My final thoughts on rankings, RPI, etc.

[TrollyMcTroller]

the BPI debate

captspac/trolly ==> argue its accuracy

Where did I ever argue its accuracy?

john/hockey ==> argue the way espn forces it as a major ranking system.

neither are gonna budge and we both make good points however......

I'm still waiting for you to substantiate that claim. (pointing to a reddit thread with a dozen people, some of whom agree with you, doesn't substantiate anything)

And I don't care if you budge or not. I'm not trying to convince you to like the BPI. I'm simply pointing out how silly the reasons are you've given for not liking it.

You claim that MoV and injuries are something that should be factored in when comparing teams, but you don't like the BPI because it factors in MoV and injuries.

You claim that ESPN is forcing it down everyone's throat, but then you turn around and say the only way it would be credible is if more analysts used it.

You claim your biggest problem with the BPI is that in a previous season it didn't rank Syracuse very high when it was blowing through it's typical lineup of shitty, instate cannon fodder, but you just couldn't understand why their BPI improved when they started their conference schedule and started playing teams didn't have to sub in their mascot.

You claim that it is was designed to predict who will be in the tournament, and since it doesn't do that as accurately as the RPI it has no worth, but you ignore that fact that it is designed to predict who should be in the tournament.

It's pretty clear at this point you just have some angst boner for ESPN, and you're trying to rationalize it any way you can.

 

[john01992]

Where did I ever argue its accuracy?

where did i say your name was catstoppac

You claim that MoV and injuries are something that should be factored in when comparing teams, but you don't like the BPI because it factors in MoV and injuries.

no dumbass my claim is that MoV and injuries are something that SHOULDNT be included. jeez talk about dumb. you don't even understand my point. which is this ==> those are eyeball test measurements and not really statistical measurements. while i dont like them being included in a mathematical ranking system the problem could be slightly diminished if other metrics such as team momentum, rivalry game, time of year, attendance, etc. get included as well. MoV & injuries are something that can play a factor, but there are 10-20 other things that play an equally important factor and I don't see why only 2 should be included while the other 18 are not.

You claim that ESPN is forcing it down everyone's throat, but then you turn around and say the only way it would be credible is if more analysts used it.

yeah analysts NOT FROM espn

You claim your biggest problem with the BPI is that in a previous season it didn't rank Syracuse very high when it was blowing through it's typical lineup of shitty, instate cannon fodder, but you just couldn't understand why their BPI improved when they started their conference schedule and started playing teams didn't have to sub in their mascot.

because syracuse went 18-1 beat 3 ranked teams including knocking off #1 louisville @ KFC and didn't do very well in the BPI. the rest of the regular syracuse went 5-7 (1-4 against ranked teams) with their wins coming against #25 notre dame, SHU, provy, depaul, & SJU only to suddenly shoot up into the top 10 of BPI.

that is 100% grade A bullshit. it means that purely because of SOS no matter what SU does they won't do well in the BPI regardless of the Ws in the first half of the season. But because they play a hard schedule in the 2nd half of the season they will jump in the BPI regardless of the Ws. again thats grade A bullshit. all SU did in that 5-7 stretch was get crushed by the elite BE teams (#5, #10, #11, #22), beat a semi decent BE team and 4 other cupcake BE teams. SU essentially got lots of credit for LOSING to quality teams. that sounds exactly like how the BCS & the SEC works and I fundamentally believe with all my heart that that is one of the most unfair things to do. It pretty much means your W/L record doesn't matter ==> just as long as you play elite teams. while some may agree that W/L < quality of opponent i just don't see how anyone can defend how much weight BPI gives to quality of opponent and so little to W/L.

You claim that it is was designed to predict who will be in the tournament, and since it doesn't do that as accurately as the RPI it has no worth, but you ignore that fact that it is designed to predict who should be in the tournament.

no numnuts.....
catstopac argues that BPI > RPI because it is better at predicting tournament RESULTS. I say that that is an unfair comparison to make because RPI is all about resume and predicting who is most likely to get into the tourny. RPI has shown that it is better at predicting tourny bids than BPI.

It's pretty clear at this point you just have some angst boner for ESPN, and you're trying to rationalize it any way you can.
it is pretty clear that you are out of the loop here and don't fully understand the arguments that posters on this board are making

 

[TrollyMcTroller]

where did i say your name was catstoppac

captspac/trolly ==> argue its accuracy

Or was that a rhetorical question?

my claim is that MoV and injuries are something that SHOULDNT be included. jeez talk about dumb. you don't even understand my point. which is this ==> those are eyeball test measurements and not really statistical measurements. while i dont like them being included in a mathematical ranking system the problem could be slightly diminished if other metrics such as team momentum, rivalry game, time of year, attendance, etc. get included as well. MoV & injuries are something that can play a factor, but there are 10-20 other things that play an equally important factor and I don't see why only 2 should be included while the other 18 are not.

Read what I wrote, you stupid shitheel, not what you prentended I said. Here it is again for you:

You claim that MoV and injuries are something that should be factored in when comparing teams, but you don't like the BPI because it factors in MoV and injuries.

You clearly agree, as you just reiterated, that those should indeed be factored in when comparing teams, which is exactly what I fucking said. And you also said that you don't like the BPI because it includes the very factors that you think are relevant when comparing teams. Which you also just fucking agreed with. And you wanna fucking complain about other people's reading comprehension? Go fuck yourself.

That being said, Sagarin and Kenpom both use scoring margin in their ratings and they are two of the most highly regarded and often quoted metrics from sports analysts.

But I understand your point just fine. I'm just demonstrating how fucking ridiculous it is. It's not your fault that you don't have a fucking clue about statistics, but it is your fault for trying to act like you do.

yeah analysts NOT FROM espn

Yeah... and? You still bitch about it being crammed down your throat. (still an unsubstantiated claim by the way) Yet you claim the only way it would be validated is if more analysts use it. (again, ignoring the fact that can't happen because of factors that have nothing to do with the BPI itself)

because syracuse went 18-1 beat 3 ranked teams including knocking off #1 louisville @ KFC and didn't do very well in the BPI. the rest of the regular syracuse went 5-7 (1-4 against ranked teams) with their wins coming against #25 notre dame, SHU, provy, depaul, & SJU only to suddenly shoot up into the top 10 of BPI.

that is 100% grade A bullshit. it means that purely because of SOS no matter what SU does they won't do well in the BPI regardless of the Ws in the first half of the season. But because they play a hard schedule in the 2nd half of the season they will jump in the BPI regardless of the Ws. again thats grade A bullshit. all SU did in that 5-7 stretch was get crushed by the elite BE teams (#5, #10, #11, #22), beat a semi decent BE team and 4 other cupcake BE teams. SU essentially got lots of credit for LOSING to quality teams. that sounds exactly like how the BCS & the SEC works and I fundamentally believe with all my heart that that is one of the most unfair things to do. It pretty much means your W/L record doesn't matter ==> just as long as you play elite teams. while some may agree that W/L < quality of opponent i just don't see how anyone can defend how much weight BPI gives to quality of opponent and so little to W/L.

Shit like that happens with the RPI all the time. Good teams in power conferences almost always see their RPI raise as conference play progresses. It certainly isn't a phenomenon unique to the BPI. Only 25% of a teams RPI rating is based on what that team did on the court. The other 75% is SoS (who you played, and how good they are, based on who they played) Didn't know that did you?

catstopac argues that BPI > RPI because it is better at predicting tournament RESULTS. I say that that is an unfair comparison to make because RPI is all about resume and predicting who is most likely to get into the tourny. RPI has shown that it is better at predicting tourny bids than BPI.

If you'll recall the original context of the discussion when BPI was brought up was rating teams not resumes. As BPI has proven to be a more accurate predictor of outcomes, it is clearly better at rating teams than the RPI, so all your bullshit arguments about predicting tournament teams and seeting, means exactly fuckall to this discussion.

 

[TrollyMcTroller]

It pretty much means your W/L record doesn't matter ==> just as long as you play elite teams. while some may agree that W/L < quality of opponent i just don't see how anyone can defend how much weight BPI gives to quality of opponent and so little to W/L.

Oh, by the way, you might want to look at Kansas's RPI rating and then have a nice hot cup of shut the fuck up.

 

[john01992]

Oh, by the way, you might want to look at Kansas's RPI rating and then have a nice hot cup of shut the fuck up.

KU is 11-4 with wins over 2 ranked teams not 5-7 ==> you STFU!!!!!!!!!

 

[TrollyMcTroller]

KU is 11-4 with wins over 2 ranked teams not 5-7 ==> you STFU!!!!!!!!!

Look at what I was responding to...

It pretty much means your W/L record doesn't matter ==> just as long as you play elite teams. while some may agree that W/L < quality of opponent i just don't see how anyone can defend how much weight BPI gives to quality of opponent and so little to W/L.

And then look at what I said.

Kansas is a perfect example of the Win/Loss record not mattering in the RPI because they beat elite teams. There are at least 40 teams in the country right now that have a better Winning% than Kansas, yet the RPI has them rated #2. Since RPI is only made up of 2 basic components (W/L and SoS) and they clearly aren't #2 based on their W/L numbers, it's pretty safe to conclude that the RPI weights SoS pretty fucking heavily. (Not to mention, I told you the exact fucking breakdown)

You know what... forget all that. You're clearly never going to be able to reconcile that pile of shit you got yourself into.

Let's go back to "The BPI incorporates too many (and not enough) useful factors" That seemed like it was going to be a fun conversation.

(Of and I'm still waiting for you to substantiate your claim that ESPN was cramming the BPI down people's throats)

 

[john01992]

Look at what I was responding to...



And then look at what I said.

Kansas is a perfect example of the Win/Loss record not mattering in the RPI because they beat elite teams. There are at least 40 teams in the country right now that have a better Winning% than Kansas, yet the RPI has them rated #2. Since RPI is only made up of 2 basic components (W/L and SoS) and they clearly aren't #2 based on their W/L numbers, it's pretty safe to conclude that the RPI weights SoS pretty fucking heavily. (Not to mention, I told you the exact fucking breakdown)

You know what... forget all that. You're clearly never going to be able to reconcile that pile of shit you got yourself into.

Let's go back to "The BPI incorporates too many (and not enough) useful factors" That seemed like it was going to be a fun conversation.

(Of and I'm still waiting for you to substantiate your claim that ESPN was cramming the BPI down people's throats)

you can lead a horse to the water but you can't make him drink. i have already explained how espn has been pushing BPI. you just refuse to listen. so why bother?

they won 11 games and 4 losses and you say the W/L columns dont matter

 

[TrollyMcTroller]

you can lead a horse to the water but you can't make him drink. i have already explained how espn has been pushing BPI. you just refuse to listen. so why bother?

You can explain it all you want. That doesn't substantiate it. You can point to a handful of people that agree with you. That doesn't substantiate it either.

I can claim that the Earth is flat. I can give you a lengthy explanation about its flatness. I can point you to a number of websites where people endorse the same theory. That doesn't mean I've done anything to substantiate my outlandish claim, and until I do, it's just a baseless, useless claim. Just. Like. Yours.

Do you know what "substantiate" means?

they won 11 games and 4 losses and you say the W/L columns dont matter

I also said there are at least 40 teams that have better W/L records, that are ranked below Kansas. I'm not so sure why that's difficult for you to comprehend. (Just kidding! I know exactly why. Because you're an idiot!)

 

[john01992]

You can explain it all you want. That doesn't substantiate it. You can point to a handful of people that agree with you. That doesn't substantiate it either.

I can claim that the Earth is flat. I can give you a lengthy explanation about its flatness. I can point you to a number of websites where people endorse the same theory. That doesn't mean I've done anything to substantiate my outlandish claim, and until I do, it's just a baseless, useless claim. Just. Like. Yours.

Do you know what "substantiate" means?



I also said there are at least 40 teams that have better W/L records, that are ranked below Kansas. I'm not so sure why that's difficult for you to comprehend. (Just kidding! I know exactly why. Because you're an idiot!)

not my fault that you cant tell the difference between 5-7 & 11-4

 

[TrollyMcTroller]

not my fault that you cant tell the difference between 5-7 & 11-4

Not my fault you can't see the similarity.

Oh wait... I know how to make it into a comparison that you can understand... The 11-4 is racist.

Your complaint was that SoS mattered too much in the BPI. I'm simply pointing out that it works the same way in the RPI. I actually gave you the percentage breakdown of how much goes into Wins and losses (25%) vs SoS (75%) That's 3:1 in favor of SoS for those of you keeping track at home. And I gave you an example of a team with a mediocre W/L record, and a remarkably high RPI.

I don't really care about your little Syracuse anecdote, as you've yet to support it with any actual facts. I've given you a concrete example. Not some hazy recollection of a system I'm clearly biased against.

 

[john01992]

like i said stupidity is rampant with this one

11-4 vs 5-7 a .733 vs a .416 winning percentage and he thinks that these are comparable examples

 

[TrollyMcTroller]

like i said stupidity is rampant with this one

11-4 vs 5-7 a .733 vs a .416 winning percentage and he thinks that these are comparable examples

And like I said. there are 40 teams with better win% than Kansas's that are rated below Kansas in the RPI. That clearly demonstrates that Kansas's on court performance is less important than their SoS (the only other component to the RPI)

Is the Kansas example (note, I say "example" because I've provided you with real numbers) as extreme as your Syracuse anecdote (note, I say "anecdote" here, because you're just relaying things like "didn't do very well" and for all I know, you're making this shit up. You've made up plenty of other bullshit since you've been posting)

No it isn't.

Does it still address the exact complaint you made about the BPI? Let's see what you wrote:

It pretty much means your W/L record doesn't matter ==> just as long as you play elite teams. while some may agree that W/L < quality of opponent i just don't see how anyone can defend how much weight BPI gives to quality of opponent and so little to W/L.

Yes. It fucking does.

The RPI uses 25% win% and the other 75% has fuckall to do with on court performance of the team being evaluated.

 

[john01992]

And like I said. there are 40 teams with better win% than Kansas's that are rated below Kansas in the RPI. That clearly demonstrates that Kansas's on court performance is less important than their SoS (the only other component to the RPI)

Is the Kansas example (note, I say "example" because I've provided you with real numbers) as extreme as your Syracuse anecdote (note, I say "anecdote" here, because you're just relaying things like "didn't do very well" and for all I know, you're making this shit up. You've made up plenty of other bullshit since you've been posting)

No it isn't.

Does it still address the exact complaint you made about the BPI? Let's see what you wrote:



Yes. It fucking does.

The RPI uses 25% win% and the other 75% has fuckall to do with on court performance of the team being evaluated.

i am amazed that I have to walk you through this.

at 11-4 the RPI shows the KU gets rewarded for playing a combination of tough sos & winning.

at 18-1 syracuse does not get rewarded for winning, only bashed for a weak sos. but at 5-7 SU obviously is terrible in both winning percentage and lacks elite wins and yet somehow they suddenly shoot up the rankings?

this is where your stupidity and complete lack of understanding comes into play. you simply are not capable of coherent thought and this is not the first time i have had this problem with you. you bring 2 examples (that you think are similar) and say "well look at this" trying to compare it. but reality is that those two examples are totally different and not comparable at all. <== and this is not the first time you have pulled crap like this.

 

[TrollyMcTroller]

i am amazed that I have to walk you through this.

at 11-4 the RPI shows the KU gets rewarded for playing a combination of tough sos & winning.

at 18-1 syracuse does not get rewarded for winning, only bashed for a weak sos. but at 5-7 SU obviously is terrible in both winning percentage and lacks elite wins and yet somehow they suddenly shoot up the rankings?

this is where your stupidity and complete lack of understanding comes into play. you simply are not capable of coherent thought and this is not the first time i have had this problem with you. you bring 2 examples (that you think are similar) and say "well look at this" trying to compare it. but reality is that those two examples are totally different and not comparable at all. <== and this is not the first time you have pulled crap like this.

First of all, I'm the only one that brought an example to this discussion. You just brought some bullshit story. I can point you to any RPI website you want and show you exactly what I'm talking about. I can write out the entire RPI formula if you want.

You can look at any of the 40+ teams that are ranked lower (many significantly lower) than Kansas and see that they weren't rewarded for winning, rather punished for having a poor SoS. It's exactly the same fucking thing. You just refuse to acknowledge it.

75% of the RPI is SoS based. SEVENTY FIVE PERCENT Of THE ARE PEE EYE IS ESS OH ESS BASED.

Do I need to write that out in crayon for you?

 

[john01992]

First of all, I'm the only one that brought an example to this discussion. You just brought some bullshit story. I can point you to any RPI website you want and show you exactly what I'm talking about. I can write out the entire RPI formula if you want.

You can look at any of the 40+ teams that are ranked lower (many significantly lower) than Kansas and see that they weren't rewarded for winning, rather punished for having a poor SoS. It's exactly the same fucking thing. You just refuse to acknowledge it.

75% of the RPI is SoS based. SEVENTY FIVE PERCENT Of THE ARE PEE EYE IS ESS OH ESS BASED.

Do I need to write that out in crayon for you?

you still don't get it..........

in RPI it clearly shows that winning games matters more whereas in BPI sos has more emphasis

granted a very unusual case like KU who has a #1 sos by a longshot can jump into the top 4. but KU has a very decent winning % to go along with that and justify it. i am saying there should be a BALANCE between the two, not that RPI should go purely based on W/L which you are proposing. look at wichita st. undefeated and #11 in RPI but #17 in BPI ==> and you wanna say that it is the BPI that gives more weight to winning % and less on sos???? look at SDSU. they have 1 loss and are 15 in RPI, while 27th in BPI

 

[gordontrue]

in RPI it clearly shows that winning games matters more whereas in BPI sos has more emphasis

No, incorrect. It doesn't get more SOS based than the RPI. The BPI does not give more emphasis to SOS, it gives more emphasis to scoring margin, pace, and missing key players.

 

[TrollyMcTroller]

And let's not forget your track record for anecdotal evidence...

What you said:

espn has developed both QBR (the system that says aaron rodgers is worse than tebow)

What actually happened:
Aaron Rodgers played an entire game. Tim Tebow played a couple of series in the 4th quarter. Tim Tebow ended up with a higher QBR for that game. Somebody saw that, and then claimed on a message board "The QBR says that Tebow is better than Rogers" (which of course is patently false, but you didn't know that, nor did you bother to verify it before you thoughtlessly regurgitated it on this board)

So with your track record on such matters, you'll forgive me if I don't believe a fucking word you say that isn't backed up by some actual facts.

 

[TrollyMcTroller]

you still don't get it..........

in RPI it clearly shows that winning games matters more

Don't tell me it does. Prove it. I've given you the numbers. They disagree completely with what you're claiming. So prove it.

granted a very unusual case like KU who has a #1 sos by a longshot can jump into the top 4. but KU has a very decent winning % to go along with that and justify it.

Why is KU an unusual case? Are you sure your Cuse anecdote wasn't an unusual case too? And what the fuck is "very decent" is that like an "extra medium" or "extraordinarily mediocre" ?

i am saying there should be a BALANCE between the two, not that RPI should go purely based on W/L which you are proposing.

I proposed no such thing you god damn liar.

look at wichita st. undefeated and #11 in RPI but #17 in BPI ==> and you wanna say that it is the BPI that gives more weight to winning % and less on sos???? look at SDSU. they have 1 loss and are 15 in RPI, while 27th in BPI

No, the BPI uses a lot more factors than simply winning%. With the RPI you can reasonably conclude that any difference between two teams with identical win% is due to SoS, as that's the only other metric used. If a team with a better winning percentage is ranked below a team with a worse win% you can draw the same conclusion. If the difference isn't win% then it must by SoS.

You can't draw the same conclusion with the BPI as it could be any number of factors causing the discrepancy. You can say it isn't win% but you certainly can't claim that it is SoS with any confidence.

 

[element1286]

I guess john shows exactly how journalism majors view statistical analysis, if it doesn't fit my preconceived notions it must be invalid. Read carefully your favorite sports writers probably approach them the same way.

 

[john01992]

i can't do this: trollys reading comprehension/coherent thought process is so terrible I can't even have a simple debate with him.

for example: No, the BPI uses a lot more factors than simply winning%

he doesn't even get what the current debate is even about.