Quote: Originally Posted By colonels19 on 1/03/2010

Quote: Originally Posted By thewizard2 on 1/03/2010

I agree margin of victory should be taken into consideration, but not at face value. For instance at 50-30 win should not be equal to a 110-90 win. There should be a percentage calculated into the formula with points scored and total points by both teams.

I used to do this, but I stopped because I didn't see the point in doing it. Its margin that really matters, not points percentage. Game A says your team has a bad offense and a good defense while Game B says your team has a good offense and a bad defense...the margins are the same...its a wash to me. I think points percentage is unnecessary

Quote: Originally Posted By schroedess26 on 1/04/2010



So in other words your giving me the no answer, answer that you hate so much. There has to have been a time when the ratings were so close that the point margin kicked in right? You probably can see a lot easier than I could in your ratings. How much does point margin count for then? I will not tell you how much point margin counts for, people with ranking systems keep their formulas close to their vests, but I will tell you that point margin is 3rd on my ranking hierarchy behind 1. W-L and 2. SOS. It has very little impact.

I don't exactly see how I didn't answer your question. You seemed to have been bringing up the points percentage concept and that some teams like to slow down and keep games in the 50s or speed up and get games in the 100s. I used a points percentage in my 2005 and 2006 college football ranking, but I discontinued it because to me it was rather pointless. People talk about how much teams won or lost by, not what % of the points they had. After 2 seasons of working with it in practice...I deemed it unnecessary for my ranking process, and that's an admitted difference of opinion/process/what a ranking should be/do on both sides. So whether its 100-80 or 50-30, its still a 20 point loss, and that's what folks talk about.

Also, why do you not think that loosing to the #1 team by 1 point in double OT is not better than win against the 300th ranked team by 1 or even 3 points? Because, in my mind, if all wins aren't rated higher than all losses, then you're suggesting that winning and losing don't matter...and winning and losing becomes some kind of huge subjective football, when the object of any competitive contest is to WIN the game/contest. I used to have some losses rate ahead of some wins and when I would look at it, I would just think "this just isn't right" because also what I was finding that another can of worms was opening up questioning which losses should be higher than which wins. I've also adopted this kind of format because I don't want a team to merely be rewarded for simply PLAYING a tough schedule. I think its great the dpatterson7 manned up and took his Quincy team through hell and back, but they only went 2-8 against top 25 teams and 12-16 overall. For the sake of argument, if you take on the 28 top teams and lose all games by 1 point, you still won ZERO games in a competition/game where the ULTIMATE GOAL is to have scored more than the other team at the end of the game. In a some losses can be rated higher than some wins format...this 0-28 team would probably make the NCAA tournament, or at the very least, the NIT and that would be incredibly ridiculous in my eyes. If I ranked dpatterson's Quincy...they might make the NIT (Top 96) but by no means would they be the #1 OVERALL seed in the NIT. Hell if not for some CT upsets, Quincy probably would have been an HD NCAA tourney team. Would you not consider the first team better than the later of the two? The first team is OBVIOUSLY the better of the 2, but you still lost to them in a contest where WINNING is the objective. What is more likely here is that Team A has a much tougher schedule than Team B (I.E. 2009 Indiana v. 2009 South Dakota) and that schedule toughness will bear out in the rankings as it did for 6 win Indiana who placed 82 ranks higher than 20 win South Dakota. Losing to the best team still gives you the best loss and beating the worst team still gives you the worst win.

Quote: Originally Posted By coach_billyg on 1/04/2010
you are absolutely absurd in using your there aren't 1 game seasons defense. also, in your claiming my scenario is extremely unlikely (which it is). anybody with the slightest inkling of what is going on understands it is reasonable to consider cases with small differences and "all else equal" or "all else held constant". would you disagree with that? there is a reason those phrases are used all the time in discussions like these. But small differences and same schedules in a ranking system are 2 entirely different premises. Small differences could have a huge impact on both teams, especially given the results of all games played. I'd rather not explain myself again, so I would ask you to read my response to schrodes(sp) to see why I rank all wins over all losses.

"When I have kids, I'll be sure to tell them to lose so guys like you will rate them higher than actual winners." - this just shows how severely you don't get it. I've done rankings both ways...with point margin and points percentage and with all wins higher than all losses and some losses rated higher than some wins...please don't insult me by saying I don't get it. I understand I haven't been the most joyous person to deal with here, but come on now.

Quote: Originally Posted By schroedess26 on 1/04/2010I will give you a real life example of a team making it into the NCAA tournament as an at large with the same profile as Quincy University that dpatterson7 controls.


Arizona in 2007-2008 went 18-15 with 17 (75-350) RPI teams on their schedule. While Quincy went 12-16 with 9 (75-350) RPI teams on their schedule. Not trying to thwart this example or anything...there is a 7 game spread in W-L between the teams though (Arizona +3, Quincy -4) And that's really the difference here. If Arizona went 12-16 and didn't win their conference tournament (like Quincy didn't a 2nd round loss), they wouldn't be in the Big Dance, and they wouldn't be the #1 OVERALL seed in the NIT.

Also 200+ RPI Teams on schedule:
Arizona: 5
Quincy: 3

Arizona SOS: 2
Quincy SOS: 1

Arizona RPI: 44
Quincy RPI: 48

Finally games vs Top 10 RPI teams:
Arizona: 4
Quincy: 7

I would say if you add in the extra games to make the schedules even Quincy would be near identical. Quincy needs to make up 5 games all of which could come from below 75 RPI which he beat 8 out of 9 times.

Also, you could take 1 game against top 10 RPI schools away which would make his loss total now 15 instead of 16 if you wish to make it exactly the same.

RPI Rankings
Arizona Profile
dpatterson7 s Team Profile

Arizona ended up as a #10 seed in the NCAA Tournament not a far cry from being #1 in the NIT.

Quote: Originally Posted By mmt0315 on 1/04/2010

Quote: Originally Posted By colonels19 on 1/03/2010
I actually just tabbed my rankings and if a team were to go 28-0 against the worst schedule possible beating all teams by 50, they would have ranked 136th in my system last year, slightly ahead of 13-17 Oregon State and behind 19-14 Kent State...can you see how this works properly?

This doesnt work. A team that goes 28-0 should have an automatic bid into the tournament regardless of their competition (you can argue where they should be seeded), which your "system" would require them to win the CT. That team would be dancing in my system, and my system doesn't require that a team wins the CT, you get the auto-bid if you win the regular season conference championship, which is inherently more fair. Your system does not make sense. The way RPI is judged wins are weighed more than losses but it also allows for a "good" loss My system allows for good losses...I'd much rather lose to Kentucky by 1 than lose to Utah Valley by 1....which doesnt take into account if you kept it close I actually do take that into account via including final point margins in my rankings. as much as it does your willingness to play good opponents on the road. Teams shouldn't merely be credited for playing GOOD SCHEDULES, that's one of the reasons I adopted this ALL WINS format. If you lose the game, you failed at the objective of the game which was/is to beat your opponent in final point margin.

Quote: Originally Posted By dalter on 1/04/2010
WHOA.

That was exactly my mantra during our discussion re: using overall ratings as SOS. I kept saying it's not about talent or who appears better, but what the team did over their 30-game schedule. You kept disagreeing. That was a completely different argument. I said that W-L and SOS didn't and shouldn't matter while compiling an OTR SOS...that's a completely different system.

My real life system works exactly how you've said and we agree there...the OTR SOS concept is ENTIRELY different altogether.

OTR SOS says that better/best opponents are determined by OTRs (Overall team ratings) and my BPI RL SOS says that the better/best opponents are determined by a combination of W-L and SOS.

Quote: Originally Posted By dalter on 1/04/2010
That's wrong Its one of the conclusions that you can draw., and another good example of how you're trying to look at what you think you know from real life rather than the reality of HD. In HD, how much you score or give up is hugely related to tempo, not how good your offense/defense is. Still the hows and whys don't matter in my system...the end result is what it is, regardless of how it was attained. I know that some of you guys don't like this concept, but 110-90 and 50-30 are the same thing in my ranking system. Point margin is still 3rd in my ranking hierarchy.

Also, I could play slowdown and beat a team 50-40, or both play uptempo and I beat him 100-80. Those margins are certainly not the same, but the teams are the same and it's the result of tempo. So you're suggesting that a team would be happier if they lost by 20 as opposed to losing by 10 because the points percentage is the same? I completely disagree and including something that factors in pace for every single team in HD is practically impossible. You guys have some idealistic views and ideas, but splitting hairs like this in a ranking system simply isn't useful or worth it. Like I've said before, things that matter in the context of a game don't necessarily matter in the context of a ranking system.

Quote: Originally Posted By a_in_the_b on 1/04/2010In real life how much your opponent or you scores is heavily related to tempo. YOu can have an uptempo pressing team who has a good defense(They force allot of turnovers and the other team shoots badly) But yet gives up a relatively high aggregate of points because there is a high amount of possessions in the game.

Quote: Originally Posted By jskenner on 1/04/2010Margin of victory can be on a relative basis, putting slowdown and uptempo teams on a much more level field. For example, a slowdown team with an avg 50-40 score would be treated the same as an uptempo 80-64 team, each with a 1.25-1 scoring ratio. There has to be a straightforward technique to include this along with winning %, home/road ratio, etc. to come up with a fairly equitable ranking.

Quote: Originally posted by colonels19 on 1/04/2010
That 10 seed in the Big Dance in RL has more factors that HD doesn't have include and for good reason.  Arizona's a high prestige team yearly tournament team with a great coach and greater importance is put on this in RL than in HD.  Arizona is in a BCS conference and how far you go in your conference tournament normally plays more of a role in RL than HD...so how far did Arizona get in their CT?  I haven't looked at the profiles, just responding here.  These are 2 really different scenarios for a number of reasons, the most important/glaring being a 7 game difference in W-L.

Quote: Originally Posted By schroedess26 on 1/04/2010

Quote: Originally posted by colonels19 on 1/04/2010
That 10 seed in the Big Dance in RL has more factors that HD doesn't have include and for good reason. Arizona's a high prestige team yearly tournament team with a great coach and greater importance is put on this in RL than in HD. Arizona is in a BCS conference and how far you go in your conference tournament normally plays more of a role in RL than HD...so how far did Arizona get in their CT? I haven't looked at the profiles, just responding here. These are 2 really different scenarios for a number of reasons, the most important/glaring being a 7 game difference in W-L.

Like I showed you though it is 7 games that did not matter to his RPI or SOS much. He could have easily won the games without problem and wound up the same as arizona but the problem is that we don't have the same amount of games in HD as in RL. He only played the tough part of the schedule not the hard part of the schedule. Arizona still was 3 games over .500 and Quincy was 4 games under .500...anyway you slice it, Arizona clearly had the better win%. Suggesting that Quincy would have done X, Y, Z differently under different scenarios, if they played more games, is irrelevant. The numbers are what the numbers are....Arizona +3...Quincy -4.

You talk about how none of the prestige or anything matters? Don't think I said none...the chasm between the haves an have nots in RL CBB is vast. Arizona has a storied tradition/past that gets them places.Well I have been told in the past that the rankings in HD take into account the prestige at the start of the season. Also, it takes into account that he was losing to the #1 ranked team 2 times as well as others. It keeps him up towards the top of the rankings. Do you know exactly how HD does its rankings?! Than you can not tell me that it does not take into account how strong of an oppenet you play like they do in real life. I never suggested such a thing.

I showed you a real life example in the last 2 years of about exactly the same schedule, SOS, RPI, and then where they end up and you would not accept that. By the way Arizona won their frist game in the Pac 10 Tournament and lost in their second. Which is the Quarter-Finals, the same as Quincy!!! Winning is most important in my system...against the virtual same SOSs and RPIs, Arizona won 18 games at a +3 clip, and Quincy won 12 games at a -4 clip...I accept that perfectly...Arizona would be rated higher. As far as the Arizona conference tourney thing...ok, so its a wash, I was merely wondering out loud about it anyhow...Arizona is still +7.

You try to bring in real life examples and when I do the exact same thing you say they can't be compared. I just compared them, read above. You don't seem to think winning and losing matter all that much, its simply who you played that counts The whole point is that it did work out like real life. Arizona wasn't 12-16 and Quincy wasn't 18-15, END OF STORY Show me an example that can't be explained through a real life example then you will get somewhere.

/END OF STOR

Quote: Originally Posted By schroedess26 on 1/04/2010Do you not get it, I showed that Quincy did not play the weaker teams a bigger schedule invovles. If they played the RPI 75+ teams like I had said they would have made up the games. This then is the same.

They play the extra 8 teams of 75+ and go even 7-1 which they did in the regular season and they are now at 19-17. This is all conjecture though...hopefully I will tab Quincy in the context of my 2009 rankings by day's end, or hell maybe the 2010s

You keep trying to say that quincy is below .500, well Arizona was as well if you take out a couple of their weak opponents. But again, you're asking me to add things to one team and take out things for another...that's all conjectureYou will never have RL and HD exactly the same #s because they do not correlate exactly since of the season length and that some teams are SIM controlled. Games played don't matter all that much since the BPI number is the total points divided by number of games played. A 4 total game difference isn't that great...you can see that in my 2009 rankings 29 and 33 aren't a big difference.