SOS - What it does and doesn't tell us

[SB Shock]

Might happen. Might not happen. You just don't know. Gotta form an opinion based on what you do know.

Well, if we are talking opinions then we need to know the following because that is how our "opinion" would be formed with only 2 games played:

1. How Did Team A and B do last year?
2. How many starters are Team A and B returning?
3. Any injuries?
4. Any coaching changes?
5. Where were the games played?

[Hurley]

It seems logical that if you're not getting the response you desire, than maybe you should change your approach.

Right now, this seems tedious and I feel like stifling discussion with cat pictures.

cats to the rescue

[Hurley]

The title of this thread is "SOS - what it does and doesn't tell us."

So far, I ain't been told $hit!

[Jamar Howard 4 President]

Well, if we are talking opinions then we need to know the following because that is how our "opinion" would be formed with only 2 games played:

1. How Did Team A and B do last year?
2. How many starters are Team A and B returning?
3. Any injuries?
4. Any coaching changes?
5. Where were the games played?

That would help you with a more informed opinion, true. However, it does not mean you are unable to form an initial opinion without that information at your fingertips.

[Jamar Howard 4 President]

The title of this thread is "SOS - what it does and doesn't tell us."

So far, I ain't been told $hit!

The thread is only 2 hrs old. Patience my friend.

[The Mad Hatter]

Might happen. Might not happen. You just don't know. Gotta form an opinion based on what you do know.

No you don't. The notion that I have to form an opinion before I have adequate evidence in not only wrong, it is harmful to productive conversation. Psychology shows that when people commit themselves to a position, there is an ingrained tendency to protect that position from external criticism:

Once something is added to your collection of beliefs, you protect it from harm. You do this instinctively and unconsciously when confronted with attitude-inconsistent information. Just as confirmation bias shields you when you actively seek information, the backfire effect defends you when the information seeks you, when it blindsides you. Coming or going, you stick to your beliefs instead of questioning them. When someone tries to correct you, tries to dilute your misconceptions, it backfires and strengthens those misconceptions instead. Over time, the backfire effect makes you less skeptical of those things that allow you to continue seeing your beliefs and attitudes as true and proper. http://www.brainpickings.org/2014/05...ffect-mcraney/

Setting up a hypothetical where you force people to commit to a position that they can't possibly know enough to reasonably commit to guarantees that the original point, which otherwise could have been made with no backlash, will be contested simply because you made people unnecessarily commit to a starting position.

[OTR Shockfan]

SOS tells us what SOME people think is going to happen with SOME teams Based on what THEY think Of all the teams they care about

Hows that LOL

[Stickboy46]

It could be a fluke? Sure, any outcome could be a fluke. However, by definition, a fluke is something that happens rarely. I'm asking you to stick to what is most likely.

I do not wish to modify my scenario to yours because yours adds complexity. I wanted to start from a very simple point that I felt we could all be in agreement... then we could move on from there into more detailed conclusions.

Then in that case, I wouldn't rank either team. Ranks off those two data points are worthless.

[Jamar Howard 4 President]

The Mad Hatter, that psychology you mentioned applies to some situations, but it is silly to try and apply it here.

Go back to your car example earlier. You said Car A leaves Cleveland at 6:00 am. Car B leaves Chicago at 9:00 am. Which car gets to their destination first?

I initially said Car A is most likely while simultaneously saying that my evaluation could change with more information. If you told me that Car A is going on a 1,000 mile trip and Car B is going on a 10 mile trip, I wouldn't "protect my initial belief from harm" and insist that I still believe Car A would arrive first. I would readily admit that this new information changes my assessment and now, based on this new information, it is most likely that Car B will get to its destination first.

[Shockeriffic]

Let's just get down to the point and stop messing around with hypotheticals. It would be helpful if SOS included what the expected record of teams of varying quality would probably be against that schedule. For example, on kenpom, Notre Dame has the 198th ranked SOS. They've played against 5 top 50 teams, some mid level teams and a lot of cupcakes. They are ranked highly because the voters don't care if a game is against 300+ Binghamton or 140ish Cal Poly. For a team that good, both should be wins.

Gonzaga has the 59th best schedule even though they've played against less top 50 teams than Notre Dame. A top 25 team would have more losses with Notre Dame's schedule but a 100 level team would have more losses with Gonzaga's. Currently SOS is only concerned with how an average team would do against a certain schedule.

Final note: Notre Dame is scared to play anyone halfway decent out of conference. They scheduled seven games against 250+ competition. They were probably disappointed that Mount St. Mary's and Northern Illinois weren't as terrible as they expected when they made that travesty.

[Stickboy46]

I think the point is JH4P should never be a teacher .. He may be good at things in life .. but this is not one of them.

[Jamar Howard 4 President]

Then in that case, I wouldn't rank either team. Ranks off those two data points are worthless.

You proposed an alternate scenario that was identical to mine, just had a larger sample size, correct? You said 6 games each instead of 2. Same results, just more of them. If my data is worthless, how does a larger sample size of worthless data help?

[The Mad Hatter]

The Mad Hatter, that psychology you mentioned applies to some situations, but it is silly to try and apply it here.

Go back to your car example earlier. You said Car A leaves Cleveland at 6:00 am. Car B leaves Chicago at 9:00 am. Which car gets to their destination first?

I initially said Car A is most likely while simultaneously saying that my evaluation could change with more information. If you told me that Car A is going on a 1,000 mile trip and Car B is going on a 10 mile trip, I wouldn't "protect my initial belief from harm" and insist that I still believe Car A would arrive first. I would readily admit that this new information changes my assessment and now, based on this new information, it is most likely that Car B will get to its destination first.

In other words, the reason you won't be committed to the answer is that you knew all along that there was no possible way you could have confidence in the answer with the information presented.

So your reason that there is no problem with making people commit to an answer is that they will all know their answer is meaningless and unreliable based on the information presented. Well if that isn't a resounding reason to engage in the hypothetical, I don't know what is.

[jdshock]

Thanks for answering the question (you selected Team A), but the part of your answer that I just quoted here is the type of thing that is blowing my mind. Pre-season rankings. Whether or not we are talking about early in the season. Etc. NONE of that has any impact on my hypothetical. I've purposefully excluded all of it to make things simple.

If new information comes into play, you are all free to explain why you think the new information changes things. However, for now, everything outside my simple scenario is left outside for a reason.

You don't have to put me and others down. You act like our idiocy "blows your mind."

Like I said, I picked Team A assuming all is equal. But all is not equal. That's never the case. If Team A is ranked #290 and Team B is tied or ranked at #289, then yeah this is enough information for me to now rank Team A ahead of Team B. If Team A is #6 before this and Team B is #5, then it just doesn't justify - to me - jumping Team A over Team B. They both did what they were supposed to do. In my first hypothetical, the games being played showed that my previous understanding of Team A was radically inconsistent with reality. In the second, my ranking could be exactly right.

Just the other day, you went off on someone for laughing at 20 point predictions over UNI. And you said something that I agreed 100% with. You said that if WSU loses, it doesn't prove the guy right for his prediction, but if WSU wins by 20 the guy is obviously wrong. It wasn't a win-win scenario. My previous understanding of Team A and Team B is important because the 2-0 that both of them went might not actually change my beliefs about the teams. Assuming a perfectly equal scenario, then yeah Team A wins. But don't act like we are so much below your intellect level just because we think this hypothetical is so un-real world that it is hard to make a decision about.

Edit to clarify: My point is that you act like this is all the information you're giving out and then you're going to release more information later on to logically get us to believe you. That's not how this analysis should work. Just post "All else equal, equal RPI, equal ranks, equal players, etc -- the teams have the following record.... who do you rank higher?"

[The Mad Hatter]

Here are two changes to the original hypothetical that would make it potentially yield results:

1. Assert that all games were on the same type of location (e.g. all home games)
2. Don't ask how you would rank the two teams, ask which of the two teams played a tougher schedule

Home/Away is too big a factor on game difficulty to not be included in an SOS discussion.

Asking who played a tougher schedule is a much narrower question than how teams should be ranked because you don't need to know anything about issues like margin of victory to determine SOS, but you do to determine ranking.