Hoya suxa

[Newsdreams]

It didn’t matter. We won.

You be very wrong. Not how it works now.

[5DollarPitcher]

2018 wants its take back. Like it or not, the RPI is dead and style points absolutely do matter.

Except they don’t. If a Horizon league team wins every game by 1 point, and goes on to win every game in the tourney by 1 point, they just won a national championship.

They don’t matter. Win the games in front of you.

[brewcity77]

Except they don’t. If a Horizon league team wins every game by 1 point, and goes on to win every game in the tourney by 1 point, they just won a national championship.

They don’t matter. Win the games in front of you.

As soon as you provide the example of this actually happening, you'll have a point. Until then, you're just defending a bad decision.

I don't get why some people get so defensive when it comes to defending Shaka. He makes mistakes. He's human. It's fine. This is the exact kind of behavior that people declared as being ProJo, yet apparently now it's back en vogue and coach is infallible.

[panda]

NET impact, probably about 6-8 spots. But as I expect us to be in, I'm more worried about our seeding.

Seeding has about an 85% correlation with the average of your resume metric average and quality metric average. Instead of climbing 4-5 in kenpom, which we would have had we held the 26-point lead, we dropped 7 spots. We also dropped in the other quality metrics.

Before GT, our average quality metric was 32.7, after GT it was 40.7. Because you are averaging that with the resume metrics, every 8 spots cut in half is 4, which equates to a full seed line drop (4 teams per line). Had MU won by 18, they would've stayed the same. Win by 26, they probably move up a full seed line. Win by 11 like they did, drop a seed line.

If Shaka changes the lineup when it gets to 20 and we hold there, no big deal, we stay where we're at. Because he let it go as long as he did, we dropped a line. I consider that a poor coaching decision, especially as we had timeouts and he brought the starters back anyway.

Just looking at before and after NET Georgetown blowouts and the first one I looked at was the 2-6 providence game. PC only moved up two spots.

UConn dropped from 15 to 17 after their blowout win on 1-25.

Seton hall moved up two spots from 43-41 following their blowout w against Georgetown on 2/1.

Unless I’m missing something, I don’t see how Marquette would’ve made such a jump in net had they won by a wider margin than they actually did.

Also it’s quite an assumption to believe they would’ve maintained the lead throughout the game.

[Its DJOver]

More people that don’t understand how NET works. How cute.

[panda]

More people that don’t understand how NET works. How cute.

Explain it to me like I’m 5. I’m happy to learn because clearly my interpretation is wrong.

[Its DJOver]

Explain it to me like I’m 5. I’m happy to learn because clearly my interpretation is wrong.

The five year old explanation is that its impossible to explain it because the formula isn't available.

However, based on what is known about NET, it can be said that the larger the data pool, the smaller effect that one singular data point can have.  This is why comparing games multiple weeks apart is not an apples to apples comparison.  A game in January accounts for a larger portion of the overall schedule (at the time) than a game in February.  Had NET been released at the time, our drop post St. Bonnies would have been even larger because it was not only a 16 point loss, but 1/6th of our total sample size.

Another misconception is that the distances between the rankings are all equal, they are not.  We currently have the 32nd best equation to the NET formula (the formula which is, again, unknown).  The margin between the 31st team, and the 33rd is unknown. This is why you can have impressive results that don't always equal a large increase in NET, as well is unimpressive results that may not account for a large decrease.  This is why losing games by less than you were expected to (like @PC) can have an increase in NET, as well as, winning games by less than you were supposed to (like Gtown) can cause your NET to drop. 

Efficiency is key.  If you lose, you want to lose as efficiently as possible (close game).  If you win, you want to win as efficiently as possible (blow out).  We had an opportunity for a blow out that would have almost surely improved our NET, and instead we had an inefficient finish to the game which resulted in a decrease in NET, despite the win.  It's impossible to know how much it effected our overall NET, but considering the scoring swing, as well as the poor opponent we were playing, it's fair to say that for a data point this late in the season, the last 12 minutes had a significant impact.

[panda]

The five year old explanation is that its impossible to explain it because the formula isn't available.

However, based on what is known about NET, it can be said that the larger the data pool, the smaller effect that one singular data point can have.  This is why comparing games multiple weeks apart is not an apples to apples comparison.  A game in January accounts for a larger portion of the overall schedule (at the time) than a game in February.  Had NET been released at the time, our drop post St. Bonnies would have been even larger because it was not only a 16 point loss, but 1/6th of our total sample size.

Another misconception is that the distances between the rankings are all equal, they are not.  We currently have the 32nd best equation to the NET formula (the formula which is, again, unknown).  The margin between the 31st team, and the 33rd is unknown. This is why you can have impressive results that don't always equal a large increase in NET, as well is unimpressive results that may not account for a large decrease.  This is why losing games by less than you were expected to (like @PC) can have an increase in NET, as well as, winning games by less than you were supposed to (like Gtown) can cause your NET to drop. 

Efficiency is key.  If you lose, you want to lose as efficiently as possible (close game).  If you win, you want to win as efficiently as possible (blow out).  We had an opportunity for a blow out that would have almost surely improved our NET, and instead we had an inefficient finish to the game which resulted in a decrease in NET, despite the win.  It's impossible to know how much it effected our overall NET, but considering the scoring swing, as well as the poor opponent we were playing, it's fair to say that for a data point this late in the season, the last 12 minutes had a significant impact.

thanks - So because there are more data points available after our Georgetown victory, the change would be even less than the movement of PC UCONN and seton hall?

[Its DJOver]

thanks - So because there are more data points available after our Georgetown victory, the change would be even less than the movement of PC UCONN and seton hall?

Not necessarily. Sample size is a factor yes, but not the only one.

Efficiently rules all as far as NET is concerned. Considering we both won the game and dropped in NET, it’s fair to say that the 15 point swing was significant.

Impossible to say how significant, it could have just been that the teams right behind us were just a hair behind us as far as the actual number that NET spits out, but significant nonetheless.

[5DollarPitcher]

As soon as you provide the example of this actually happening, you'll have a point. Until then, you're just defending a bad decision.

I don't get why some people get so defensive when it comes to defending Shaka. He makes mistakes. He's human. It's fine. This is the exact kind of behavior that people declared as being ProJo, yet apparently now it's back en vogue and coach is infallible.

The point is - you play to win the game. You win the game in front of you - nothing else matters. Can keep that philosophy all the way to a Final Four or National Championship.

And you bet your sweet ass I’m gonna be the most insufferable Pro Shaka person (whatever the term is now) on this board and possibly in the entire world. Wojo sucked, Shaka is good (and in record time) - I would like to twist that knife for former Projos as much as humanly possible.

[Vander Blue Man Group]

The point is - you play to win the game. You win the game in front of you - nothing else matters. Can keep that philosophy all the way to a Final Four or National Championship.

And you bet your sweet ass I’m gonna be the most insufferable Pro Shaka person (whatever the term is now) on this board and possibly in the entire world. Wojo sucked, Shaka is good (and in record time) - I would like to twist that knife for former Projos as much as humanly possible.

Except the problem is our seeding could be adversely impacted which could make winning as many games as we’d like more difficult.  So unfortunately, by how much you win the game DOES matter, no matter how much you whine that it doesn’t.

[brewcity77]

thanks - So because there are more data points available after our Georgetown victory, the change would be even less than the movement of PC UCONN and seton hall?

It's really not that easy to quantify. Kenpom, for example, is based on Adjusted Efficiency Margin, which is the difference between your adjusted offensive and defensive ratings.

Before Butler, we were at +16.75, which was 24th overall. We lost and fell to +15.86, 32nd. Then we beat Georgetown but because of the margin fell to +15.44, 39th.

We lost +1.31 in adjusted efficiency in 2 games, but had we flipped those results by equally positive scores and gone to +19.06, we would've only gone up to 19th, because there are bigger gaps in the top-20. For instance, climbing from 39 to 23 requires an improvement of +1.79 in AdjEM, but climbing from 15 to 14 requires an improvement of +1.91. It's why giving any definitive rank jumps are difficult, particularly once you get into the top-20. But because of that narrow gap from 39-23, we can move between there comparatively easily.

[panda]

Not necessarily. Sample size is a factor yes, but not the only one.

Efficiently rules all as far as NET is concerned. Considering we both won the game and dropped in NET, it’s fair to say that the 15 point swing was significant.

Impossible to say how significant, it could have just been that the teams right behind us were just a hair behind us as far as the actual number that NET spits out, but significant nonetheless.

Understood - given the unknown and most likely small impact, Shaka probably made the correct basketball decision. And in the grand scheme of things will most likely help the overall goal of the team.

[Boston Warrior]

How correlated is net rank to seeding?

[DoctorV]

How correlated is net rank to seeding?

This guy!
Very correlated.
Find the NET rankings online today and check out the reveal.

Speaking of, what time is that?

[brewcity77]

How correlated is net rank to seeding?

It's a decent analogue, but the best comparison is the average between the resume averages and quality averages on the team sheet. Teams land within a seed line of that average about 85% of the time.

For instance, Marquette has the following ranks currently:

Resume -- KPI 21, SOR 29, Average 25
Quality -- BPI 45, Kenpom 33, Sagarin 36, Average 38

The average of those two numbers is 31.5, which means Marquette would typically be on the edge of the 8 (teams 29-32) and 9 (teams 33-36) lines. Marquette is generally being seeded about a line or two higher than that because of our volume of Q1 wins.

Before Georgetown, that average was 25.6, which would usually be around the 6 (teams 21-24) and 7 (teams 25-28) lines.

[CountryRoads]

Not necessarily. Sample size is a factor yes, but not the only one.

Efficiently rules all as far as NET is concerned. Considering we both won the game and dropped in NET, it’s fair to say that the 15 point swing was significant.

Impossible to say how significant, it could have just been that the teams right behind us were just a hair behind us as far as the actual number that NET spits out, but significant nonetheless.

The NET is pretty flawed. Look at Iowa for example. 0-6 in Q1 with 8 losses overall and their NET is 21. How is that possible? Because they won their 8 buy games by an average of about 30 this year. Drop that average to 15 and they are probably on the outside looking in at this point. All because of how bad they beat up on inferior opponents.

[Its DJOver]

Understood - given the unknown and most likely small impact, Shaka probably made the correct basketball decision. And in the grand scheme of things will most likely help the overall goal of the team.

Grand scheme of things will it make a large difference, probably not. Missed opportunity would probably be the words I would describe it as, when you have the chance to step on a teams throat and kick them when they’re down you have to take it. In approximately 1/35 of our pre ncaa tourney net calculations, we didn’t seize that chance. Guess that just means we’ll have to blow out Creighton.

[Its DJOver]

The NET is pretty flawed. Look at Iowa for example. 0-6 in Q1 with 8 losses overall and their NET is 21. How is that possible? Because they won their 8 buy games by an average of about 30 this year. Drop that average to 15 and they are probably on the outside looking in at this point. All because of how bad they beat up on inferior opponents.

Flawed yes, I haven’t seen anyone say it’s perfect. Significant improvement over previously used metrics, also yes.

No algorithm is perfect, this one is better than ones previously used.

[Elonsmusk]

Scholl better call Shaka into his office for a reprimanding from Brewcity77 for his horrendous coaching that has landed him on the Naismith Coach of the Year Watchlist.

Scoop is filled with absurd takes, but whining about Shaka's coaching in the Georgetown game is among the most absurd.

1 seed line in the NCAA isn't going to decide if MU wins its first round game.  Matchups are everything in the NCAA tournament. If we get a 6 seed playing an 11 with physical big men, strong on the boards, that won't be a good matchup.  If we get an 8 seed with a matchup against a 9 that is weak on the boards and more perimeter oriented, that would likely yield a better probability for a win.

This season has been a massive success and energized the fanbase that was relegated to apathy after 7 years of the underachieving meathead.  Enjoy this and chill the fug out about the NET.

[Goose]

Elon

Half of these folks thought Shaka put together a NIT bubble team and now are criticizing coaching moves, which makes me laugh. I would recommend that most on here should just be thanking Shaka for a great season and enjoy the run. I want this season to keep rolling along, but I am perfectly fine with Shaka mixing things up, regardless of what it does to their seeding.

[tower912]

Goose, I am relaxed and enjoying the ride.   But it appears to me that these are the same criticisms that would have been made about previous coaches in similar situations.   As it should be.

[Vander Blue Man Group]

Elon

Half of these folks thought Shaka put together a NIT bubble team and now are criticizing coaching moves, which makes me laugh. I would recommend that most on here should just be thanking Shaka for a great season and enjoy the run. I want this season to keep rolling along, but I am perfectly fine with Shaka mixing things up, regardless of what it does to their seeding.

You realize thanking Shaka and being thrilled at what’s he’s accomplished this year and questioning a decision he made during a game are not mutually exclusive, right?

[Goose]

Tower

We have an experienced HC, not a journeyman assistant or a wet behind the ear's assistant and things have to be looked at differently. I went hard on here about the first Creighton game because I felt Shaka blew it, playing a group of young guys against a bunch of stiffs is not high on my list of things to get upset over. Plus, I believe they climb the seeding ladder beginning tomorrow and not worried about a supposed bracket in mid-February. Win games and those twelve minutes will mean nothing in three weeks.

Vander
If that lineup had performed better, everyone would be doing cartwheels because of their performance. I had zero problem with the lineup, modest problems with performance and major problems watching it.

[Elonsmusk]

Elon

Half of these folks thought Shaka put together a NIT bubble team and now are criticizing coaching moves, which makes me laugh. I would recommend that most on here should just be thanking Shaka for a great season and enjoy the run. I want this season to keep rolling along, but I am perfectly fine with Shaka mixing things up, regardless of what it does to their seeding.

Once again Goose...your perspective is right on target.