I'm definitely interested.

well, it sounds like people are pretty much OK with my rpi/post season weighting for base rank. honestly, i thought that was the simplest piece, as i could check some of the components simply on a fair number of examples to make sure it was reasonable (like making sure different combinations of 3 post seasons that were worth equal amounts were about equal in difficulty). and, it sounds like we are in agreement that the all-time dynasties should weigh every season equally, and current dynasties will weigh more recent seasons more heavily, using method B described above. if anybody has any objections to those (and i welcome them - often people doing stuff like these lists don't want the "hassle" of people nit picking their ideas, but that is exactly what i am looking here in this planning stage, so have at it!)... then speak now, or forever hold your peace! (no really, you are welcome to complain later, but the chance you have of changing my opinion after the program is finished is much lower).

anyway, on to steps 2 and 3 - conf rank and how to incorporate it into team rank (actually, step 3 actually also included how to weight seasons, but we decided that already). i want to run these guidelines i am going to use to devise the conference component by you all. i am thinking of including conference tournament success, not just the championship. but, because there are very often easy paths to the top (and i am not up for generating the seeding and all that based on conf standings, which is partially impossible because you don't know the first tie breaker, head to head), i feel you have to diminish it quickly. so, if the conference tournament championship is worth 1, then i am thinking of the rest being on this scale:
champ : 1
runner up : .5
ct final 4 : .2


so, the next item is, how much do you count a conf championship as? well, for the average best conference in a division, i am thinking of making it be worth THREE points. that might sound like a lot, considering the difference between a 1st round and sweet 16 is only 3 points. but, keep in mind, most tournament teams will get some points from their conference tournament as well. also, i have spent a lot of my career in some really exceptional conferences. so i thought, for example... in those conferences, if you told me one of the teams won the CT and made the second round of the NT, and the other didn't even make the top 4 in the CT and made the sweet 16, and then asked me which team would i expect to be better, there is no question. i would pick the CT champ. so i felt the gap should be significantly higher than that gap, which is 1.8. i thought about a half dozen of the other scenarios and 3 seemed to be a reasonable figure in all of them, to me. so, what do you guys think of 3, not as as the high end limit, but what you'd get for a CT champ in the average #1 conf in a world?

ok. last item. how does it scale to other conferences? well, i was thinking that in the average worst conference, the CT barely means a damn thing. but, i didn't want to make it too hard for lower schools to compete with top conf schools. i feel like a balancing factor is that top confs tend to get screwed royally in the seeding, so i think that helps a bit. but anyway, in an effort to keep the CT exciting for everybody hoping to make the list, no matter how crappy a conf, i felt like using 1 point was fair. but, i think i am going to scale linearly between those two values BASED ON THE SCORE FOR THE CONFERENCE. that is very important because if you plot the scores for the conferences, it is definitely not going to be linear. the #15 in a conference in a division with 30 worlds is going to be below the half way mark between the #1 and #2 conference.


OK, I think that pretty well sums up my ideas on the conf stuff. once we decide this, all that will remain is world strength, and then i will actually run the numbers and put out the end result. BUT, i am not going to do it until i get some feedback on these remaining items! i want at least 5, hopefully 10 responses to this brief survey before i will post my proposal for world strength!

1) do you agree with using more than just CT champ in the dynasty rankings? do you think 1, .5, .2 is a reasonable progression for weighting the ct champ, runner up, and other final 4 participants, respectively?

2) do you agree with a score of 3 for the average best conference champion winner? even if you think its reasonable, and feel its slightly high or low, please say so! and if you think its unreasonable, if you would, i'd appreciate if you'd toss a figure in the wind :)

3) do you agree with a score of 1 for the average worst conference champion winner?

4) do you feel a linear scaling by conference score (the sum of the scores of the 12 member schools) is a reasonable way to fill out the rest of the values?

anybody? this is all uncharted territory, i'm really not comfortable with a system until i get at least a couple "yeah, that looks ok"s. and for how much thought and time i've put in this so far i'm not willing to just slap the last 2 pieces together! really, somebody has to think i am an idiot for weighting the CT so much, right? so... tell me about it!

hmmm, i dunno, doc,.... you see , this is where i keep coming back to the idea that the NT results and final rpi tell you all you need to know about a team, the rest is just fluff.

i like that you are devising a system for valuing conferences so that a weak CT does not count the same as a strong CT.

im kinda reading as i go here....

hmmmm...

yeah, i would just have a hard time valuing a CTwin as worth more than an NT win. just cant see the argument.

I think i would want the CT to be worth 2 points at most and even that will create scenarios that bother me. I think the weakest cofnerences CT winner should be awarded no more than .5 a point or so.

anyways, try to get some other feedback, i might be the minority here.

okay, doc, here is a scenario for you to chew on...

TeamA, is in a tough conference and wins the CT, thier rpi against a tough schedule is 32

TeamB is in a weak confernece (the weakest) and won thier CT. thier rpi against a weak schedule is 31.

the two teams are matched up in the first round of the NT. a beautiful 8-9 matchup in which two really different type teams get to prove on the court which is better.

Team B wins. but the fall in round2 to a strong #1 seed.

which team had the better year?

(if the coincidence that they face each other is too much for ya, then have them play other teams but teamA still loses in r1 and team b still loses in r2.

Quote: Originally posted by coach_billyg on 2/18/2010anybody? this is all uncharted territory, i'm really not comfortable with a system until i get at least a couple "yeah, that looks ok"s. and for how much thought and time i've put in this so far i'm not willing to just slap the last 2 pieces together! really, somebody has to think i am an idiot for weighting the CT so much, right? so... tell me about it!

Why don't you just go ahead and put the rankings together the best way you see fit? It should be fairly obvious from an eyeball test whether they make any sense to you or not, and tweak it from there. There's really no need to write a novel on, when it comes to down it, completely arbitrary and subjective weightings.

Quote: Originally posted by oldave on 2/18/2010okay, doc, here is a scenario for you to chew on...TeamA, is in a tough conference and wins the CT, thier rpi against a tough schedule is 32TeamB is in a weak confernece (the weakest) and won thier CT. thier rpi against a weak schedule is 31.the two teams are matched up in the first round of the NT. a beautiful 8-9 matchup in which two really different type teams get to prove on the court which is better.Team B wins. but the fall in round2 to a strong #1 seed.which team had the better year? (if the coincidence that they face each other is too much for ya, then have them play other teams but teamA still loses in r1 and team b still loses in r2.

well, i like the example. i think i would probably go with the team who won if they played each other... i tend to feel it takes a lot to justify putting 1 team over a team who beat them when it really mattered. but if they played other teams... i don't know. i would think it was kind of close. i mean if like, the 1st round team won the big east, and the other team won <insert a conf so bad i don't know its name>, i would probably go with the big east champ. that is a tough tournament to win, and i think it is a lot harder to win that than to win an 8/9 matchup.

Quote: Originally posted by lostmyth2 on 2/18/2010

Quote: Originally posted by coach_billyg on 2/18/2010anybody? this is all uncharted territory, i'm really not comfortable with a system until i get at least a couple "yeah, that looks ok"s. and for how much thought and time i've put in this so far i'm not willing to just slap the last 2 pieces together! really, somebody has to think i am an idiot for weighting the CT so much, right? so... tell me about it!

Why don't you just go ahead and put the rankings together the best way you see fit? It should be fairly obvious from an eyeball test whether they make any sense to you or not, and tweak it from there. There's really no need to write a novel on, when it comes to down it, completely arbitrary and subjective weightings.

i guess thats reasonable. i am torn between making CT significant and leaving it out altogether though, and honestly if its not that big a factor in a 10 year list, i don't think people are going to be able to point it out just by eyeballing the list. i guess i was just hoping there would be an 80% opinion and i wouldn't have to come up with a rational reason for something that is, like you said, very subjective.

Personally, I would leave CT out all together. By adjusting CT value based on quality of conference you bring in the question of subjectivity. The only way I can think to avoid that would be to base the value of a CT championship on the actual RPI of the conference. But again this brings about the problem of the historic conference RPIs not being available.

Again, I would not include CT as part of the ranking.

If you want to include it, I would not try and handicap the value, but instead would just make it a very small part of the ranking so as not to distort the numbers much.

cbg,

as discussed on sitemail, i tend to think you over-value the CTs, and think the 3 points might be a bit much.

i also see lots of problems in matchups and such in how you value them relative to conference strength. For instance, a conference could have 5 good teams and 7 mediocre to bad teams, and your conference ratings would put them pretty low. but making the finals would still seem to be a feat on par with a better-rated conference.

i just think there are too many variables to get this even close to fair. and, of course, counting them equally is a total joke.

also, in the oldave example above, no way either of those teams ends up in an 8-9 game. The Big East champ has now been upset as a 5 seed, and the small-conference team just won as a 6 or 7. Now who's had a better season? Whose fans are happier? Whose coach?

i know there are prolly quite a few folks who agree with LM and think that doc should just publish it and do away with all the novelwriting and stuff.

but i for one like the way doc is doing it. allows him to get some good feedback whichs allows him to try to get the thing "right" (or close to it) on the first try.

and allows those who are interested an inside view of this neat project

Quote: Originally Posted By wronoj on 2/19/2010also, in the oldave example above, no way either of those teams ends up in an 8-9 game. The Big East champ has now been upset as a 5 seed, and the small-conference team just won as a 6 or 7. Now who's had a better season? Whose fans are happier? Whose coach?

hmm, yeah, i didnt put a whole lot of thought into the seeds, my main point was that the two schools would be pretty evenly matched in most peoples minds, but they would be on opposite ends of the spectrum as to how they got there.

to me, the NT is where all the bets get settled. and maybe it is better to assume they didnt play each other.

i just think that the team that goes further in the NT has allways had the better season

i agree, OD.

was half nitpicking, half joking on the seeding thing.

Quote: Originally posted by wronoj on 2/19/2010cbg,

as discussed on sitemail, i tend to think you over-value the CTs, and think the 3 points might be a bit much.

i also see lots of problems in matchups and such in how you value them relative to conference strength. For instance, a conference could have 5 good teams and 7 mediocre to bad teams, and your conference ratings would put them pretty low. but making the finals would still seem to be a feat on par with a better-rated conference.

i just think there are too many variables to get this even close to fair. and, of course, counting them equally is a total joke.

well, based on the feedback, i am thinking of making the CT worth significantly less, or maybe none at all.

however, in your example with 5 good teams and 7 mediocre to bad, i am not sure why the conference ratings would put them pretty low. the team ranking i proposed really does little to distinguish teams outside of the competitive zone, so using a conf rank that is just the sum or average of the 12 schools mostly only looks at the quality schools in a conference. so a conf with more than 5 good teams, or 5 better teams, would rank higher. but a conf with 3 good teams and 9 that were better than the 7 but still not very good, would not be ranked higher, as they would in conference rpi. so, i guess i am wondering who you feel would rank above the 5 good team conf who should not?