MIAA Seedings

[Mia A]

I've read complaining about seedings here and wondered if it's true the system the MIAA uses is flawed. For a quick check, I examined the different divisions and regions for both boy's and girl's to find how many of the top 4 seeds made their respective semi-finals.

On the boy's side, 39 of the 48 top seeds made the semis. That's 81%. Pretty good. The West and South went almost to form with only 1 top team in each being upset. Central had 2 teams upset. D1 + D2 North had all the upsets. 3 went as expected, but only 1 D1 and 2 D2 North top seeds made it through.

The girl's are a different story. The West and Central brackets tracked well. 9 of 12 top seeds in the Central and 10 of 12 in the West got through. Only 6 of 12 of the South teams made their respective final 4. And again the North had all the upsets with only 3 of 12 top teams making the cut. Still, the current system predicted 58% of the final 4 teams correctly. If only I picked horses that well.

Looking at boy's and girl's combined the west, with admittedly fewer teams, predictions were the closest with 21 of 24 (88%) of top 4 teams getting through, followed by Central with 19 of 24 (79%) and then South 17 of 24 (71%). Overall only the North can truly gripe about seedings. Boy's and girl's combined the MIAA North seedings only got 10 of 24 (42%) of the semifinal teams correct.

Talking Soccer seems to attract more girl's supporters, so perhaps the fact that the seedings have less predictive value for them, is why there seems to be so much whinging here online. My sense is other than the North, MIAA gets it right.

Why is the North different? Do North leagues vary greatly in their relative strength? Is it like this every year?

[Unregistered]

I've read complaining about seedings here and wondered if it's true the system the MIAA uses is flawed. For a quick check, I examined the different divisions and regions for both boy's and girl's to find how many of the top 4 seeds made their respective semi-finals.

On the boy's side, 39 of the 48 top seeds made the semis. That's 81%. Pretty good. The West and South went almost to form with only 1 top team in each being upset. Central had 2 teams upset. D1 + D2 North had all the upsets. 3 went as expected, but only 1 D1 and 2 D2 North top seeds made it through.

The girl's are a different story. The West and Central brackets tracked well. 9 of 12 top seeds in the Central and 10 of 12 in the West got through. Only 6 of 12 of the South teams made their respective final 4. And again the North had all the upsets with only 3 of 12 top teams making the cut. Still, the current system predicted 58% of the final 4 teams correctly. If only I picked horses that well.

Looking at boy's and girl's combined the west, with admittedly fewer teams, predictions were the closest with 21 of 24 (88%) of top 4 teams getting through, followed by Central with 19 of 24 (79%) and then South 17 of 24 (71%). Overall only the North can truly gripe about seedings. Boy's and girl's combined the MIAA North seedings only got 10 of 24 (42%) of the semifinal teams correct.

Talking Soccer seems to attract more girl's supporters, so perhaps the fact that the seedings have less predictive value for them, is why there seems to be so much whinging here online. My sense is other than the North, MIAA gets it right.

Why is the North different? Do North leagues vary greatly in their relative strength? Is it like this every year?

Seeding in the North, South and Central is based on winning percentage, West bases it on the Walker system.

[Unregistered]

I understand the Walker system. The question is why doesn't the winning percentage seeding, which does a decent job elsewhere, predict the North division as closely?

[Unregistered]

not to difficult to figure out, north has the most teams overall per div and hence the seeding are not as signifigant and upsets more likely as they are more closely matched. next question.

[Unregistered]

not to difficult to figure out, north has the most teams overall per div and hence the seeding are not as signifigant and upsets more likely as they are more closely matched. next question.

actually, the biggest flaw in the system are the regions. North and South are both large, lots of competitive teams from multiple competitive leagues, have to win 4 (or 5) games to reach state semis...then North/South winners play each other to reach the finals. Central has same 4-5 teams from 1 league essentially who are larger than the rest, coast into regional semi's (to play the same league teams again), then only 1-2 tough games before playing the Western champ to reach the state finals. Compared to a strong North or South team, a strong Central or Western team has a much easier path to the State Finals game.

[Unregistered]

I understand the Walker system. The question is why doesn't the winning percentage seeding, which does a decent job elsewhere, predict the North division as closely?

Seems like the North has more of those little catholics and tech and charter schools; those are the schools that through everything off.

[Unregistered]

Seems like the North has more of those little catholics and tech and charter schools; those are the schools that through everything off.

Maybe but if you look at D3N Boys this year, the percentage seeding method is holding pretty true and that is where those "little catholics and charters" were found - nonetheless, the 1 and 2 seeds are finalists, having out out the 3 and 4 seeds in yesterday's semis - so I'm unconvinced - but I concede there were more upsets in years past....

[HS Harry]

Let me preface the below by stating it deals specifically with the north region and the girls side:

In the past, most leagues where single division where each team played every conference opponent twice (home and away). The leagues were comparably equal in strength (especially as to D2) so every top team played a relatively equal number of stronger, mediocre and weaker teams. So the D1 and d2 seeds basically comported to actual strength.

During the last decade the soccer demographics also shifted a little west leaving leagues like the GBL, NEC and MVC comparativley weaker (in D1) to the Midd and DCL. This left a 6-7 year period where no D1 team from a weaker conference beat a team from the other two (or Masco, Newton North or Framingham) in the tourney. So you would occasionally have teams from the GBL (like Peabody) or MVC (like Andover) that would be a higher seed but lose to a team from the DCL or Baystate. But those results where entirely predictable and didn't involve many higher seeded teams.

Over the last three years there has been lots of intra-conference movement (leaving the GBL and Boston City even weaker) and the establishment of weighted divisions within the conferences. This leaves schedules unbalanced and avails teams of more non-league games, further unbalancing schedules. So despite the fact that Danvers and Gloucester are both d2 teams from the NEC, their schedules couldn't be more different. And that helped Gloucester to gain a higher seed because of a significantly weaker schedule. Same can be said for Concord Carlilse, Stoneham and Newburyport - all of who played out of the weaker side of their respective leagues. It is even worse in D3 where some of the leagues are simply not very good.

Ultimately to win a title you have to beat some good teams, but the problem is in the route and its not just because of the mis-seeding of top teams. The complete arbitrary nature of the bracket is the problem. Shift a team one or two spots (often done by coin-flips) and the bracket loses any competitive symmetry. Chelmsford has betaen two solid teams to get to the semis, but still doesn't face the type of difficulty that #1 Peabody does and North Andover's route is the easiest.

With computers there is no great difficulty in creating a RPI style/ objective seeding mechanism for these tournaments. They already have tried it with excellent success in the girls lacrosse tourney. But personally I'd rather see them address the officiating issues first!!

[Unregistered]

actually, the biggest flaw in the system are the regions. North and South are both large, lots of competitive teams from multiple competitive leagues, have to win 4 (or 5) games to reach state semis...then North/South winners play each other to reach the finals. Central has same 4-5 teams from 1 league essentially who are larger than the rest, coast into regional semi's (to play the same league teams again), then only 1-2 tough games before playing the Western champ to reach the state finals. Compared to a strong North or South team, a strong Central or Western team has a much easier path to the State Finals game.

this may be correct for wmass but not always for cmass. plenty of times the cmass team plays the same number of games as the emass champ and you are wrong about cmass and who wins. div 1 is dominated by the same teams but div 2 and 3 are always between several different leagues like the swcl, dvc, mid wach etc. and not one league.

[Unregistered]

I understand the Walker system. The question is why doesn't the winning percentage seeding, which does a decent job elsewhere, predict the North division as closely?

that system is DYNoMITE!!