Ranking D3 BBall Conferences

[sac]

I kind of like this cherry picking idea...........this brings the MIAA right back into the race.

That's gotta feel really crappy for those players!  "We're playing for the status quo this year!  Ok guys, average on three... one, two three!"

+k

[Gregory Sager]

Geez, what kind of goals are those?  Do you only want to play the first half of games well and the second half doesn't matter, as long as you were leading at halftime?

That's gotta feel really crappy for those players!  "We're playing for the status quo this year!  Ok guys, average on three... one, two three!"

If they play the slate of allowable games (25) then they really CAN'T end up .500!

The impression I got was that President Bennett was talking about the entire spectrum of Earlham athletics, not just the men's basketball program. But I agree with you, and I wondered the same thing about his quote. Worse, what sort of impact will it have upon recruiting? I mean, this quote was in the papers. If I was the coach of another D3 school, and I was vying with Earlham for a high-school prospect, I'd tell the kid, "Do you really want to compete for a school whose self-declared goal is to win half of its games? Or would you rather compete for my school, where our goal is to win all of our games?"

I kind of like this cherry picking idea...........this brings the MIAA right back into the race.

If I had called it tulip-picking instead, it would've vaulted you guys right into the lead.

[Mr. Ypsi]

Greg, I believe Michigan leads the nation in cherry production, too!

[Gregory Sager]

It probably does. If Tim Allen ever gives up that gig doing the voiceovers for the "Pure Michigan" tourism commercials, you and sac can fight it out to be his successor.

[John Gleich]

It probably does. If Tim Allen ever gives up that gig doing the voiceovers for the "Pure Michigan" tourism commercials, you and sac can fight it out to be his successor.

No disrespect to you two Michiganders, but I hope Allen keeps it going... just about the only thing I like more than those "Pure Michigan" commercials is the spoof ad that a suburban mini-golf place did on it over the summer...  Now that was great!

... Back to your regularly scheduled conversations...

[Hugenerd]

Greg, I honestly had already done those calculations.

You could've saved me a lot of time spent wrestling with the calculator function on my cell phone.

The reasons I didnt post them are as follows: First off, Case Western and Emory absolutely kill the UAA, if you take into the top 6, the numbers are nearly all square.

You already know what I'm going to say in response to that, don't you?  Case Western Reserve and Emory are just as much a part of the UAA as are Wash U and Rochester. You can't leave them out.

Secondly, in a previous post, you yourself stated that a good top-to-bottom metric would be comparing the first teams, the second teams, the third teams, and so on; therefore, I did exactly that analysis.

Yes, I did say that, and, yes, you did exactly that. But averaging out the data you presented to give both league ranking totals and mean average of each league's eight teams are good top-to-bottom metrics, too. In fact, they might be even better ways of top-to-bottom comparison than the slot-by-slot method, since they fit the CCIW and UAA into the overall context of four-year college basketball.

Also, I had already stated that over the past 3 years they CCIW's bottom 3 teams have been much better than the UAA's bottom 3.  Over those 3 years (l listed 4 years, so I am only looking at 2007-2009 here), the bottom 3 spots alone have account for +2017 ranking spots for the CCIW (that is 672 spots per year, or 224 spots per position).  Average that over all 8 teams, that is 84 spots per position, taking into account only the difference in the bottom 3 teams.  However, in those same 3 years, they UAA has been +1927 or 642 spots per year, or 80 spots per team.  So therefore, the UAA has been equally strong in its first 5, compared to the CCIW, as it has been week compared to the bottom 3.  Therefore, along with the fact that comparing the 1st place teams, 2nd place teams, third place teams, etc. were based on yours and Ypsis comments, I did not present that data.

Well, I'm making progress. I've got you to move from basing your argument upon the top team to basing your argument upon the top five teams.

While, according to Massey, four years ago the CCIW was better on average than the UAA, although handicapped by two bottom teams that were significantly weaker than anything the UAA had to offer, I would agree that in two of the past three years there has been rough parity between the two leagues in terms of their top fives while the UAA's bottom three were noticeably weaker than their CCIW counterparts. (In 2007-08, the CCIW actually had a weaker bottom three, on average). Nevertheless, you're still cherry-picking the data by drawing two arbitrary cutoff points (in terms of both time span and number of standings slots) rather than taking the data -- and thus, the two leagues themselves -- as a whole.

I still would rather be strong in the first 5 spots than in the final 3: edge UAA.  What that analysis truly shows is how weak the bottom 3 have been for the UAA in comparison to their CCIW counterparts.

"Edge: UAA"? Even if you cherry-pick the numbers in the manner you suggested -- by dropping out the bottom three from each league over the past three seasons -- the UAA gains only 2,067 points on the CCIW, and that's not enough to close the 2,455-point gap between the two leagues. If anything, that only brings the UAA into rough parity with the CCIW; it doesn't give the UAA an edge at all.

It's all still dancing around the bottom line, though, which is that you can't hide any of your league's teams and pretend that they don't exist. A league still has to be held accountable from the penthouse to the basement.

Greg,  I said this in my post, but you apparently missed it.  You averaged the entire 4 years in your initial analysis, while I based my analysis on the last 3 years (I listed 4 years in my initial post because that is how far back I could get the Massey data, I would have gone back further if I could get more data, but I have basing my argument on the past 3 years in as early as my first post on this board, and therefore I will stick with 3 years).  The UAA is +1927 overall over the past 3 years for the top 5 spots compare to the CCIW.  That means that, on average, the top 5 teams are ranked +128 spots higher compared to their respective CCIW team (1927/3 seasons/5 spots), while the bottom is -2017, or -224 spots per bottom 3 team (2017/3 seasons / 3 teams).  This means that, overall the conference are very close on average over all 8 teams; however, the reason for this is because the bottom 3 for the UAA has been so bad.  In otherwords, on average, the UAA has been significantly better than the CCIW in the top 5 spots and the UAA has been nearly equally significantly worse in the bottom 3.   

I was not cherry picking, I was just breaking up the data into the top 5 of the conference vs. the bottom 3 (over the past 3 seasons), which I believe is valid.  If the top 5 spots, on average, are better in the UAA, than that is still the majority of the conference.  The reason your numbers are different is because 4 years ago, which is not in the last 3 years, the CCIW had better numbers than the UAA.

[Gregory Sager]

Well, you gave four years' worth of data, so four years' worth of data is what I used. And why would three years be a more suitable sample size than four years, anyway? Four years is a good, solid number to use in D3 sports, since it's the timespan of a matriculating class. Heck, I'll go as far back with the numbers as anyone is willing to provide them, because the CCIW only gains by that analysis. But, still, four years is a better sample size than three.

Even if you cancel out 2005-06, the CCIW still comes out ahead in the total numbers. If the UAA comes out very, very slightly ahead at some arbitrary point -- be it the top team, the top three, the top five, whatever -- but the trend isn't borne out over the entire eight teams, then what good is it? As PS said, you're just looking at part of the league rather than the whole league. And the discussion is about how whole leagues stack up against each other.

[Hugenerd]

Well, you gave four years' worth of data, so four years' worth of data is what I used. And why would three years be a more suitable sample size than four years, anyway? Four years is a good, solid number to use in D3 sports, since it's the timespan of a matriculating class. Heck, I'll go as far back with the numbers as anyone is willing to provide them, because the CCIW only gains by that analysis. But, still, four years is a better sample size than three.

Even if you cancel out 2005-06, the CCIW still comes out ahead in the total numbers. If the UAA comes out very, very slightly ahead at some arbitrary point -- be it the top team, the top three, the top five, whatever -- but the trend isn't borne out over the entire eight teams, then what good is it? As PS said, you're just looking at part of the league rather than the whole league. And the discussion is about how whole leagues stack up against each other.

I reported the data as far back as I could, but Massey only has four years prior in d3 (d1 goes back further).  I stuck with 3 years for my analysis because that is what I have been using for the last several dozen posts.  And again, the overall numbers are ahead for the CCIW because of the bottom 3 teams, the top 5 for the UAA almost make up the complete difference (the total difference is 90 total, or 30 per year, or less than 4 spots per team).  Using your own metric, which is matching up the 1's, the 2's. etc.  The UAA is ahead in the top 5 over the past 3 years, while the CCIW is well ahead in the bottom 3.   The question is "which is the better conference" and because we can interpret that differently, I still take the UAA because of their superior strength in the majority of spots in the conference.

And some justification for 3 years.  Although a class is 4 years, even a spectacular freshman class doesnt usually contribute in its first year. For example, if we take the case of last year's POY Jimmy Bartolotta, he averaged only about 10 ppg in his freshman year, before averaging 21, 23 and 28 ppg, respectively.  And regardless of whether you "buy" that or not, I was using 3 years initially, and I will continue to use it now.  And as was predicted by many, including myself, absolutely no head-way in this argument.

[Gregory Sager]

I reported the data as far back as I could, but Massey only has four years prior in d3 (d1 goes back further).  I stuck with 3 years for my analysis because that is what I have been using for the last several dozen posts.  And again, the overall numbers are ahead for the CCIW because of the bottom 3 teams, the top 5 for the UAA almost make up the complete difference (the total difference is 90 total, or 30 per year, or less than 4 spots per team).  Using your own metric, which is matching up the 1's, the 2's. etc.  The UAA is ahead in the top 5 over the past 3 years, while the CCIW is well ahead in the bottom 3.   The question is "which is the better conference" and because we can interpret that differently, I still take the UAA because of their superior strength in the majority of spots in the conference.

Again, though, you're cherry-picking the data. You're not using all of the data available. And even if one concedes to your insistence upon using the last three years (rather than the last four, or last year all by itself -- it's just as arbitrary to use the last three years as the last year, or the last two years), your reasoning that the UAA wins five slots to three really doesn't follow. If the edge is barely discernible in the UAA's direction in five spots, while it's a gulf in the CCIW's direction in the other three, that trumps the whole slot-by-slot thing. Why? Because the way I'm using the data examines the league as a whole, not as a collection of eight disparate teams that are not interrelated. In other words, however good the slot-by-slot method may be, the total points / mean average method is better.

I'm not sorry that I originally proposed the slot-by-slot method, mind you. But part of wisdom lies in realizing when it's time to get rid of your perfectly good mousetrap because a better one's now on the market.

I reported the data as far back as I could, but Massey only has four years prior in d3 (d1 goes back further).  I stuck with 3 years for my analysis because that is what I have been using for the last several dozen posts.

OK, I can buy the idea that your motives were pure. But why continue to draw that arbitrary line in that one exact spot, when the line can either be drawn to cover last year only or by using the full four years of available data?

And again, the overall numbers are ahead for the CCIW because of the bottom 3 teams, the top 5 for the UAA almost make up the complete difference (the total difference is 90 total, or 30 per year, or less than 4 spots per team).

They still don't make up the difference, however, even though you are using the numbers to the UAA's optimal advantage by disregarding 2005-06.

Using your own metric, which is matching up the 1's, the 2's. etc.  The UAA is ahead in the top 5 over the past 3 years, while the CCIW is well ahead in the bottom 3.

Yes, that was my metric. But this is a better one, again because it: a) examines the leagues as a whole; and b) puts them into the greater numerical context of all four-year college basketball: D1, D2, D3, NAIA-1, NAIA-2, USCAA, and NCCAA. It's one thing to say that Team A from League 1 is better than Team X from League 2, while Team Y from League 2 is better than Team B from League 1. But if there's a qualitative difference there -- if the edge of Team A over Team X is miniscule, while the edge of Team Y over Team B is vast -- then that needs to come out in a good analysis, if the data to determine such a thing is available.

As I said, however good the slot-by-slot thing is, the total points / mean average thing is better.

And some justification for 3 years.  Although a class is 4 years, even a spectacular freshman class doesnt usually contribute in its first year. For example, if we take the case of last year's POY Jimmy Bartolotta, he averaged only about 10 ppg in his freshman year, before averaging 21, 23 and 28 ppg, respectively.

Aw, c'mon, you're really reaching with that one. For one thing, this discussion is not about examining which league had the better class of '09. Four years is simply a convenient block of time for data-gathering purposes because it's a familiar collegiate measurement.

I'm not a statistics prof (although we are blessed to have one in our midst ), but I'm pretty sure that it's axiomatic in statistics discussions that, for comparative purposes, the more data one can use to make comparisons, the better. I just don't see the point of using three years when we have four years' worth of data -- and "I'm using three years because that's what I started with" doesn't really wash. That's especially true when you're the one who provided the four years' worth of data.

And as was predicted by many, including myself, absolutely no head-way in this argument.

Well, I give you credit for both obstinence and persistence , but you're really stretching to make your point. I'm not really stretching to make my point at all. The data you've provided is on my side.

[Mr. Ypsi]

I reported the data as far back as I could, but Massey only has four years prior in d3 (d1 goes back further).  I stuck with 3 years for my analysis because that is what I have been using for the last several dozen posts.  And again, the overall numbers are ahead for the CCIW because of the bottom 3 teams, the top 5 for the UAA almost make up the complete difference (the total difference is 90 total, or 30 per year, or less than 4 spots per team).  Using your own metric, which is matching up the 1's, the 2's. etc.  The UAA is ahead in the top 5 over the past 3 years, while the CCIW is well ahead in the bottom 3.   The question is "which is the better conference" and because we can interpret that differently, I still take the UAA because of their superior strength in the majority of spots in the conference.

Again, though, you're cherry-picking the data. You're not using all of the data available. And even if one concedes to your insistence upon using the last three years (rather than the last four, or last year all by itself -- it's just as arbitrary to use the last three years as the last year, or the last two years), your reasoning that the UAA wins five slots to three really doesn't follow. If the edge is barely discernible in the UAA's direction in five spots, while it's a gulf in the CCIW's direction in the other three, that trumps the whole slot-by-slot thing. Why? Because the way I'm using the data examines the league as a whole, not as a collection of eight disparate teams that are not interrelated. In other words, however good the slot-by-slot method may be, the total points / mean average method is better.

I'm not sorry that I originally proposed the slot-by-slot method, mind you. But part of wisdom lies in realizing when it's time to get rid of your perfectly good mousetrap because a better one's now on the market.

I reported the data as far back as I could, but Massey only has four years prior in d3 (d1 goes back further).  I stuck with 3 years for my analysis because that is what I have been using for the last several dozen posts.

OK, I can buy the idea that your motives were pure. But why continue to draw that arbitrary line in that one exact spot, when the line can either be drawn to cover last year only or by using the full four years of available data?

And again, the overall numbers are ahead for the CCIW because of the bottom 3 teams, the top 5 for the UAA almost make up the complete difference (the total difference is 90 total, or 30 per year, or less than 4 spots per team).

They still don't make up the difference, however, even though you are using the numbers to the UAA's optimal advantage by disregarding 2005-06.

Using your own metric, which is matching up the 1's, the 2's. etc.  The UAA is ahead in the top 5 over the past 3 years, while the CCIW is well ahead in the bottom 3.

Yes, that was my metric. But this is a better one, again because it: a) examines the leagues as a whole; and b) puts them into the greater numerical context of all four-year college basketball: D1, D2, D3, NAIA-1, NAIA-2, USCAA, and NCCAA. It's one thing to say that Team A from League 1 is better than Team X from League 2, while Team Y from League 2 is better than Team B from League 1. But if there's a qualitative difference there -- if the edge of Team A over Team X is miniscule, while the edge of Team Y over Team B is vast -- then that needs to come out in a good analysis, if the data to determine such a thing is available.

As I said, however good the slot-by-slot thing is, the total points / mean average thing is better.

And some justification for 3 years.  Although a class is 4 years, even a spectacular freshman class doesnt usually contribute in its first year. For example, if we take the case of last year's POY Jimmy Bartolotta, he averaged only about 10 ppg in his freshman year, before averaging 21, 23 and 28 ppg, respectively.

Aw, c'mon, you're really reaching with that one. For one thing, this discussion is not about examining which league had the better class of '09. Four years is simply a convenient block of time for data-gathering purposes because it's a familiar collegiate measurement.

I'm not a statistics prof (although we are blessed to have one in our midst ;)), but I'm pretty sure that it's axiomatic in statistics discussions that, for comparative purposes, the more data one can use to make comparisons, the better. I just don't see the point of using three years when we have four years' worth of data -- and "I'm using three years because that's what I started with" doesn't really wash. That's especially true when you're the one who provided the four years' worth of data.

And as was predicted by many, including myself, absolutely no head-way in this argument.

Well, I give you credit for both obstinence and persistence , but you're really stretching to make your point. I'm not really stretching to make my point at all. The data you've provided is on my side.

You rang, sir?!

Yes, as long as it is accurate (Massey ain't perfect, but it's probably the best we've got) and relevant (and I can't think of any reason why 3 years is better than 4). more is better than less.

(Of course, if the question were 'is the CCIW or UAA better NOW', none of this would be relevant and we could go to reading tea leaves! )

I kinda like the idea of going back MORE than four years; historically we wouldn't even be having this discussion (though that would take away the fun!).

Hugenerd is correct that this is not really going anywhere (obviously because he is unwilling to yield ), but I'll stick with my position that lately (3-4 years?) the top three conferences are WIAC, CCIW, UAA (and I think at least one of those years #1 was CCIW), in no clear order - with the proviso that over the last 5-15 years it is WIAC, then CCIW, then OAC or UAA or NESCAC (ODAC is coming on strong [especially because of Strong ], but wouldn't make the top five for the longer period).

[Hugenerd]

In all honesty, I think what would provide really compelling data is if we had the rankings only for d3 and did this type of averaging.  Since we are trying to compare among d3 conferences, it would make sense that we only use d3 massey rankings (it shouldnt matter if schools from all these other divisions are ranked among the d3 schools we care about).  Why does it matter how good the D1, D2, NAIA-1, NAIA-2, USCAA, and NCCAA schools are compared to d3?  I was under the impression that we were comparing only d3 schools.  I am sure this will not help your argument, because it will diminish the gap between the top 5 and the bottom 3, so I am sure you will find an escuse as to why the other method is better.  Either way, unfortunately, when I looked at the website, it was not readily possible to sort the data in terms of division (that is why I used overall ranking only to compare level by level, ie first spot, second spot, etc., because the level by level would be unchanged by having the other teams sorted in).  If anyone wants to take the lead on this, that would be great, but it seems like it would take a whole lot of time and may not be worth it.

I honestly dont think that comparing only d3 schools is a "stretch."  The other schools from the other divisions are irrelevant to this argument.

[sac]

Discussing the latest bumper crop of Tart Cherries is becoming more and more appealing than carrying on this discussion which has gone nowhere.

Greg, I believe Michigan leads the nation in cherry production, too!

I do have to once again correct Mr. Y, sort of.........Michigan leads the nation in Tart Cherry production for certain.  Just over 90% of the cherry crop in Michigan is of the Tart variety.....aprox 230 million lbs.  (thats about 210 million more than #2 Washington or 80% of US production, we simply dominate)    Without Tart cherries we are only 4th in the Cherry pecking order, behind  Washington, California and Oregon.   So with  Tart Cherries we are #1, consider sweet cherries our Case Western or Emory........or Alma.


Oh but things are changing.........
http://www.freshplaza.com/news_detail.asp?id=48962

-----------------
sidenote.............Amazingly I can file this under my 'learn something new everyday' category...........the State of Utah is third in Tart Cherry production........and that is one of the many reason I love the internet.

[sac]

Take a deep breath

http://www.youtube.com/watch?v=OWzJFiAbi98&feature=player_embedded

[KnightSlappy]

In all honesty, I think what would provide really compelling data is if we had the rankings only for d3 and did this type of averaging.  Since we are trying to compare among d3 conferences, it would make sense that we only use d3 massey rankings (it shouldnt matter if schools from all these other divisions are ranked among the d3 schools we care about).  Why does it matter how good the D1, D2, NAIA-1, NAIA-2, USCAA, and NCCAA schools are compared to d3?  I was under the impression that we were comparing only d3 schools.  I am sure this will not help your argument, because it will diminish the gap between the top 5 and the bottom 3, so I am sure you will find an escuse as to why the other method is better.  Either way, unfortunately, when I looked at the website, it was not readily possible to sort the data in terms of division (that is why I used overall ranking only to compare level by level, ie first spot, second spot, etc., because the level by level would be unchanged by having the other teams sorted in).  If anyone wants to take the lead on this, that would be great, but it seems like it would take a whole lot of time and may not be worth it.

I honestly dont think that comparing only d3 schools is a "stretch."  The other schools from the other divisions are irrelevant to this argument.

Wouldn't it be better to use the 'rating' category, rather than the rank?

[Hugenerd]

In all honesty, I think what would provide really compelling data is if we had the rankings only for d3 and did this type of averaging.  Since we are trying to compare among d3 conferences, it would make sense that we only use d3 massey rankings (it shouldnt matter if schools from all these other divisions are ranked among the d3 schools we care about).  Why does it matter how good the D1, D2, NAIA-1, NAIA-2, USCAA, and NCCAA schools are compared to d3?  I was under the impression that we were comparing only d3 schools.  I am sure this will not help your argument, because it will diminish the gap between the top 5 and the bottom 3, so I am sure you will find an escuse as to why the other method is better.  Either way, unfortunately, when I looked at the website, it was not readily possible to sort the data in terms of division (that is why I used overall ranking only to compare level by level, ie first spot, second spot, etc., because the level by level would be unchanged by having the other teams sorted in).  If anyone wants to take the lead on this, that would be great, but it seems like it would take a whole lot of time and may not be worth it.

I honestly dont think that comparing only d3 schools is a "stretch."  The other schools from the other divisions are irrelevant to this argument.

Wouldn't it be better to use the 'rating' category, rather than the rank?

Yes, it probably would be.  The reason I originally used rank was because I was only comparing the first team in each conference, 2nd team, etc. so rating was unnecesary.  However, when you start averaging these numbers, the rating is likely a more fitting metric than averaging ranks.

**It is also not possible to do this currently because, for some reason, the massey site does not list the ratings of the teams for 2008, but only ranks them if you click the 'csv' link.