Top 25 talk

[Gregory Sager]

If that's true, then I don't really remember this methodological rift among the voters being a topic of discussion before. It seems to me to be an interesting subject for debate.

[Mr. Ypsi]

If that's true, then I don't really remember this methodological rift among the voters being a topic of discussion before. It seems to me to be an interesting subject for debate.

Wow - I can't let an opening like that go by!

The whole point of the Posters' Poll is debating WHY we rank the way we rank (sometimes vs. one another; often why we think college X is overrated or team Y is underrated on the real poll).  While I haven't yet tallied the confirmations on that thread, I think we only have 7 or 8 so far confirmed to participate - I invite one and all who DO take a national interest in d3 hoops.

[Gregory Sager]

If that's true, then I don't really remember this methodological rift among the voters being a topic of discussion before. It seems to me to be an interesting subject for debate.

Wow - I can't let an opening like that go by!

The whole point of the Posters' Poll is debating WHY we rank the way we rank (sometimes vs. one another; often why we think college X is overrated or team Y is underrated on the real poll).

That's a good point, Chuck. While I know that Pat isn't thrilled about the Posters' Poll, Chuck is right that one of the benefits of it is that participants openly discuss their picks, both defending their own choices and challenging the choices of others. It really opens up the whole methodology question in a way that doesn't take place here in the official Top 25 room, since Pat, Bob, and D-Mac are the only voters who've publicly identified themselves.

[KnightSlappy]

If that's true, then I don't really remember this methodological rift among the voters being a topic of discussion before. It seems to me to be an interesting subject for debate.

I think a healthy top 25 poll will have several methodologies represented by the voters.

There's the "who I feel is best" method that tends to overrate "big name" teams based on tradition. (Wooster, Hope, Calvin, WashU, Wheaton) This method is good at preventing the week-to-week overreactions that may otherwise drop a strong team (Wash U) too far for a fluke (seemingly) loss.

There's the W-L method where W-L record, especially for undefeated teams, weighs very very heavily.

There's the SOS method that takes schedule difficulty and number crunching very seriously.

There's the head-to-head or common opponent method that weighs direct results against each other.

There's the margin of vicory method that rewards big wins and doesn't trust slim margins.

But most people probably combine these methods (and other methods that I didn't immediately think of) to form their own method (one that can justify almost any team at almost any ranking).

It would be boring if every voter used the same method. There would be very little splitting of the first place votes and very few ORV teams. Why would you have a poll if everyone knew (and agreed with) the placement of the teams?

[Hugenerd]

I think that most voters use the sliding scale, but it appears their sliding scale is even simpler than just W/Ls, I think they only look at previous week's results (one week of W/Ls) and the previous week's rankings when making their decision.

Let me use Wheaton (IL) as an example (and remember, this is only an example, I probably could have done the same analysis with other teams).  Wheaton has lost 3 games yet are still ranked tied for#13.  With 3 losses at this point in the season and being ranked so high, you would expext those losses to be to highly ranked teams.  Yet, only one of their losses is to a ranked team (WashU) while another is to ORV 33 (Hope, who crushed them by 22) and the third is to a team that has not received a vote all season.  With these losses, you would expect some big time wins to justify their ranking, so lets look at that.  Of their wins, only Trine, Chicago, and McMurry currently have winning records.  Chicago may have looked good at the time, resulting in a bump up of their ranking at the time, but Chicago has now lost 4 in a row and 5 of 7, so is that win really impressive looking back now?  The Trine win isnt impressive either, so that leaves us only with the McMurry win.  McMurry may be a great win, but right now they are tied for ORV 38.  By the way, the Pomona-Pitzer team that Wheaton beat in 4 OTs this weekend was crushed by WashU by 26 points.  So what exactly is keeping Wheaton (IL) ranked so highly?  In my opinion the only factor is ptheir preseason ranking.  If voters would look at each game result each week of all teams considered, instead of only one week of results + their previous opinion of the team, I think the rankings would be different than they are.  How much, Im not sure, but I definitely feel some teams are still hanging on based on their preseason rankings while others are too low based on not being on anyone's radar early in the season.

For that reason, I think a team like Eastern Mennonite is ranked too low.  They haven't lost to a single d3 opponent, and have a win over a current top 12 team on the road, yet only broke into the rankings this week.  This happened for two reasons: 1) they beat the current top 12 team when they were not ranked top 12 at the time, and 2) they were not ranked in the preseason.  I have no problem with either of these points, but you would think that as time goes on and you gain more knowledge about each team, you could apply that knowledge to change your opinion of that team.  You would think that when VWU keeps moving up someone would say to themselves: "Geez, this Eastern Mennonite team hasnt lost to anyone in our division and they beat a top 12 team on the road, they probably deserve a higher ranking right now until proven otherwise," rather than discounting their good win by saying, "I didnt think they were good at the beginning of the season, and I am still skeptical now, they probably just got lucky on the road so I will leave them off my ballot."  I mean, is a win over VWU really more impressive now than a few weeks ago, just because they are ranked higher now and were presumably underranked at the time?  Similarly, now that JCU clearly isnt as good as everyone thought, was Wilmington's jump from no votes to #10 in the country justified.  I dont think so, but everyone has a right to their opinion.

We are far enough a long in the season that one can vote on teams based not only on personal opinion.  Take a case in the northeast that has recently been discussed: Middlebury and Brandeis.  Middlebury has destroyed everyone they have played, but played a relatively weak schedule.  I can say, based on opinion (just for the sake of argument), Middlebury is ranked too high, but how would I prove that?  In my opinion, Middlebury is justifiably a top 10 team at this point, because they havent lost and won convincingly.  Even if someone ranked them first, I would not agree with it, but I dont think I could come up with a solid argument as to why that choice was not justified.  Brandeis on the other hand has played a relatively weak schedule also, yet lost a game to a previously highly ranked UMD team (who is no longer receiving any votes in the poll).  Well I would argue that that loss is now a "bad" loss for a top 10 team, but where are the good wins to offset that loss?  Brandeis really doesnt have any.  Well in response, one could say, "It's only one game, they may have had a bad day, even WashU has one "bad" loss".  My response to that would be that although WashU and Brandeis only have a bad loss. I think that the one loss means more with respect to my opinion of Brandeis because they have not played nearly as many quality opponents.  WashU has played several games against top 25 teams this year and won all of them, Brandeis has only played one such game and lost it.

Clearly, regardless of how you fill out your poll, there will be a lot of personal opinion involved, especially in d3 where there is a relative disconnect between regions.  However, the things I dont understand are why some head-to-head results are discounted while others are weighed heavily and how significant the time-value of wins are.  What I mean by this is how much emphasis is put on when a win occurred with respect to the ranking of the team at that time.  If we know a team is not as good now as was thought or a team is better now than what we thought, the value of a win against that team should be adjusted to account for what we currently know and not just ignored as something that happened in the past and is out of the time frame of relavence.

I honestly dont care that much which method voters use. My main disappointment is that it seems like a lot of voters only look at a single week at a time when voting, when I think they should look at the whole body of work each week.

[WUPHF]

Are the voters provided with a more detailed look at the Division III schools prior to the vote each week or are the voters expected to do research in the same way the average fan would?

[Mr. Ypsi]

Are the voters provided with a more detailed look at the Division III schools prior to the vote each week or are the voters expected to do research in the same way the average fan would?

They get an extensive look at the top 50 or so schools.  Pat was kind enough to share such a release with me once.

[ScotsFan]

Interesting that Wooster has only one more loss than Wheaton and of Wooster's 4 losses, only one wasn't highly ranked at the time Wooster lost the game.  And you can't really say that Wheaton has a signature win on their resume to warrant their lofty ranking either just like Wooster which seems to be the reason why the Scots have dropped out of the poll altogether.  I guess I just don't get what is so special about Wheaton that they can continue to hover around the top 10 with 3 losses and teams like Eastern Mennonite struggle to make it into the poll period? 

That said, these are just but a few of the reasons I don't envy those who put in the time and the effort to make the D3hoops.com top 25 poll the most accurate poll we have going.  The top 5 are hard enough to figure out.  It's far more fun to sit back as a fan and debate why team x is way overrtated and team y should be ranked higher.

BTW, this is the same reason why I haven't thrown my hat into the pollsters' poll.  I like to keep an eye on the national D3 scene, but I'm not even close to being well versed enough on each and every team out there I feel should be ranked higher than they are and those others I feel are a bit overrated.

[Titan Q]

There's the "who I feel is best" method that tends to overrate "big name" teams based on tradition. (Wooster, Hope, Calvin, WashU, Wheaton) This method is good at preventing the week-to-week overreactions that may otherwise drop a strong team (Wash U) too far for a fluke (seemingly) loss.

I don't think that is necessarily a fair assumption.  I can only speak for myself - I'm on record as being in the "who I feel is best" camp - and this week I switched my #1 vote from Wash U to Randolph-Macon.  Obviously, Wash U would have been the "big name" pick.  

For me, this approach simply means taking all available information and trying to sort out the teams in order, 1 through 25.  When I finalize my ballot, I believe my #1 is a better team than my #2, #2 better than #3, etc (who would win a neutral court game kind of thing).  

I think you are referring to another dynamic - "big name bias."  That certainly exists...most often in November and December (even early January), when there may not be enough current-season data to really evaluate a "surprise" team...so you end up going with a more known commodity.   But I honestly think this is something entirely separate from the voting philosophy of trying to determine who is best, in order.

[Titan Q]

Are the voters provided with a more detailed look at the Division III schools prior to the vote each week or are the voters expected to do research in the same way the average fan would?

They get an extensive look at the top 50 or so schools.  Pat was kind enough to share such a release with me once.

The voters also talk.  For example, this week I contacted a head coach (who votes) who faced Illinois Wesleyan.  I basically just asked, "Where do you see IWU?"  He shared his thoughts on the Titans and how he sees them stacking up relative to the Top 25 poll.

So there is a lot of that that goes on behind the scenes as well.   

[Titan Q]

I honestly dont care that much which method voters use. My main disappointment is that it seems like a lot of voters only look at a single week at a time when voting, when I think they should look at the whole body of work each week.

I know a lot of the voters pretty well, and talk to several of them regularly.  I honestly can't think of any that seem to just look at things week to week.  Everyone I talk to works very hard to look at the big picture.  

That doesn't mean voters don't fall asleep at the wheel or make mistakes.  I was extremely disappointed this season in the preseason poll to see votes for Elmhurst, for example.  It didn't take much research to learn that Elmhurst would be in complete rebuilding mode this year.  I know I have made a lot of mistakes in my 11 years as a voter...not intentional, but there is just a lot to process.

The panel that votes in the D3hoops.com poll knows and loves Division III basketball, and cares a lot about creating an accurate Top 25 poll.  You'd be surprised how much thought goes into it collectively week- to-week.

[Hugenerd]

I honestly dont care that much which method voters use. My main disappointment is that it seems like a lot of voters only look at a single week at a time when voting, when I think they should look at the whole body of work each week.

I know a lot of the voters pretty well, and talk to several of them regularly.  I honestly can't think of any that seem to just look at things week to week.  Everyone I talk to works very hard to look at the big picture.  

That doesn't mean voters don't fall asleep at the wheel or make mistakes.  I was extremely disappointed this season in the preseason poll to see votes for Elmhurst, for example.  It didn't take much research to learn that Elmhurst would be in complete rebuilding mode this year.  I know I have made a lot of mistakes in my 11 years as a voter...not intentional, but there is just a lot to process.

The panel that votes in the D3hoops.com poll knows and loves Division III basketball, and cares a lot about creating an accurate Top 25 poll.  You'd be surprised how much thought goes into it collectively week- to-week.

Well in that case, I stand corrected, but sometimes it just doesnt seem that previous weeks results are weighed at all.  I appreciate the effort that voters put into the voting process, it is undoubtedly a difficult task.  But as scotsfan stated, it is hard to understand the positioning of some teams compared to other teams sometimes.

[Titan Q]

But as scotsfan stated, it is hard to understand the positioning of some teams compared to other teams sometimes.

And that will probably always be the case with any poll voted on by people.  Until next year...

Word is, starting in 2010-11 Pat will fly each voter to a neutral site each Tuesday for a press conference.  There will be a really long table, with 25 voters stretched across, each having a microphone.  A big D3hoops.com logo will serve as the backdrop.  D3 fans will be invited and a copy of all 25 ballots will be distributed (with names attached)...the voters will be thoroughly grilled.

(We're just trying to get the location pinned down.  Pat wants us to come to Minnesota...I'm holding out for Vegas.)

[Hugenerd]

But as scotsfan stated, it is hard to understand the positioning of some teams compared to other teams sometimes.

And that will probably always be the case with any poll voted on by people.  Until next year...

Word is, starting in 2010-11 Pat will fly each voter to a neutral site each Tuesday for a press conference.  There will be a really long table, with 25 voters stretched across, each having a microphone.  A big D3hoops.com logo will serve as the backdrop.  D3 fans will be invited and a copy of all 25 ballots will be distributed (with names attached)...the voters will be thoroughly grilled.

(We're just trying to get the location pinned down.  Pat wants us to come to Minnesota...I'm holding out for Vegas.)

I wasnt trying to say that all 25 voters needed to debate each spot in the poll, I just find a few of the teams rankings illogical given their results and given the results of similar teams.

To follow up with the example of Wheaton, they currently have 259 points, meaning on average they are ranked 14.5 on ballots.  I am sure that this average represents some distribution, with some people voting them in the 20s, some in the teens, some in the top 10.  I am just curious, with their 3 losses, who would put them in the top 10 or close to that number, when there are teams, even from "power" conferences, with less losses and losses to stronger opponents?  That is all I am trying to say.  (Again, not trying to pick on Wheaton, but they just seem like the best example currently.)

[Pat Coleman]

Voters gets sent about three weeks worth of info on each team each week, plus are reminded of each loss. Here's Wheaton's emailed resume, for example:

No. 8 Wheaton (Ill.) (8-3, 0-0):
Dec 09    AWAY    Washington U. (9-1, 0-0)    L    64-62
Dec 12    HOME    Grinnell (1-7, 0-2)    W    127-100
Dec 29    AWAY    Hamline (6-4, 2-3)    L    69-68
Dec 30    HOME    Monmouth (1-9, 0-3)    W    84-53
Jan 02    AWAY    Pomona-Pitzer (4-6, 0-0)    W    100-102  4OT
Jan 06    AWAY    Carthage (8-3, 0-0)        -
Losses: Hope,Washington U.,Hamline,

Of course, if you take the entire break off, you may not have much of a resume to work from:

No. 13 Brandeis (7-1, 0-0):
Jan 04    AWAY    Curry (3-6, 0-0)        -
Losses: Mass-Dartmouth,