The Mad Scientist Top 25 Ranking Debate Topic

colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
Quote: Originally Posted By a_in_the_b on 12/23/2009
If you can see that overall rating does not accurately measure the difference between teams save in the most general terms and at the extremes of ability, you don't need to see it formally done to know it won't work.
Your statement is inherently false. Overall team ratings can be improved and tweaked to become even more accurate measures of team talent and judging which teams are better and worse.

If you can't see how overall team ratings COULD trump/be better than W-L and SOS across a myriad of different varying schedules in games that have CONCRETE ratings, then you're being entirely closed-minded and you're flat wrong. You've just been conditioned with W-L and SOS for so long that you won't even CONSIDER concepts that differentiate from your train of thought...that's weak if you ask me. I get what you all are saying...my real life rankings are based off of wins, losses, and strength of schedule, and one of my rankings for this game would be as well...but none of you want to see that, acknowledge that, or give me kudos for that...you just want to see the perceived "wrongs" in my argument because at the heart of it, you guys are all just haters that can't stand me because of my staunch loyalty to concept and the way that I debate, not necessarily the argument/debate itself. I'm not saying overall ratings is the only way to do it, I'm saying its a viable way...that's what you all are missing really.
12/23/2009 8:21 PM
namshub
Posts: 280 (0)
Hall of Famer
48 3 307 2 37
Rookie
colonels, i'll bite. there is no perfect ranking system so someone is always going to be able to pick at your formula. also, the way you come across does make one want to poke holes in your approach.

having said that, i do agree that as a general proposition, better core ratings does make for a better team. that is a very limited statement qualified by IQ and number of scholarship players on a roster.

on another note, not only are rankings currently flawed but the current RPI system is extremely flawed as well so those that want to use that measure as the end all for the NT you have many holes in your argument as well.

one thing i certainly hope WIS considers when modifying its ranking system and tourney selection criteria is the interplay making the NT has on job promotion. you start keeping small conference teams and mid-majors in ghost conferences out of the tourney on a consistent basis and not allowing these coaches to qualify to move up because they don't have NT experience you will have a mess on your hands.

12/23/2009 8:25 PM
a_in_the_b
Posts: 1653 (0)
Hall of Famer
1 18 666 3
Pro
Alright. YOu just admitted you lost by pulling out the "Haters" argument. Once someone pulls that out. . .

The lamest counter argument of all time.
12/23/2009 8:26 PM
colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
Quote: Originally Posted By a_in_the_b on 12/23/2009Alright. YOu just admitted you lost by pulling out the "Haters" argument. Once someone pulls that out. . .

The lamest counter argument of all time.
I haven't lost anything pal, I'm still here, I'm still fighting, I'm still debating. You pick 1 word out of a 200 word response and harp on it....I think that says something about your tactics as well. You can't argue the contents of my post, thus you just find the most "absurd" thing you think you can and make the tiniest little jab you can about it....very weak.
12/23/2009 8:29 PM
namshub
Posts: 280 (0)
Hall of Famer
48 3 307 2 37
Rookie
colonels, i believe that a senior laden team is the outlier (the 50 pt. lower rated team that can consistently beat a higher rated team) that really skews the overall rating concept you propose. how would you take IQ or SR./JR. heavy teams into account in your proposal?
12/23/2009 8:33 PM
colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
Quote: Originally Posted By namshub on 12/23/2009
colonels, i'll bite. there is no perfect ranking system so someone is always going to be able to pick at your formula. also, the way you come across does make one want to poke holes in your approach. I'll agree with that, but these guys want to debate this way, so I'm giving them what they want/asked for.

having said that, i do agree that as a general proposition, better core ratings does make for a better team. that is a very limited statement qualified by IQ and number of scholarship players on a roster. The overall rating needs to be tweaked no doubt...I suggested player minutes/team minutes playing a part as well as incorporating things like IQ and FT Shooting. The overall rating as is, is by no means perfect, but its still useful, but you do make good points as to its shortcomings.

on another note, not only are rankings currently flawed but the current RPI system is extremely flawed as well so those that want to use that measure as the end all for the NT you have many holes in your argument as well. Agreed...not really a big fan of the RPI, but having done my own rankings for a while now, I think its important to produce something that you believe in, and feel is the best way IYO to rank teams. Everyone that does a ranking is and should be on trial for it, but undertaking the challenge you have to know that there are people that will love your system, hate your system, and be in the middle here. To be honest, I'm a bit surprised at the backlash here (doesn't mean I'm going to back down at all)...its fair to not like my concept, but to not accept it as a logical, intelligent, innovative way to do rankings has been a bit surprising to me.

one thing i certainly hope WIS considers when modifying its ranking system and tourney selection criteria is the interplay making the NT has on job promotion. you start keeping small conference teams and mid-majors in ghost conferences out of the tourney on a consistent basis and not allowing these coaches to qualify to move up because they don't have NT experience you will have a mess on your hands. Fair enough. WIS should do what the majority of its users want it to do...everyone has their likes and dislikes. Thank you for the post by the way, its nice to civilly discuss things with people in this thread. Take care.



12/23/2009 8:38 PM
colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
Quote: Originally Posted By namshub on 12/23/2009
colonels, i believe that a senior laden team is the outlier (the 50 pt. lower rated team that can consistently beat a higher rated team) that really skews the overall rating concept you propose. how would you take IQ or SR./JR. heavy teams into account in your proposal?
I'll spell it out more than I have done currently. My plan is to do something like this...

Let's say Larry Wade is a 500 core rating (everything but WE-ST-DU, including IQ and FT Shooting, rated 0 (F) to 100 (A+), having a 7.7 rating difference between grade levels....F 0, D- 7.7, D 15.4, D+ 22.1, and so on) and that Wade plays 500 minutes out of 2000 possible minutes played.

I would multiply .25....percentage of minutes played by his rating and use that number, along with the number of the other 11 players to get an adjusted core team rating, based upon usage.

If your 270 walk on plays 20 out of 2000 minutes (1% of available minutes) then only 2.7 of his points will be applied to the adjusted team total while 125 of Larry Wade's points would be applied because he played many more minutes. If you're just looking at the 2 guys...125 out of 127.7 (125+2.7) 98% is a lot fairer assessment than 500 out of 770 (500+270) 65%.

That's the raw idea I have in my head...I would need WIS' help with this if I were to move forward though....I could do it for my team and teams where guys submitted me their rosters and asked me to rank them. If all we get out of this is a more solid overall rating, then I'd be satisfied.
12/23/2009 8:44 PM
fd343ny
Posts: 2254 (0)
Hall of Famer
7 634 36
Veteran
when a bunch of smart people do not understand what someone is explaining, it is possible that they are closed minded or it is possible that the explanation is inadequate or it is possible that that which is being explained is just wrong

there may be other explanations also, but colonel, please consider whether repeatedly saying that folks dont understand you begins to suggest that you aint splaining real well
12/23/2009 10:02 PM
fd343ny
Posts: 2254 (0)
Hall of Famer
7 634 36
Veteran
having read the thread, I am not convinced that arithmetic along the lines that you describe would better model real team strength than a results based analysis - unless very complex adjustments to raw ratings points were made - too complex to be manageable

I think your methodology would make a lot of sense in a simpler game - take the extreme (not actual) case where every player had a numerical strength rating - integers from 1 thru 10 - and where there were no choices as to off/def/depth chart etc - and where the engine allocated playing time - THEN, strength of a team would be pretty much as you describe

because this game is far from what I described above (thankfully, cause I wouldnt want to play such a game), I dont see a way to represent strength by mere appplication of ratings (again, without real complex adjustments)

part of the disagreement here may go to what one means by strength - I tend to think of it as a Sagarin rating would - what team is more likely to win the next game. Another meaning of strength might be which team has better talent in some abstract sense. Those are not the same. One could create a talent ranking, but would it have predictive value? would it measure odds of winning and competitive strength as the team is actually used?

see for example:

"My point is, teams with better core ratings are still really the better teams, regardless of what happens in the games. Underperformance, coach quality, and offense/defense aren't issues here."

I dont think that is what others mean by strength - I think others mean ability to win games - not some abstraction of core ratings - better core ratings means to me that the team has better ratings, not that it is a better team. The better team needs to consider other - tougher to calculate - elements (depth, IQ, suitability of talent for off/def and the like)

I dont have an answer, I dont believe in name calling in these forums and I think much of the tone of this thread has been disappointing and has obscured some interesting questions

thats my take
12/23/2009 10:10 PM
fd343ny
Posts: 2254 (0)
Hall of Famer
7 634 36
Veteran
one more point - I would strongly urge WIS not to devote its energy to such an effort. I think there are a bunch of things that need fixing in this game and a bunch of improvements that make good sense. The approach proposed does not seem to me likely to produce useful results and it would, in my opinion, be a shame if WIS spent its limited resources on this sort of an approach to rankings

that being said, it would be great if colonel took the data from say one division in one world and produced rankings. Doesnt seem to me that it would - as he describes it - need any special WIS access or support, since the ratings of every player are there for all to see - as are the results of every game.

and if there is something out there in public in the real world that you have produced in the way of rankings that we can look at, please point us toward it - would be interesting to see it - sorry if it was referenced somewhere in the thread but I didnt find it if it was there.
12/23/2009 10:15 PM
colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
Quote: Originally Posted By fd343ny on 12/23/2009

having read the thread, I am not convinced that arithmetic along the lines that you describe would better model real team strength than a results based analysis - unless very complex adjustments to raw ratings points were made - too complex to be manageable

I think your methodology would make a lot of sense in a simpler game - take the extreme (not actual) case where every player had a numerical strength rating - integers from 1 thru 10 - and where there were no choices as to off/def/depth chart etc - and where the engine allocated playing time - THEN, strength of a team would be pretty much as you describe

because this game is far from what I described above (thankfully, cause I wouldnt want to play such a game), I dont see a way to represent strength by mere appplication of ratings (again, without real complex adjustments)

part of the disagreement here may go to what one means by strength - I tend to think of it as a Sagarin rating would - what team is more likely to win the next game. Another meaning of strength might be which team has better talent in some abstract sense. Those are not the same. One could create a talent ranking, but would it have predictive value? would it measure odds of winning and competitive strength as the team is actually used?

I dont have an answer, I dont believe in name calling in these forums and I think much of the tone of this thread has been disappointing and has obscured some interesting questions

thats my take

Thank you for posting. I admit I've gotten out of line a few times with the name calling, it wasn't my intent but it happened and I am sorry for it. I considered deleting it from the posts I made, but decided not to because I don't want to play it like I never said/did it. I think my frustration in said situations was understandable, but I really shouldn't have stooped to the level of my "peers" here. Its frustrating when you'll stand in front of the firing squad, answer and take on everything that everyone says, and those same people on the firing squad don't reciprocate...that's weak to me. Anyhow....

I'm not big on predictive rankings, all of my rankings are retrodictive because I think rankings should tell what has happened, not guess at what will happen, because those guesses can be wrong. Everyone has their opinions of what rankings should be/do and I like retrodictive rankings because I like to deal in what happened, not what if. I also think retros are more logical...predictives are more for spread/betting line purposes.

Its easy to tell someone else they're wrong when you don't have or have ever thought of a viable solution to a perceived problem. I think all the angst against my system/premise is EXTREMELY premature and unwarranted...but people will think what people will think.
12/23/2009 10:25 PM
colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
Quote: Originally Posted By fd343ny on 12/23/2009
when a bunch of smart people do not understand what someone is explaining, it is possible that they are closed minded or it is possible that the explanation is inadequate or it is possible that that which is being explained is just wrong I'm not sure everyone arguing against me is "smart" in the context of the argument that we're talking about here, and that's funny because I've been told that I am dumb/sub par in elements of the game that have no pertinence to the issues at hand.

there may be other explanations also, but colonel, please consider whether repeatedly saying that folks dont understand you begins to suggest that you aint splaining real well I can't help it if people keep getting caught up in crap that doesn't matter. Folks here seem to want to pick out the minutist parts/words of my post and try to call me out about it...many of the guys still in this thread have just adopted a gang mentality...I'll keep going though, regardless of the situation.

12/23/2009 10:29 PM
colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
Quote: Originally Posted By fd343ny on 12/23/2009
one more point - I would strongly urge WIS not to devote its energy to such an effort. I think there are a bunch of things that need fixing in this game and a bunch of improvements that make good sense. The approach proposed does not seem to me likely to produce useful results and it would, in my opinion, be a shame if WIS spent its limited resources on this sort of an approach to rankings That's a fair point...I'm not expecting much to come of this situation. Its a matter of principle and respect now...I will not be down-trodden by folks that simply don't know any better.

that being said, it would be great if colonel took the data from say one division in one world and produced rankings. Doesnt seem to me that it would - as he describes it - need any special WIS access or support, since the ratings of every player are there for all to see - as are the results of every game. I'd love to, even just for the hell of it...just see what it spits out...I have 3 sets...we could see which is the best.

and if there is something out there in public in the real world that you have produced in the way of rankings that we can look at, please point us toward it - would be interesting to see it - sorry if it was referenced somewhere in the thread but I didnt find it if it was there. www.bpisports.com there's a lot of different things on there...my first CBB rankings of the year will be up in 2 Mondays. I've spent 7+ years thinking/testing/changing my college football rankings and developing the thoughts and processes I see fit in a good ranking system. In my eyes, my rankings are the best out there...but that's really for the public to judge...not everyone will love them, I understand that.

12/23/2009 10:33 PM
lostmyth2
Posts: 2 (0)
Hall of Famer
191 3 2
Prospect
They are extremely pertinent. The fact that you are trying to come up with a ratings-based system when you so clearly do not understand how ratings actually contribute to the strength of a team is the crux of the issue.

Let me give a extremely clear example. You want to convert free throws into a 0-100 scale where F = 0 and A+ = 100, and then add this to other ratings. Seriously? F for free throws is a zero? Any half-decent coach could instantly tell in a second how inane this would be.
12/23/2009 10:37 PM
colonels19
Posts: 4399 (0)
Hall of Famer
78 90 3 1 168 2 1 11
Prospect
Quote: Originally Posted By lostmyth2 on 12/23/2009
They are extremely pertinent. The fact that you are trying to come up with a ratings-based system when you so clearly do not understand how ratings actually contribute to the strength of a team is the crux of the issue.

Let me give a extremely clear example. You want to convert free throws into a 0-100 scale where F = 0 and A+ = 100, and then add this to other ratings. Seriously? F for free throws is a zero? Any half-decent coach could instantly tell in a second how inane this would be.
If you don't set the lowest grade (failing no less) at the lowest possible score for a given rating (1?), then what's the sense in using it at all? If the low number of FT Shooting is say, 50, then it skews the entire rating system. I'm trying to be logical and fair in this instance....what's your solution?
12/23/2009 10:41 PM
◂ Prev 1...37|38|39|40|41...75 Next ▸
The Mad Scientist Top 25 Ranking Debate Topic
△Top△Bottom

Search Criteria

Terms of Use Customer Support Privacy Statement

© 1999-2026 WhatIfSports.com, Inc. All rights reserved. WhatIfSports is a trademark of WhatIfSports.com, Inc. SimLeague, SimMatchup and iSimNow are trademarks or registered trademarks of Electronic Arts, Inc. Used under license. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.