at or above ..333, but below .500 3 points
Loss at home versus a team at or above ..333, but below .500 2 points
Loss on the road versus a team at or below ..333 1 point
Loss at home versus a team at or below ..333 0 points
Points for tied contests are calculated by taking the points which would have been awarded with a win, adding the points which would have been awarded with a loss and dividing by two.
The strength-of-schedule index is calculated by adding the total number of points and dividing by the number of games.
A neutral game is defined as being in neither team's locale and, as such, is awarded as if it were an away game.
The strength-of-schedule index is calculated for ratings at the time of the ranking calls and using final results for selection purposes.
In-region head-to-head competition.
In-region results versus common regional opponents.
In-region results versus regionally ranked teams.
Opponents are considered ranked once they appear one time in the sport's official rankings.
Conference postseason contests are included.
Secondary Criteria - For ranking and selection.

Out-of-region head-to-head competition.
Overall Division win-loss percentage.
Results versus common non Division opponents.
Results versus Division teams ranked in other regions.
Overall win-loss percentage.
Results versus common out-of-region opponents.
Should a committee find that evaluation of a team's win-loss percentage during the last 25 percent of the season is applicable (i.e., end of season performance), it may adopt such criteria with approval from the championships committee.

[Sonny]

Nice idea lex. Why don't you suggest it to the MCLA board and see if they take you up on your idea?

[lex]

Thanks Sonny, I will, but I don't have that information. Who is on the board? You can send it in a private message if you don't want to waste forum space for the answer.

[Sonny]

Thanks Sonny, I will, but I don't have that information. Who is on the board? You can send it in a private message if you don't want to waste forum space for the answer.

The handy dandy FAQ has a link to the MCLA board:
http://www.collegelax.us/faq.php

[Ben Clark]

Maybe it would help if the pollsters had some guidance to go by. It seems this league is so wrapped up in the number of points a team wins by. A win is a win. Period. I think, if these polls determine the fate of a team in the post-season, pollsters should have to vote based on some kind of criteria. As far as I know, there aren't any.

Here are the NCAA guidelines, I suggest the MCLA adopt something similar so there is some kind of process invovled:

Primary Criteria - For ranking and selection (all contests leading up to NCAA championships).
Win-loss percentage against regional opponents.
Strength-of-schedule index (only contests versus regional competition).
Win on the road versus a team at or above ..667 14 points
Win at home versus a team at or above ..667 13 points
Win on the road versus a team at or above ..500, but below .667 12 points
Win at home versus a team at or above ..500, but below .667 11 points
Win on the road versus a team at or above ..333, but below .500 10 points
Win at home versus a team at or above ..333, but below .500 9 points
Win on the road versus a team below ..333 8 points
Win at home versus a team below ..333 7 points
Loss on the road versus a team at or above ..667 7 points
Loss at home versus a team at or above ..667 6 points
Loss on the road versus a team at or above ..500, but below .667 5 points
Loss at home versus a team at or above ..500, but below .667 4 points
Loss on the road versus a team at or above ..333, but below .500 3 points
Loss at home versus a team at or above ..333, but below .500 2 points
Loss on the road versus a team at or below ..333 1 point
Loss at home versus a team at or below ..333 0 points
Points for tied contests are calculated by taking the points which would have been awarded with a win, adding the points which would have been awarded with a loss and dividing by two.
The strength-of-schedule index is calculated by adding the total number of points and dividing by the number of games.
A neutral game is defined as being in neither team's locale and, as such, is awarded as if it were an away game.
The strength-of-schedule index is calculated for ratings at the time of the ranking calls and using final results for selection purposes.
In-region head-to-head competition.
In-region results versus common regional opponents.
In-region results versus regionally ranked teams.
Opponents are considered ranked once they appear one time in the sport's official rankings.
Conference postseason contests are included.
Secondary Criteria - For ranking and selection.

Out-of-region head-to-head competition.
Overall Division win-loss percentage.
Results versus common non Division opponents.
Results versus Division teams ranked in other regions.
Overall win-loss percentage.
Results versus common out-of-region opponents.
Should a committee find that evaluation of a team's win-loss percentage during the last 25 percent of the season is applicable (i.e., end of season performance), it may adopt such criteria with approval from the championships committee.

But you could just have a computer do that, couldn't you? It would take the human element out of the polls. I'm not sure what the point of that would be. To end the complaining? We all know better than that - people will complain about anything.

[Beta]

But you could just have a computer do that, couldn't you? It would take the human element out of the polls. I'm not sure what the point of that would be. To end the complaining? We all know better than that - people will complain about anything.

Yeah and as we all well know...the bcs is FLAWLESS and no one ever challenges it.

[ME]

I agree that these rankings are very strange. I am disapointed to see teams with big wins such as ASU and Utah only move up a few spots. How is U of A still ranked a head of ASU that doesn't make any since at all after they beat Oregon and U of A didn't. Or why is Utah still ranked behind UCSB after beating them? It seems as if both those teams should have gone up a few more spots than the few that they did.

[lex]

But you could just have a computer do that, couldn't you? It would take the human element out of the polls.

I agree with your post, but the idea is to start with a formalized process and then add a subjective element. That's where the pollster's would come in. Right now, there is no guidance at all. Teams deserve better than that. Even if the league doesn't adopt the NCAA's exact criteria, it should at least adopt some criteria, and then, as in the NCAA, publish that criteria.

[lex]

Based on the chatter on collegelax, I suspect that some pollster rely solely on wins and losses, while others rely heavily on margin of win and margin of loss. While I don't tout one of those over the other, the simple fact is there are X amount of pollsters relying on vastly different criteria. As a starting point, the league should decide what to base rankings on and then educate everyone on that criteria. It's always better to inform people about what they're playing for. (i.e. should a team run up a score instead of putting in their second string in order to satisfy the pollsters - a strategy I disagree with)
As this season shows, teams from all over the country are becoming more competitive, and the league owes it to the teams spending so much time and money to compete in the MCLA to provide a formalized process for what determines their end of season fate.

[Rob Graff]

I agree with your post, but the idea is to start with a formalized process and then add a subjective element. That's where the pollster's would come in. Right now, there is no guidance at all. Teams deserve better than that.

Your NCAA rules argument fails on a number of grounds - one of which Ben has pointed out. Another is pointed out by Beta - that a primarily computer based system has made things arguably worse in college football. I see no reason why the same would not happen in the MCLA.

Last year, before the "message board poll" came out, it was trumpeted (again, before it came out) as the "correct" poll because a few thought the USLIAMDIA poll was incorrect. As it turned out, both polls effectively tracked each other - so much so that there was little to no interest in a Message board poll this year.

Also - there has been general agreement that the AQ/AL combination has yielded an national tourney field that properly represents both the long term interest of improving lacrosse all over the country and the short term interest in getting the best teams there. Are there teams that were left out that have arguments that they should have been there? Absolutely. But that's more a result of the AQ than improper polling. And I think that the consensus is that the AQ is a good and necessary thing for the growth of the sport in the MCLA. And we're not the only organization that uses AQ's at the expense of potentially more talented teams. Exhibit 1 - the NCAA basketball tournament.

And, if you want a "mechanistic" approach to polling, the pollsters can refer to Lax power - which I know some pollsters do review and include in their calculus.

I think you are underestimating the attention that the 40 Division A pollsters give to their poll. We know the importance of it, we know the importance of getting it right. And from all I've spoken to, all are giving it the time it deserves.

[x1dschm]

Maybe it would help if the pollsters had some guidance to go by. It seems this league is so wrapped up in the number of points a team wins by. A win is a win. Period. I think, if these polls determine the fate of a team in the post-season, pollsters should have to vote based on some kind of criteria. As far as I know, there aren't any.

Here are the NCAA guidelines, I suggest the MCLA adopt something similar so there is some kind of process invovled:

Primary Criteria - For ranking and selection (all contests leading up to NCAA championships).
Win-loss percentage against regional opponents.
Strength-of-schedule index (only contests versus regional competition).
Win on the road versus a team at or above ..667 14 points
Win at home versus a team at or above ..667 13 points
Win on the road versus a team at or above ..500, but below .667 12 points
Win at home versus a team at or above ..500, but below .667 11 points
Win on the road versus a team at or above ..333, but below .500 10 points
Win at home versus a team at or above ..333, but below .500 9 points
Win on the road versus a team below ..333 8 points
Win at home versus a team below ..333 7 points
Loss on the road versus a team at or above ..667 7 points
Loss at home versus a team at or above ..667 6 points
Loss on the road versus a team at or above ..500, but below .667 5 points
Loss at home versus a team at or above ..500, but below .667 4 points
Loss on the road versus a team at or above ..333, but below .500 3 points
Loss at home versus a team at or above ..333, but below .500 2 points
Loss on the road versus a team at or below ..333 1 point
Loss at home versus a team at or below ..333 0 points
Points for tied contests are calculated by taking the points which would have been awarded with a win, adding the points which would have been awarded with a loss and dividing by two.
The strength-of-schedule index is calculated by adding the total number of points and dividing by the number of games.
A neutral game is defined as being in neither team's locale and, as such, is awarded as if it were an away game.
The strength-of-schedule index is calculated for ratings at the time of the ranking calls and using final results for selection purposes.
In-region head-to-head competition.
In-region results versus common regional opponents.
In-region results versus regionally ranked teams.
Opponents are considered ranked once they appear one time in the sport's official rankings.
Conference postseason contests are included.
Secondary Criteria - For ranking and selection.

Out-of-region head-to-head competition.
Overall Division win-loss percentage.
Results versus common non Division opponents.
Results versus Division teams ranked in other regions.
Overall win-loss percentage.
Results versus common out-of-region opponents.
Should a committee find that evaluation of a team's win-loss percentage during the last 25 percent of the season is applicable (i.e., end of season performance), it may adopt such criteria with approval from the championships committee.

Lex can you post what the rankings would be right now based on your system? It would be interesting to see how different your poll would be from the "coach's poll." Include the number of points each team has.