Talk:2011 National Invitation Tournament

Team rankings
Sorry but I strongly disagree on your decision to replace Sagarin's rankings with the RPI. Sagarin is highly reputable, and you can find evidence of it by reading Jeff Sagarin's Wikipedia article (which is linked in the NIT'11 page). The RPI is actually an extremely simple formula which the committee uses mostly to categorize a team's record (for example W/L record vs RPI-top-50, etc). It never has been intended to rank teams. If you want to keep both the RPI and Sagarin I have no problem with it, though. Here is a direct quote from Jeff Sagarin's article:
 * "Jeff Sagarin is an American sports statistician well-known for his development of a methodology for ranking and rating sports teams in a variety of sports. His ratings have been a regular feature in the USA Today sports section since 1985, have been used by the NCAA Tournament Selection Committee to help determine the participants in the NCAA Men's Division I Basketball Championship tournament since 1984, and have been a part of the Bowl Championship Series since its inception in 1998."

There is strong evidence of reputability there. Now, if you want to question the Sagarin article itself, that is a possibility, but on the basis of what is written there, considering Sagarin a reputable ranking is quite justified. Then, again, ADDING the RPI might be OK, also. — Preceding unsigned comment added by Virgrod (talk • contribs) 12:33, 15 March 2011 (UTC)
 * You can't quote a Wikipedia article to back up your assertion, you need an outside reliable source. You also did not address the other two rankings that were added.  Whatever people may think about the RPI, it is the only statistical ranking recognized and produced by the NCAA.  The goal of these articles is not to push a particular point of view (POV), but to say, without citation, that something is reputable is exactly that.  If you want to discuss this further, I suggest starting a disucssion on at the college basketball Wikiproject in order to get a community consensus on this issue.  Thanks. - Masonpatriot (talk) 15:08, 15 March 2011 (UTC)

So, you dispute that "His ratings have been a regular feature in the USA Today sports section since 1985"? That is an easily verifiable fact...do you accept it? If you do, you must accept reputability, because it isn't that easy to keep a spot on a major national newspaper for over 25 years running... over that length of time, if his ratings were flawed or tendentious in any way it had been already found out, and removed... Furthermore, do you dispute that his ratings have been "part of the Bowl Championship Series since its inception in 1998"? If you dispute it, we can try to verify it... if you do not dispute it then that is further evidence of reputability considering how important and how much money there is in the BCS. I do not have a problem with you ADDING the RPI. I believe you ARE justified to include the RPI, as trivial as it is by design, because it is indeed used somehow... ... but it seems your intention is to REMOVE Sagarin. I equally strongly believe that you are NOT justified to remove Sagarin. Removing Sagarin would be justified only if there was no evidence that Sagarin's work is reputable...as I mentioned above there is tangible evidence of reputability that can be verified. I do not see why any debate on this issue couldn't be conducted right here. It seems to me that this is the very purpose of the discussion section. I will for the time being restore Sagarin, and encorage you to ADD the RPI if you see fit. — Preceding unsigned comment added by Virgrod (talk • contribs) 16:12, 15 March 2011 (UTC)
 * Keep in mind that you are linking to an article about a PERSON, not his rankings, while the RPI actually has a WP page (and is actually NCAA sponsored). I will bring this up at the College Basketball Wikiproject to see if we can get some other voices here.  On Wikipedia we move forward by establishing consensus, not by pushing an individual POV. - Masonpatriot (talk) 16:21, 15 March 2011 (UTC)

I respectfully suggest that it is you who is pushing a POV... you apparently do not see value in Sagarin's work, or possibly consider it tendentious of flawed for some reason, which is fine with me as a personal opinion... there is however tangible evidence that indicates that Sagarin's work is in fact highly appreciated and respected by the sports community, and this makes it a reputable source, as have been mentioned above... not the only source, though, but a reputable one for sure. The reference to his page is because it contains the necessary information... the person who made the page could have easily called the page Sagaring's ratings instead, since most of the material there is about his work. Any discussion on any board should not be framed as an either or proposition (RPI OR Sagarin). There is no reason why both the RPI and Sagarin could not coexist. Both ratings can be justified for a reason or another. So, this argument exists only because apparently you feel that Sagarin's work is not worthy of inclusion (which I dispute). --Virgrod (talk) 17:01, 15 March 2011 (UTC)
 * IMO, the article should only list the rankings that are used by the NIT selection committee to select teams. My quick look is that they do not list any computer rankings used to make selections.  Therefore, I don't think this article should list any rankings.  The only one I'd leave is RPI, since that is used by the NCAA basketball selection committee.
 * It's original research and synthesis to list computer rankings that are not used by the selection committee. — X96lee15 (talk) 17:02, 15 March 2011 (UTC)

Sorry, but why is it necessary that the rating be officially used by the selection committee in order for them to be included in this article? This in an information article and the criterion should be whether the information given is relevant and reasonably reliable/credible. Possible users of the ratings include people who may want to predict the outcomes, or estimate the best match-ups, as well as getting a feeling for how fair the selection may have been. For any such use Sagarin ratings could be useful, provided they are reputable/authoritative (meaning they are not just the product of a hobbyist playing with a computer in his basement). There is actual evidence that Sagarin's ratings are viewed as authoritative by professional people who should know, such as the sports editors of a major national newspaper, or those running the very important BCS selection, who have been trusting the Sagarin ratings for decades now. Notice also that even if the comittee does not "officially" use these ratings, individual members are free to read newspapers, check websites, etc and the information they gather do influence how they perceive a team and ultimately vote.--Virgrod (talk) 17:39, 15 March 2011 (UTC)
 * You are turning this article into something that it's not. This article is about this year's NIT... that's all.  RPI is sponsored/sanction by the NCAA and is used in the selection for NCAA/NIT tourney slots (the CIT states that they use it as well).  My arguement has NOTHING to do with whether or not the Sagarin ratings are any good... I don't care if they are, because it's not relevant to the NIT.  Keep in mind that you link to a page about Jeff Sagarin, NOT his ratings, which do not have a WP article.  You also do not provide any support for the other two ratings that are listed, just Sagarin.  This article is just about the 2011 NIT, not about "Possible users of the ratings include people who may want to predict the outcomes, or estimate the best match-ups, as well as getting a feeling for how fair the selection may have been."  WP is not an engine for editorial statements and conjecture.  Like X96lee15 stated, your position relies on original research, and this discussion is verging on WP:OWN.  As a new editor and contributor, I encourage you to review those two Wikipedia polices.  Thanks. - Masonpatriot (talk) 17:49, 15 March 2011 (UTC)

The other ratings have a link to their corresponding Wipedia page, where an interested user can find further information on them. Elo's ratings have been used in the Chess world for almost a century, and Pomeroy's ratings are widely cited (see their page) as a simple google search can confirm. That is why we use links, to send the reader to other parts of the wikipedia where relevant information has already been given, so we do not have to keep on repeating the same information. I do not get at all the "original research" claim...The Sagarin ratings have been widely available for decades, and can be easily found in the appropriate locations (linked in the JS article). If those pages have deficiencies (they may well have) then they should be questioned there, not here. The ratings cited are only for the specific teams that are playing in the NIT'11 and that makes them relevant to the NIT'11 article... whether the committee used these ratings to make the selection or not. It is no different than posting each team W/L record, or their average attendance figure, etc... It is relevant information about the teams playing in the tournament. By the way, JS wikipedia article states that JS's ratings are used by the NCAA. I have not yet found direct evidence of that, for which I am not emphasizing it, but the USA News and the BCS use is well known an easily verifiable. --Virgrod (talk) 18:16, 15 March 2011 (UTC)
 * It's original research for a couple of reasons, and I will cite from the policy what X96lee15 was referencing:
 * "Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources. If one reliable source says A, and another reliable source says B, do not join A and B together to imply a conclusion C that is not mentioned by either of the sources. This would be a synthesis of published material to advance a new position, which is original research.[5] "A and B, therefore C" is acceptable only if a reliable source has published the same argument in relation to the topic of the article."
 * Your argument fails even that because nothing you've cited is backed by reliable sources, all you've cited are other Wikipedia pages (of which Sagarin's page is completely unsourced and has now been tagged as such and Pomeroy's is on the verge of being deleted due to notability concerns). Additionally, you have not addressed the fact that your arguments are about the validity of various computer rankings, and that they have nothing to do with the NIT (which is what the article is supposed to be about).  As stated above by me and X96lee15, the RPI is the only ranking officially used by the NCAA (and is PRODUCED by the NCAA).  Other wikipolicies that are relevant include WP:Verifiability:
 * "The threshold for inclusion in Wikipedia is verifiability, not truth. The policy says that all material challenged or likely to be challenged, including quotations, needs a reliable source..."
 * ...and WP:NPOV:
 * "An article should not give undue weight to any aspects of the subject but should strive to treat each aspect with a weight appropriate to its significance to the subject... Keep in mind that, in determining proper weight, we consider a viewpoint's prevalence in reliable sources, not its prevalence among Wikipedia editors or the general public."
 * By piling on statistical rankings, you are placing undue weight on such matters in an article that is supposed to be about the 2011 NIT. At this point I agree with X96lee15 that we whould not have any statistical rankings aside from the RPI listed.  I plan on making this change to the article, but prior to that I suggest taking this discussion to the college basketball Wikiproject, where a section on this has already started.  - Masonpatriot (talk) 19:15, 15 March 2011 (UTC)
 * Going briefly off topic, thanks for pointing out Pomeroy's article. I've added a few source to at least try to make it non-PROD-able, though I don't know that it would survive a deletion discussion in its current state. cmadler (talk) 14:16, 16 March 2011 (UTC)

Here are a couple of relevant links (these should be added to the JS article): Official BCS page confirms use of Sagarin (and others). This is a recent technical paper (already published in a peer-reviewed academic journal from the Berkeley Electronic press) written by a Univ of Michigan researcher which _explicitly_ indicates that the NCAA selection committee (at least at the time of the article's writing, 2006) use Sagarin's ratings. In section 2.2 it says:
 * The NCAA selection committee also uses another objective rating method to help distinguish between teams. Jeff Sagarin, who graduated with a degree in mathematics from the Massachusetts Institute of Technology and an MBA from Indiana University, has been supplying the selection committee with his unique computer ratings of the teams since 1984. In short, his overall ratings for NCAA basketball teams are based on a synthesis of two different ratings: Sagarin’s personal modification of the chess rating system developed by Elo (1978), which only takes wins and losses into account, and a rating method developed by Sagarin known as the PURE POINTS method, which takes a team’s scoring performance into account. It is worth  mentioning here that Sagarin’s computer ratings are respected enough by the NCAA to play a role in the calculation of the now infamous BCS computer ratings in  college football.

I found this paper by performing a google scholar search (this only searches "scholarly" sources) and obtained 114 hits, which suggests that these ratings are indeed taken seriously even by sports researchers. I feel that the conclusion that these ratings are authoritative, or at least that they are considered authoritative by people well positioned to make such assessment is inescapable.--Virgrod (talk) 19:32, 15 March 2011 (UTC)
 * Again, that is all well and good, and completely relevant to Jeff Sagarin (and I honestly hope you improve his WP article), but it is still irrelevant to the 2011 NIT. Brining up the BCS is not persuasive as the football and basketball selection processes are completely different.  Honestly, I'm not sure why this is such a sticking point.  If you feel this strongly, I encourage you to give the same treatment to 2011 NCAA Men's Division I Basketball Tournament and you will likely get the same response from other editors.  Wikipedia is not just a collection of statistics, especially those that are not directly related to how teams are selected for this tournament (which have NOT established).  What people think about Jeff Sagarin is completely irrelevant.  Again, my plan remains to edit this page to reflect my concerns, and am curious why you are unwilling to bring more editors into this process. Masonpatriot (talk) 20:37, 15 March 2011 (UTC)

Sorry but the citation from the published paper right there written by a Univ. of Michigan researcher (which is peer-reviewed; i.e., people who are supposed to know how to do research, document sources, etc have carefully read the paper and accepted its content) indicates that the Sagarin ratings ARE used by the NCAA committee. Now you can argue that it say nothing about the NIT committee, however the NIT is now an NCAA event, and there is no reason to believe that the NCAA would use it for one event and not for the other very similar one. The BCS is relevant to this debate simply because it helps establish the reputability of the source... which was initially questioned. A rating system used by an organization like BCS -- which also has been published for over 25 years by a major national newspaper, which also has been referenced over 100 times in the scientific literature -- IS reputable. Secondly, and as I have already indicated, the ratings given are for the teams in the tournament, and may have been used in their selection (although this is not critical). They are an important piece of information about the specific teams that are playing in this tournament. I find puzzling that so much effort need to be invested simply to include a widely known and accepted reputable measure about the teams performance. I find immensely puzzling that you do not find acceptable to simply ADD the RPI (perfectly alright with me), and then allow the reader to choose which of the two numbers s/he finds more credible/useful (if any).--Virgrod (talk) 21:37, 15 March 2011 (UTC)
 * First, I understand what peer review means. Second, you fail to address (again) the underlying point of why this at all matters in an article about the 2011 NIT.  All your arguments do is bolster the fact that Jeff Sagarin's article could be much better (even though you speak as though you've reached a forgone conslusion without the burden of providing citations... citations also need to include why something is relevant to THIS article).  Consensus on college basketball tournament articles currently does not include this data within the article (even the RPI).  Your suggestion may carry the day, but again, it is not YOUR decision, it is something that is decided by consensus, which is why I suggested opening this discussion up.  The fact that you dismissed the only other editor that has chimed in is dishartening.  I understand your perspective, but Wikipedia is governed by WP policies (of which I've cited many, and those arguments also haven't been addressed), and no one WP:OWNs any article.  At this point, I think I'm going to involve some of the WP dispute resolution resources, since we'll just go back and forth. - Masonpatriot (talk) 00:46, 16 March 2011 (UTC)


 * Comment - I am with X96lee15 in that I don't think this article needs any statistical ratings. Frankly, people aren't coming to an article about the 2011 NIT to see what RPI/Sagarin/Pomeroy/etc. ratings the participants had.  It's basically irrelevant.  What people will come here to see is who the participants were, what the results are, major storylines, etc.  The ratings don't add anything and neither does the bickering over them.  This is turning into a Wikipedia circle jerk at its best/worst.  Rikster2 (talk) 01:19, 16 March 2011 (UTC)
 * Comment - At this point, I agree, and it is more consistent with existing articles. Your last comment hit the nail on the head... too bad I let myself get sucked in! - Masonpatriot (talk) 01:31, 16 March 2011 (UTC)

Can you please explain to me why my quotation from the Univ. of Michigan published paper fails to address the relevance issue:
 * The NCAA selection committee also uses another objective rating method to help distinguish between teams. Jeff Sagarin, who graduated with a degree in mathematics from the Massachusetts Institute of Technology and an MBA from Indiana University, has been supplying the selection committee with his unique computer ratings of the teams since 1984.

I have also added in addition to that very relevant and explicit quotation the argument that the Sagarin rating given are a performance indicator of the specific teams involved in this specific tournament and hence can be useful information to the reader. We appear to have fallen into a "moving goalpost" situation. First you questioned reputability, which is rather surprising since the material in question have been included for over 25 years running in a major newspaper. Then you demanded references, rejecting the relevant Wikipedia articles on this subject as inadequate. You also argued that only the RPI was used by the committee and hence only the RPI could be included in this article. I have addressed all those points including providing references to official websites, and to the Univ. of Michigan scientific paper, and mentioning that over 100 scientific publications cite the Sagarin ratings. So, frankly it appears that you simply made up your mind long ago, and no amount of evidence or argument will convince you... when I address an objection, you'd simply find new objections. I honestly fail to see why all this is helpful or necessary. --Virgrod (talk) 02:45, 16 March 2011 (UTC)

Arbitrary section break
I noticed that you already decided that this issue has been "settled", apparently because 2 people joined your position. I very strongly disagree with you on this matter, and find your overall approach of shifting from a line of objection to another to another as the previous ones were addressed in an effort to ultimately uphold your original position at all cost to be non-constructive and unhelpful to the general objective of this organization. I have provided a great deal of information in addressing each of your various and evolving concerns. I am still waiting for the answer to my question above on how the quotation from the scientific paper does not address the relevance issue. Rather than answering, you simply declare your own position as the "winner". You indicated that no one should behave as the owner of an article, with which I agree. Yet your approach of imposing your point of view at all cost, by constantly shifting from a rationale to another to another seems consistent with someone who believes to own this article.--Virgrod (talk) 12:31, 16 March 2011 (UTC)


 * While the Berkely Electronic Press article does confirm that the NCAA selection committee uses Sagarin's rankings (or at least, they did from 1984 through 2005 or 2006, which generally predates the time when the NCAA began operating the NIT), there are several problems with using that as the basis for including Sagarin's rankings in this article. First, as far as I can tell, although the NCAA now operates the NIT, there is a separate selection committee for each tournament. Second, IF the NCAA selection committee is now picking the NIT, we have no source tying Sagarin's rankings to the 2011 NIT. Third, IF Sagarin's rankings were used in picking the 2011 NIT, unless things were done in a meaningfully different manner than other years, that should probably be covered at National Invitation Tournament. Finally (at least for now), IF Sagarin's rankings were used in picking the 2011 NIT, as far as I can tell they were one of many such pieces of information -- the committee can also consider RPI, overall record, conference record, schedule strength, recent versus early-season performance, and on and on -- and listing out the rankings for each team would give that criterion undue weight in this article. cmadler (talk) 14:13, 16 March 2011 (UTC)

Hi, Adler: I do find significant merit in your arguments. But let me first clarify the source issue. The author of the paper mentioned above published another (peer-reviewed) follow-up paper in the same journal 2 years later where he makes an identical statement about Sagarin (same section): [http://www.bepress.com/jqas/vol4/iss2/8/ West, Brady T. (2008) "A Simple and Flexible Rating Method for Predicting Success in the NCAA Basketball Tournament: Updated Results from 2007," Journal of Quantitative Analysis in Sports: Vol. 4 : Iss. 2, Article 8. DOI: 10.2202/1559-0410.1099]. This suggests that the statement continued to be true at least for 2 additional years after the first paper was published, and there is no evidence that indicates it has stopped being true since.

Nevertheless, we should all agree that there is no way for anyone to force a committee member to consider or ignore a criterion. Whatever the NCAA guidelines indicate, a committee member, as a person, has his/her own mind, and can use whichever (combination of) criteria s/he considers fit to reach his/her decision on specific teams (just the RPI, just Sagarin, just winning percentage, his/her own personal ranking system, a combination of many things, etc. etc.). So, ultimately we cannot assess the degree to which Sagarin, the RPI or any particular factor was used to allocate at-large bids for the NIT or the NCAA tournaments (unless we ask each member, and s/he chooses to give us a truthful answer).

On the other hand, we may perhaps agree that providing *some* objective team "performance indicator" (not necessarily Sagarin or the RPI) for the teams in this and other similar tournaments is sensible, consistent with policies, and more importantly useful to the reader, for whom we are ultimately doing all this. The need for this is easily seen by noticing that the only indicator of team performance or strength currently given is the win-loss record. But the W/L record (or winning percentage) is almost useless information when several teams have played vastly different schedules, as is the case in NIT'11 and similar tournaments. Each team played a different schedule, and nearly all played in different conferences, so how each did in its own particular schedule is almost entirely irrelevant.

So, providing our reader, for his/her convenience, at least 1-2 objective, reputable, authoritative performance indicators that can be objectively compared across teams is useful information to the reader. If you do agree on this premise then we could move the debate to try to determine _which_ is/are the most sensible such performance indicator(s), for our purpose (and, again, Sagarin adn the RPI need NOT be included).

I look forward to further input by you and others. --Virgrod (talk) 16:04, 16 March 2011 (UTC)


 * That's all well and good, but I don't necessarily agree that additional "performance indicators" are needed in this article. In the long run, we have three performance indicators in this article: W/L record (coupled with conference, which makes up over half the season and gives some indication as to schedule strength), seed in this tournament, and (in a couple weeks) performance in the NIT. Readers desiring additional information about any team can easily click through to an article detailing (among other things), the team's results for the entire season. Further, if we were to agree that some top-to-bottom ranking number needed to be given, picking one would be tremendously problematic. With top-25 rankings, the Coaches' Poll and the AP Poll are clearly the dominant rankings, and we can safely use one or both without any problem. But with top-to-bottom 345-team rankings, there are a whole bunch of options (Pomeroy, Sagarin, RPI, etc.), none of which are necessarily dominant. cmadler (talk) 17:18, 16 March 2011 (UTC)

I agree that in the long run we may not need any indicator, since we will have known the results already (although some may still find interesting/useful to look retrospectively at the relationship between the performance indicator(s) and the results). Nevertheless, in the short run having at least one objective indicator that (approximately) mean the same for all teams -- unlike the W/L record which is totally meaningless -- would seem useful to some readers, if not all... I can only speak for myself, but from my years following sports and participating in sport forums I know that sports fans do like to get a feeling for the disparity between teams, and to project/speculate on results, etc. Obviously, a fan can find *ALL* the information on this page elsewhere (for example the official tournament site would be an excellent source, as well as sites by many reputable media organizations). So, if we accept the criterion that the user can find objective indicators elsewhere, we could as well make a strong case that most of the information on this page is redundant (a link to the official site and/or a couple others reputable media sites might be all is needed). Nevertheless we do have the page, presumably to give the reader useful/convenient information on a singe place. I do acknowledge the challenge of choosing one objective indicator among many, for the reasons you gave, but we cannot assume we will never agree on one or two. For example, some reputable people (e.g., college professors) provide "meta indicators" which simply try to "average" some of the better ones to provide a single indicator. Something like that MIGHT lead to a consensus. So we can frame the question in a different direction: IF we *did* agree on a reasonable objective performance "consensus" indicator, and there is a volunteer (not necessarily you or me) who agrees to provide his/her labor to add the information to the page for the benefit of whoever cares, would any harm be done by adding this information?--Virgrod (talk) 17:52, 16 March 2011 (UTC)

Concrete constructive proposal
There has't been input to this discussion for quite some time. Since the tournament continues, it is obvious that this discussion is time limited. So, I would like to make a concrete, and I hope, also constructive proposal. First, I accept the premise that singling out Jeff Sagarin's rankings for inclusion over the other alternatives is NOT justified. Even though there is tangible evidence that Sagarin's sport ranking system has the long-running backing of notable institutions (e.g., USA Today, and the BCS National Championship Game), has been widely cited by scientific publications, and has been in recent years taken into account by NCAA tournament selection committees, still it is difficult to conclusively establish that it is "the best" or most authoritative such system.

My proposal is to include a median ranking calculated by Carson-Newman College's Mathematics faculty member Kenneth Massey and provided here. Notice that I do NOT propose to use Massey's own sport ranking system, which is respected in its own right (for example, it is also utilised by the BCS National Championship Game). What I proposed is the median ranking that he calculates, after considering the results of nearly 50 Sports rating system.

Notice that no claim is made that the median ranking is explicitly used by the NCAA. As indicated above, each committee member can reach his/her decision by considering any number of factors, whether or not they are explicitly listed by the NCAA. The purpose of the median ranking is simply to allow the interested reader to get a quick feel for the "strength" of the teams involved, and the possible disparity between them, without having to do a lot of research. For reasons discussed above, the W/L records are nearly useless for this purpose. This may be quite useful especially to the casual fan. Of course, some readers may find this information useless. But they can simply ignore it.

Finally, I propose to use the median as opposed to the simple statistical mean (basically what Massey calls the "consensus" ranking), because the statistical mean is more susceptible to be influenced by "outliers" (for example the median of 8, 9, 10, 11 and 100 is 10; but the mean ("average") is almost 30. Evidently 10 is more "representative" than 30 for the values in this example. On the other hand, this is an extreme example just to make a point; often the mean and the media are closer to each other)

I do hope that everyone agrees this is a sensible thing to do, which at least a significant percentage of our readers will appreciate, while not bothering too much to the percentage that won't--Virgrod (talk) 00:53, 18 March 2011 (UTC)
 * I appreciate the time you've taken to put this proposal together Virgrod, but I don't think we can go it because it is original research and synthesis of information. — X96lee15 (talk) 01:03, 18 March 2011 (UTC)
 * As an aside, I definitely disagree with the assertion that "this discussion is time limited". There is no deadline, and Wikipedia is a work in progress. This is an encyclopedia, not a newspaper; we don't have to "finish" this article the day the tournament ends. cmadler (talk) 13:19, 18 March 2011 (UTC)

Sorry, but I do disagree that by simply posting the median ranking that Prof. Massey calculates and post -- which is publicly verifiable information -- *we* would be doing any original research or even synthesis of information. Perhaps, Prof. Massey himself is doing original research (which he can and probably should as a college professor). But we would be simply using him as a source. It is no different than quoting material from a book or research article, give proper credit to the source. The original research is done by the authors of the book/article and that is fine, not by the Wikipedia editor that is simply reporting the material in the book/article. Likewise, Prof. Massey computes the median ranking himself and post it for everyone to use (*he* does "information synthesis"). When we report what he calculated, his website becomes a reliable source for us, which we credit. Notice that we do not calculate anything at all, and therefore do neither research nor synthesis ourselves. — Preceding unsigned comment added by Virgrod (talk • contribs) 01:23, 18 March 2011 (UTC)
 * I'm not ready to say that we should definitely use it, but I also don't necessarily think that using Massey's ranking comparison presents an OR or SYN problem for us. My first thought is that it would be akin to the way we cite review aggregators Rotten Tomatoes and Metacritic in film articles. But, if I may extend that comparison for a moment, presenting the rankings here would perhaps be akin to listing Metacritic scores (or, for that matter, box office receipts) next to each nominee's listing in, say the 83rd Academy Awards article. In that case, people wanting to know more about the individual films can click through to those articles, and in this case, I think people wanting to know more about the individual teams can click through also. cmadler (talk) 13:37, 18 March 2011 (UTC)
 * The issue we stubmle upon, though, is by selecting any particular ranking, we are putting undue weight on whichever one is chosen. Since there is no definitive ranking and no one ranking (aside from the RPI) is officially adopted by the NCAA, I think staying with the status quo is the best course.  All the scholarly works on Google don't change that.  I agree that this type of information may be more appropriate on team-specific pages.  - Masonpatriot (talk) 15:19, 18 March 2011 (UTC)

Masonpatriot, I am not sure you understood my proposal... I am NOT proposing to choose any specific ranking, and to put undue weight on it... what I am proposing is simply a statistical figure (the median) of the various reputable rankings which is made available by Prof. Massey. Therefore no undue weight of any kind is been given to any specific ranking... to calculate this statistic, Prof. Massey considers all reputable rankings, nearly 50, that meet certain reasonable criterion...this does include the RPI, the major polls, Sagarin, Massey's own (only one of many), and many others... not one specific ranking is singled out as *the* ranking. I would also like to emphasize that while one cannot prove that the "average" ranking will be interesting to EVERY reader, one would have a much harder time to prove that it will be UN-interesting to *all* or even *most* readers. The fact that major publications (not only USA Today, also the NY Times, and many others) have been publishing rankings for decades, and that so much scientific research has been done on them (with at least one scientific journal entirely devoted to them) confirms that there is a lot of interest in this subject both among the regular public/fans (i.e. our readers) as well as among the academic circles. As indicated also above, the idea of adding a single number that tries to represent a rough estimate of "strength" of the team has nothing to do with which criteria each committee member uses to choose teams (which is impossible to know). It is to give the reader an opportunity to get an idea of team strength without doing much research, clicking, reading etc. However, the reader who wishes to do further research can still do so. I am yet to hear any argument on how including the "average ranking" provided by Prof. Massey would *harm* the page in any way; that is, how it will be a disservice to the reader in any way. That should be the criterion to exclude information: does it harm the content of the page? I once again get a disturbing feeling that some of us have extreme and unreasonably hardened views on this matter, and no amount of reasoning or evidence would make them change. No amount of time or effort anyone expend addressing their concerns would make any difference. When a line of attack seems exhausted or addressed, they simply shift to a brand new one. I once again fail to see why such attitude is helpful to the interest of this organization, and most importantly, the reader.--Virgrod (talk) 15:55, 18 March 2011 (UTC)

On the subject of OR and information synthesis, I agree with Adler that it is a non issue here. It seems some of us are not considering the complete policy statement. There is nothing wrong with including in an article original research (quite a few do). The problem is to include original research "not already published by reliable sources". For example, a person may try to use the WP to publish for the first time a brand new way to rank teams. That would be AGAINST the OR policy. The correct approach for the new ranking proponent is to publish it first in an existing reputable publication (say an academic journal or well established sports periodical or website). Once published, then anyone can refer to it, by properly crediting the source. In our case, the source would the the website of the Carson-Newman College, which seem reasonably reliable for our purpose. With respect to information synthesis, again, some of us are leaving out an important part of the policy. There is NOTHING wrong with combining sources. In fact, most WP articles do so, when the article is based on bits and pieces of multiple sources. The problem is "Synthesis of published material that advances a position"... emphasis on the second half of the sentence. For example, one can post in a WP article an average of the forecast NYC temperature by various weather services. This is clearly information synthesis. But it does NOT violate the policy because such average would serve to advance no specific position or agenda of any kind (2nd part of the sentence above).--Virgrod (talk) 16:19, 18 March 2011 (UTC)


 * Just out of curiousity, can you point to any instances of "major publications" putting team rankings (top-to-bottom rankings, not just top-25) on a tournament bracket (NCAA or NIT)? To answer another point, Wikipedia places the burden on the editor who wants to add information rather than on the editor who wants to exclude information. Rather than asking whether including such a ranking would harm the page, the correct question is whether it helps the page, or more correctly, whether it helps a reader's understanding of the topic. In this case, I think it's possible, but it would need to be more than just listing the numbers next to the teams. For example, assuming sources can be found (I know they could be for the NCAA tourney, I'm guessing they could be for the NIT), the article could mention the highest-ranked teams in the NIT, which probably were on the bubble and just missed the NCAA tourney. It could get into the mix of auto-bids versus at-large bids. There's probably an NIT "bubble" similar to the NCAA bubble, and so it could comment on teams that barely made the NIT, teams that just missed out. It could mention teams that seem to be mis-seeded, either too high or too low. And so on. Assuming that sources can be found, I think there's a very strong case for including that kind of information. But you're talking about just throwing a bunch of numbers into charts/brackets, which is of far less value. cmadler (talk) 16:29, 18 March 2011 (UTC)

I know that USA Today and the NY Times publish complete rankings, but don't know if they put it by/near the brackets... strictly as a matter of reader convenience it seems they should in these cases. I have seen also Vegas lines by/near the brackets, which is somewhat similar to a ranking. In our case the idea was to put the ranking not inside the graphical brackets, but in a table that currently appears above the graphical brackets which lists the tourney teams with some minimal info about each (conference, W/L record, etc). The "average ranking" would be exactly one additional column in that table. I do agree with your suggestions on material that could be included on the general issues of invitations (at-large, autos, bubbles, etc). In fact, I got interested in those rankings for the very sort of reasons you mentioned: some debate on whether some of the auto-qualifiers were "too weak", and whether some of the teams not getting at-large invites were much stronger than those in, etc. I think it is a very interesting angle, and I would be glad to contribute in that general direction, along with others. On the general issue on the "burden of proof" to include/exclude material, I have been providing arguments/evidence for inclusion. But ultimately, a decision need to be made, and I argue that after all the arguments have been considered, the ultimate question is whether or not the additional material would be detrimental; that is, whether the new article with the additional info would be by some reasonable criterion "worse" than it was before.--Virgrod (talk) 17:13, 18 March 2011 (UTC)
 * But this article isn't the proper place for the "debate" you reference. As stated above, the place to discuss that in the section of the NIT article regarding selection process.  Also, I question the long-term necessity of such numbers, because once the tournament happens, we have results and statistical rankings are pretty meaningless.  A teams record is different, because it records events that actually happened.  Also, cmadler explained that the question isn't "the ultimate question is whether or not the additional material would be detrimental," but you just state the opposite as if that makes it so.  Again, I support X96lee15's position that this information constitutes original research and synthesis of information.  - Masonpatriot (talk) 17:31, 18 March 2011 (UTC)

Please read my comments and Adler's above about original research and synthesis of information. I spent sometime writing them and I would appreciate the courtesy. Neither OR nor IS is forbidden. The problem is to include, and I quote (feel free to verify this), material "not already published by reliable sources". By posting an average computed by a college professor and published at the college website, we would be doing nothing that satisfies the OR definition. Likewise, as discussed above (please, please, do read it) information synthesis is indeed allowed. What the policy forbids is (this is a direct quotation): "Synthesis of published material that advances a position"... emphasis on the second half of the sentence. If you believe that by simply posting an average computed by a college professor, and made available at a College website we would be breaking either policy, please, explain HOW we would be breaking either policy, do not just post the link to the policy. Secondly, the proposal to discuss "selection controversies" was actually made by Adler, and I do believe it would be a very interesting angle, and one of current interest to our readers. Furthermore, since the controversies would be about THIS year's selection, and the specific teams involved, this page would indeed be the right place to discuss it. In fact, numerous WP articles do have a section on "controversies" related to the subject of the article. Finally, if you have some reasonable arguments explaining why adding one single column to an existing table to include the Carson-Newman-provided "average ranking" would make the article worse, or would harm the interest of its readers in any way, please, provide those arguments. So far it seems your arguments are of the sort "the reader can find that elsewhere" which is true, but it would also be true about the entire article (brackets, results, w/l records, etc, etc can all be found at the official NCAA site, and at several other reliable places). In that case, we should replace the entire article with one or several external links. In fact, by adding the Carson-Newman "average ranking" column, as well as the selection "controversies" section suggested by Adler, we would be providing at least *something* that the official site does not already provide.--Virgrod (talk) 19:47, 18 March 2011 (UTC)

Point of order
As others have pointed out, as an encyclopaedia article, the article's production may continue way past the point where the tournament has ended. Nevertheless, from a practical viewpoint, it seems fair to say that the material added to the article after the contest has finished will have a lesser readership, and therefore much fewer readers can benefit from this. At that point it may still be useful to readers interested in retrospective/historical views of the tournament (perhaps for school-related work), whereas during the development of the tournament this article may be of interest to the average fan also. On the other hand, we do not necessarily want to rush the debate, and reach the wrong conclusion. But it is undeniable that for as long as the material under discussion is excluded from the article, the numerous readers who may find that material useful cannot use it. And even if we eventually agree that the material should be included, for the majority of the readers it may be too late, since we cannot go back in time. From this point of view, we can at least agree that enough good arguments have been provided above to at least justify to:

TEMPORARILY add exactly one (1) column to the background information table that is above the brackets (NOT to the graphical brackets themselves) to provide the median value of about fifty (50) reasonable rankings that is computed by Carson-Newman College's Mathematics Prof. Kenneth Massey and published here.

The idea is that WHILE we settled this debate, and considering that the contest is in progress, the readers of the article can at least consider Prof. Massey's median of reasonable rankings as they see fit. Those who don't find this average useful can simply ignore it, since it is just one column in one table that contains background team information, and it has no effect on the brackets themselves.

It is understood that he inclusion of this column with the average value computed by Prof. Massey would be WITHOUT PREJUDICE on any decision on the permanence of this information; that is, eventually it may be decided that such material be removed.

Finally, it is emphasized that the median published by Prof. Kenneth Massey is not "yet another ranking" but a statistical figure obtained after considering ALL available rankings (both poll-based and computer-obtained) that meet certain reasonable criteria. No ranking receives favorable or unfavorable treatment of any kind, and all of those mentioned above (RPI, Sagarin, Pomeroy, AP, etc) are considered.Virgrod (talk) 16:55, 20 March 2011 (UTC)


 * The more I think about it, the more I think this discussion should be taking place at Wikipedia talk:WikiProject College Basketball. If this information should be included here, it should also be included at every other such article (NCAA men's & women's, NIT, WNIT, etc. for all years), or if the information is excluded, it should be excluded on every article (which I believe is the status quo). So if, as it seems, you feels strongly about this, I suggest that you raise the issue there as a change to occur on all such articles. cmadler (talk) 15:08, 21 March 2011 (UTC)

Yes, I agree, in principle (it seems that Prof. Massey does not compute "average ranking" for women teams, but perhaps someone else does). I have focused on the M-NIT'11 simply because it so happens that I am following that tournament more closely, because of the specific teams involved in it this year (the NCAA-T includes many more teams also, which means more manual work to insert the data). But clearly my arguments would work equally well (or fail equally) with respect to the other tournaments for which the concerned data may be available. In fact, one could even consider previous years, if the concerned data is available (and it seems to be). But once again, I do hope that in view of the timeliness issues mentioned above, the additional column with the average rankings is allowed to remain online at least while this particular contest is in progress, and while we continue to seek a final consensus.--Virgrod (talk) 03:19, 22 March 2011 (UTC)


 * As noted above, the "timeliness issue" is a red herring. I have reverted the addition of the Massey median ranking until a consensus can be reached through the WikiProject regarding the inclusion of this information. cmadler (talk) 15:05, 22 March 2011 (UTC)

Addler, as indicated in my answer to X96 below, your previous comment did not sound to me as a rejection of my proposal to temporary include Pro. Massey's average of 50 independent rankings while the debate goes on. Nevertheless, if you have a good reason to object, I can understand, even if I do not know what the reason is. As fro moving this discussion to the place you suggest, I do not object, as I indicate above. On the other hand, I do not really know in concrete terms what should we do. Copy/paste the comments above in another place?

A purely technical point is that the format that Wikipedia provides for these discussions is way, way, too rudimentary...we should have a simple message-board-type interface, with threads, individual "posts", basic search, etc. The present system might be OK for a very short discussion, but it is definitely unhelpful for somewhat longer debates. Since there are quite a few FREE and open-source message-board software, it should not be too difficult for the WP technical people to adapt one such software to serve as a foundation to support discussions such as this (basically every WP article should have a basic BBS attached to it, for discussing the article). If you have some connection that could help for this, please, use it, as it may be really useful for everyone. --Virgrod (talk) 03:48, 23 March 2011 (UTC)


 * I think the message-board interface you're looking for is coming eventually, see LiquidThreads. If you'd like to help with that effort, that page will help you get started. As to whether the Massey mean (or any similar ranking) should be listed, my answer remains that I do not believe there is any particular urgency in adding it right this moment, that consensus needs to be in favor of adding it (failure to reach consensus means we retain the status quo, which in this case is to omit such a ranking), and I think this consensus needs to be obtained on a WikiProject-wide basis. To start that, you should go to Wikipedia talk:WikiProject College Basketball, click the "new section" tab to start a new section, and post a brief explanation of why you feel that articles about college basketball tournaments would benefit from including this information. cmadler (talk) 12:39, 23 March 2011 (UTC)

What was determined?
So I can't tell where this was left. There are too many words on this page. I don't like the rankings. What do they mean? How were they used to seed the teams? They seem very out of place. There's no correlation of the rankings to the seeds. — X96lee15 (talk) 13:25, 22 March 2011 (UTC)
 * X96, my point of order proposal (please, read it, it is still there) was to temporarily include Prof. Massey-computed average rankings, with the idea of eventually confirming them or removing them when the debate eventually ended. The reasons are still there. The only comment I received was by Adler (see above). His comment did not sound to me as a rejection of the idea of temporarily including the rankings. Nevertheless, it now seems that he does object to the temporary inclusion. Anyhow, please, do keep in mind that we should not focus on whether or not we like Prof. Massey's average of 50 rankings. The focus should be whether or not a significant portion (not necessarily all or even most) of our readers will find the information useful. Whether the ranking were or not utilized by a committee member or another -- which really is impossible to determine -- is besides the point. In fact, your (correct) observation that the seeds seem inconsistent with the average rankings is a solid reason to include them: that observation is definitely something our readers may want to know, would you not agree?--Virgrod (talk) 03:30, 23 March 2011 (UTC)


 * I'm not totally against it, but I think it needs to be discussed from a WikiProject-wide perspective at the WikiProject talk page. cmadler (talk) 15:06, 22 March 2011 (UTC)