Gotta love Kenpom

[Piratz]

Seton Hall should be in the NCAA Tournament let’s rejoice.

Been a crazy season. We didn’t pass the eye test in the non-con or half the time on the road with 5 BE losses by 20+ yikes. Then we had 5 BE home wins by 10+. 5 games over .500 but only +10 total point differential. We are one of those gritty teams with good record, mediocre metrics. It happens. I’ll take it.

 

[HallX2]

ThEy ArE MoRE eFfiCiEnT WhEn ThEy LoSe

Well let’s hope the tourney isn’t limited to the most efficient in losing.

 

[LeftCoastPirates]

How do u defend xavier and maryland at .500 w a higher net? How is st johns and nova 30 spots higher?

I did say there are outliers, but I’ll play along.

Remember it’s a computer no human evaluation for all these one off subjective variables people want it to consider. And for the last time it’s NOT a ranking system.

1. Maryland is 72 in the NET, so what the heck are people complaining about? That’s average at best. Look at most teams in the 70-80 range and you are like MEH.
But from a computer perspective 12 of their 15 losses have been by single digits and 10 have been 5 points or less.
So the predictive metrics are saying Maryland is a more competitive team than its .500 record.

2. Xavier at 62 is merely a reflection of their overall NET SOS being 6th in the country. They are 7-13 in Q1/2… we are 9/10 in Q1/2. From a predictive analysis perspective they split with SHU showing that from a competitive perspective Xavier is not a loser type team like DePaul / Gtown.

3. St John’s. They are going to finish 11-9 in the BE, only two games worse than us. So we beat them twice and people want that to matter. But it’s a computer so it’s not isolating those two games in a bubble.
SJU is 7-8 road neutral (stronger value for those games) and in those games they were really competitive. 4 point loss @ UConn, 1 point loss at CU, 3 point loss @ PU. 7 of their nonconference games were vs teams 150 or better. And they actually beat 48. And only 2 were absolute dogs in the high 200+ range. Conversely we only had 4 games inside the top 150 and lost ALL of them.
The predictive metrics say SJU can hang with you road or away and not get blow out, hence on a given night they probably are expected to do better (not necessarily win) than SHU especially on the road.

4. Villanova. Everyone wants to point to their Big 5 losses and bury them. Well Drexel (122) and St Joes (101) are ranked well ahead of Missouri (158) who was our best OOC win. Penn at 202 is another black eye. But they make up for it by beating Maryland (72) Memphis (71) Texas Tech (35) and UNC (10). They did their work in the nonconference and we didn’t.
And when they have won in BE play they have been very efficient. 8 double digit victories. We best Gtown by a combined 10 points. Nova beat them by 44. They also have a Providence road win and one @ Creighton as well. The only difference is we have a UConn win they don’t. But they only lost by 1 to them, so the computer likes that.

From a computer perspective these teams are where they should be. Or slightly elevated for the factors mentioned above.

 

[Piratz]

ThEy ArE MoRE eFfiCiEnT WhEn ThEy LoSe

Basically. Bc if you’re good you don’t get blown out. And when you get blown out often you just have fundamental flaws.

There are teams like SHU almost every year.

 

[Fishjam]

Seton Hall should be in the NCAA Tournament let’s rejoice.

Been a crazy season. We didn’t pass the eye test in the non-con or half the time on the road with 5 BE losses by 20+ yikes. Then we had 5 BE home wins by 10+. 5 games over .500 but only +10 total point differential. We are one of those gritty teams with good record, mediocre metrics. It happens. I’ll take it.

You hit upon a very good point - Point Differential. For all of the Metrics used by KP and NET it basically comes down to that. They give weight to strength of opponent and home/road/neutral and Win or Lose but it heavily weighs point differential. It's what all the other metrics boil down to.

Here are the KP & NET rankings and just the Big East point differentials.

2/3 UConn +263

10/11 Creighton +154
14/14 Marquette +142

27/37 St Johns +95
29/31 Villanova +80

59/61 Providence +12
56/63 Seton Hall +10
54/62 Xavier +2
62/65 Butler -44

198/204 GTown -257
315/322 DePaul -454

 

[dehere23]

I did say there are outliers, but I’ll play along.

Remember it’s a computer no human evaluation for all these one off subjective variables people want it to consider. And for the last time it’s NOT a ranking system.

1. Maryland is 72 in the NET, so what the heck are people complaining about? That’s average at best. Look at most teams in the 70-80 range and you are like MEH.
But from a computer perspective 12 of their 15 losses have been by single digits and 10 have been 5 points or less.
So the predictive metrics are saying Maryland is a more competitive team than its .500 record.

2. Xavier at 62 is merely a reflection of their overall NET SOS being 6th in the country. They are 7-13 in Q1/2… we are 9/10 in Q1/2. From a predictive analysis perspective they split with each other showing that from a competitive perspective Xavier is not a loser type team like DePaul / Gtown.

3. St John’s. They are going to finish 11-9 in the BE, only two games worse than us. So we beat them twice and people want that to matter. But it’s a computer so it’s not isolating those two games in a bubble.
SJU is 7-8 road neutral (stronger value for those games) and in this games they were really competitive. 4 point loss @ UConn, 1 point loss at CU, 3 point loss @ PU. 7 of their nonconference games were vs teams 150 or better. And they actually beat 48. And only 2 were absolute dogs in the high 200+ range. Conversely we only had 4 games inside the top 150 and lost ALL of them.
The predictive metrics say SJU can hang with you road or away and not get blow out, hence on a given night they probably are expected to do better (not necessarily win) than SHU especially on the road.

4. Villanova. Everyone wants to point to their Big 5 losses and bury them. Well Drexel (122) and St Joes (101) are ranked well ahead of Missouri (158) who was our best OOC win. Penn at 202 is another black eye. But they make up for it by beating Maryland (72) Memphis (71) Texas Tech (35) and UNC (10). They did their work in the nonconference and we didn’t.
And when they have won in BE play they have been very efficient. 8 double digit victories. We best Gtown by a combined 10 points. Nova beat them by 44. They also have a Providence road win and one @ Creighton as well. The only difference is we have a UConn win they don’t. But they only lost by 1 to them, so the computer likes that.

From a computer perspective these teams are where they should be. Or slightly elevated for the factors mentioned above.

Stop being objective. With hard data to back it up.

Who wants to guess that if the NET or certain other metrics were more favorable to Seton Hall this year, there wouldn’t be such an outcry as the past few weeks?

Ultimately it doesn’t matter. But that’s the reason why this is generating so much discussion here. I certainly would much prefer to get in the dance with less favorable metrics than not get in with more favorable ones. So it’s all good from the Hall perspective given Wednesday’s W.

 

[Fishjam]

Interesting when you look at our KP ratings over the last 9 years since we became relevant, we've been rated between 51-60 6 times and 20-29 3 times. The last 4 seasons being extremley similar.

56 - 58 - 52 - 54 - 20 - 60 - 26 - 51 - 29

 

[LeftCoastPirates]

Interesting when you look at our KP ratings over the last 9 years since we became relevant, we've been rated between 51-60 6 times and 20-29 3 times. The last 4 seasons being extremley similar.

56 - 58 - 52 - 54 - 20 - 60 - 26 - 51 - 29

Also not really eye opening.

29 - whitehead team
26 - KC / Delgado/ Desi / Ish senior year
20 - Powell - COVID cancellation

You think these teams were more efficient offensively and defensively than those others. 🤔

If a computer were predicting a more competitive SHU team, shocking that it would favor those 3 teams over the others recently.

 

[shupat08]

Basically. Bc if you’re good you don’t get blown out. And when you get blown out often you just have fundamental flaws.

There are teams like SHU almost every year.

So then Nova, StJ, and X are "good"? But we have better overall and BE records than them? And took 4 of 6 from those teams?

 

[sussexcopirate]

I did say there are outliers, but I’ll play along.

Remember it’s a computer no human evaluation for all these one off subjective variables people want it to consider. And for the last time it’s NOT a ranking system.

1. Maryland is 72 in the NET, so what the heck are people complaining about? That’s average at best. Look at most teams in the 70-80 range and you are like MEH.
But from a computer perspective 12 of their 15 losses have been by single digits and 10 have been 5 points or less.
So the predictive metrics are saying Maryland is a more competitive team than its .500 record.

2. Xavier at 62 is merely a reflection of their overall NET SOS being 6th in the country. They are 7-13 in Q1/2… we are 9/10 in Q1/2. From a predictive analysis perspective they split with SHU showing that from a competitive perspective Xavier is not a loser type team like DePaul / Gtown.

3. St John’s. They are going to finish 11-9 in the BE, only two games worse than us. So we beat them twice and people want that to matter. But it’s a computer so it’s not isolating those two games in a bubble.
SJU is 7-8 road neutral (stronger value for those games) and in those games they were really competitive. 4 point loss @ UConn, 1 point loss at CU, 3 point loss @ PU. 7 of their nonconference games were vs teams 150 or better. And they actually beat 48. And only 2 were absolute dogs in the high 200+ range. Conversely we only had 4 games inside the top 150 and lost ALL of them.
The predictive metrics say SJU can hang with you road or away and not get blow out, hence on a given night they probably are expected to do better (not necessarily win) than SHU especially on the road.

4. Villanova. Everyone wants to point to their Big 5 losses and bury them. Well Drexel (122) and St Joes (101) are ranked well ahead of Missouri (158) who was our best OOC win. Penn at 202 is another black eye. But they make up for it by beating Maryland (72) Memphis (71) Texas Tech (35) and UNC (10). They did their work in the nonconference and we didn’t.
And when they have won in BE play they have been very efficient. 8 double digit victories. We best Gtown by a combined 10 points. Nova beat them by 44. They also have a Providence road win and one @ Creighton as well. The only difference is we have a UConn win they don’t. But they only lost by 1 to them, so the computer likes that.

From a computer perspective these teams are where they should be. Or slightly elevated for the factors mentioned above.

The teams you mentioned are all greatly inflated in NET versus their RPI. You can keep defending the NET but there is a reason it replaced RPI. That reason being the major conference teams want more teams in the dance and the NET gives them a crutch to do it.

 

[hallstorm]

lol. I don’t think the metrics and data based systems are necessarily objective despite what posters say. It’s simply a difference of opinion.

Because it’s all about qualifying for the NCAA tournament and that is about winning a 1 game playoff every time. Not a series that plays the odds or the more efficient team.

The metrics will always favor the more talented teams. It’s built that way. So a player that blows out teams 2 out of 3 times and then loses a nail biter is loved by the metrics.

But it doesn’t account for clutch play, injuries during the season or stepping up against the big dogs. It doesn’t like the giant killers come tourney time.

Maybe we are that. Maybe not.

I think the metrics are created because everything is done that way nowadays. And it’s like the WAR stat in baseball or the PER in the NBA.

But I think it needs serious refinement.

 

[hallstorm]

I’ll say one other thing:

I’m no longer a tennis fan, but I used to be.

I always wondered how Agassi would lose an in-between set 6-0, 6-1 or something and then go on to win the match. It’s like he took off a set to regroup or just decided it wasn’t worth expanding too much energy in order for the optics to look better.

It was a strategy that worked for him🤔 clearly.

Metrics HATE that kind of strategy.

 

[Halldan1]

You hit upon a very good point - Point Differential. For all of the Metrics used by KP and NET it basically comes down to that. They give weight to strength of opponent and home/road/neutral and Win or Lose but it heavily weighs point differential. It's what all the other metrics boil down to.

Here are the KP & NET rankings and just the Big East point differentials.

2/3 UConn +263

10/11 Creighton +154
14/14 Marquette +142

27/37 St Johns +95
29/31 Villanova +80

59/61 Providence +12
56/63 Seton Hall +10
54/62 Xavier +2
62/65 Butler -44

198/204 GTown -257
315/322 DePaul -454

Not surprising. But I thought winning the games playing the exact same schedule in conference was more important than point differential. Otherwise forget about the 40 minutes games and just list the teams as regards points for vs points against.

Same goes for the OOC metrics. The only results that matter are the point differentials???

Any wonder why I so hate metrics.

Going by memory......Thinking of the first UConn game. Wasn't the Hall up by nearly 30 points in the second half and then foot off the gas we won by 15.

In the Wagner game Sha emptied the bench very early and the end of the bench played major minutes while Copeland pressed that entire time keeping the score somewhat respectable.

Have we reached the point where keeping your best players on the court nearly the entire game to win by 30 is more important than respecting your opponent when the outcome is no longer in doubt and you want to reward your bench? I'm looking at you Jay Wright, Greg McDermott and others.

Unfortunately that is what's happening in the metrics era.

 

[Halldan1]

And I say the above because metrics are probably going to force Holloway to try and win by 25+ points or maybe drop lower on the bubble.

Is that what sports should be about?

 

[Seton75]

... And when you get blown out often you just have fundamental flaws....

Like the 1960 Pirates.

 

[hallwins]

I’ll say one other thing:

I’m no longer a tennis fan, but I used to be.

I always wondered how Agassi would lose an in-between set 6-0, 6-1 or something and then go on to win the match. It’s like he took off a set to regroup or just decided it wasn’t worth expanding too much energy in order for the optics to look better.

It was a strategy that worked for him🤔 clearly.

Metrics HATE that kind of strategy.

Feel like we subconsciously did that in our last 3 games. Compete, conserve, win.

 

[Hall84]

St John's has always lost well.
Lots of practice the last 20 years!

 

[Fishjam]

Not surprising. But I thought winning the games playing the exact same schedule in conference was more important than point differential. Otherwise forget about the 40 minutes games and just list the teams as regards points for vs points against.

Same goes for the OOC metrics. The only results that matter are the point differentials???

Any wonder why I so hate metrics.

Thinking of the first UConn game. Wasn't the Hall up by nearly 30 points in the second half and then foot off the gas we won by 15.

In the Wagner game Sha emptied the bench very early and the end of the bench played major minutes while Copeland pressed that entire time keeping the score somewhat respectable.

Have we reached the point where keeping your best players on the court nearly the entire game to win by 30 is more important than respecting your opponent when the outcome is no longer in doubt and you want to reward your bench? I'm looking at you Jay Wright, Greg McDermott and others.

Unfortunately that is what's happening in the metrics era.

That's the argument for adjusting the metrics. Give more weight to winning or losing and less weight on metrics compiled in lopsided games.

 

[dehere23]

And I say the above because metrics are probably going to force Holloway to try and win by 25+ points or maybe drop lower on the bubble.

Is that what sports should be about?

I just heard Pitino say the same thing as you in a press conference, even though SJU has pretty favorable metrics. Said against Depaul they played their bench more than the starters because that's what you do in those types of games, but if he had kept his starters in and won by like 50 or 60 instead of by 30, it would have impacted metrics.

 

[Halldan1]

And the beat goes on.

After last night's blowout not surprisingly the Pirates dropped a spot in Kenpom where winning doesn't matter if the win isn't done at the liking of the computer metric popularized to the highest degree nowadays by Ken Pomeroy.

54 Xavier 15-16 9-11 +13.89

57 Seton Hall 20-11 13-7 +13.44

 

[shupat08]

And the beat goes on.

After last night's blowout not surprisingly the Pirates dropped a spot in Kenpom where winning doesn't matter if the win isn't done at the liking of the computer metric popularized to the highest degree nowadays my Ken Pomeroy.

54 Xavier 15-16 9-11 +13.89

57 Seton Hall 20-11 13-7 +13.44

But they finished the season under .500 in such an efficient way Dan. That PROVES they are a good team. Unlike us, just winning 20 games 13 in the BE in an inefficient, bad way.

 

[sussexcopirate]

And the beat goes on.

After last night's blowout not surprisingly the Pirates dropped a spot in Kenpom where winning doesn't matter if the win isn't done at the liking of the computer metric popularized to the highest degree nowadays by Ken Pomeroy.

54 Xavier 15-16 9-11 +13.89

57 Seton Hall 20-11 13-7 +13.44

Dan, I can't believe you don't understand this by now. It's because Xavier lost by an average of 9.8967 points in their losses and was undefeated on Wednesday tipoffs at 645 PM while Seton Hall only beat Klingan because he got hurt. It's just metrics, baby!

 

[hallwins]

Dan, I can't believe you don't understand this by now. It's because Xavier lost by an average of 9.8967 points in their losses and was undefeated on Wednesday tipoffs at 645 PM while Seton Hall only beat Klingan because he got hurt. It's just metrics, baby!

Actually the system doesn't take into account such injuries, like our 2 losses without Khadary! Or our win at PC without Harper.

Metrics have a place for sure and they are reliable, but to continue to defend without admitting any holes in the system, makes no sense. There are always flaws. Like when did you play the team that handed you the poor metric loss?

Some post the other day actually stated:

We beat St John's twice so you want to say we are better.

Of course we do! Because we are.

You know who else would say that?
- Rick Pitino.

 

[Halldan1]

More Kenpom

8 losses in their last 11 games with the three victories at home against Rutgers, Maryland and Ohio State and they fall from #11 to #21

21

Wisconsin

B10

19-12

+19.00

 

[Fishjam]

Hey, at least we separated from the other BE teams in the NET. Moved up from 63 to 61 jumping ahead of Providence and Xavier.

61 Seton Hall
64 Providence
65 Xavier
66 Butler

 

[Fishjam]

BTW, interesting tidbit on KenPom. We began the season ranked #56 by KP. After our entire 31 game season we are now #57.

 

[sami]

I did say there are outliers, but I’ll play along.

Remember it’s a computer no human evaluation for all these one off subjective variables people want it to consider. And for the last time it’s NOT a ranking system.

1. Maryland is 72 in the NET, so what the heck are people complaining about? That’s average at best. Look at most teams in the 70-80 range and you are like MEH.
But from a computer perspective 12 of their 15 losses have been by single digits and 10 have been 5 points or less.
So the predictive metrics are saying Maryland is a more competitive team than its .500 record.

2. Xavier at 62 is merely a reflection of their overall NET SOS being 6th in the country. They are 7-13 in Q1/2… we are 9/10 in Q1/2. From a predictive analysis perspective they split with SHU showing that from a competitive perspective Xavier is not a loser type team like DePaul / Gtown.

3. St John’s. They are going to finish 11-9 in the BE, only two games worse than us. So we beat them twice and people want that to matter. But it’s a computer so it’s not isolating those two games in a bubble.
SJU is 7-8 road neutral (stronger value for those games) and in those games they were really competitive. 4 point loss @ UConn, 1 point loss at CU, 3 point loss @ PU. 7 of their nonconference games were vs teams 150 or better. And they actually beat 48. And only 2 were absolute dogs in the high 200+ range. Conversely we only had 4 games inside the top 150 and lost ALL of them.
The predictive metrics say SJU can hang with you road or away and not get blow out, hence on a given night they probably are expected to do better (not necessarily win) than SHU especially on the road.

4. Villanova. Everyone wants to point to their Big 5 losses and bury them. Well Drexel (122) and St Joes (101) are ranked well ahead of Missouri (158) who was our best OOC win. Penn at 202 is another black eye. But they make up for it by beating Maryland (72) Memphis (71) Texas Tech (35) and UNC (10). They did their work in the nonconference and we didn’t.
And when they have won in BE play they have been very efficient. 8 double digit victories. We best Gtown by a combined 10 points. Nova beat them by 44. They also have a Providence road win and one @ Creighton as well. The only difference is we have a UConn win they don’t. But they only lost by 1 to them, so the computer likes that.

From a computer perspective these teams are where they should be. Or slightly elevated for the factors mentioned above.

Don’t know why you keep trying to fight this. Some people will just never get it. My favorite is when statements are made that basically ask how Ken Pomeroy ranks something when it’s a friggin mathematical model.

No formula is perfect. There will be anomalies on both sides. You could be sure as shit if we were in Villanova’s shoes, the same people would be talking about how we’re better than our record.

 

[HALL85]

Not surprising. But I thought winning the games playing the exact same schedule in conference was more important than point differential. Otherwise forget about the 40 minutes games and just list the teams as regards points for vs points against.

Same goes for the OOC metrics. The only results that matter are the point differentials???

Any wonder why I so hate metrics.

Thinking of the first UConn game. Wasn't the Hall up by nearly 30 points in the second half and then foot off the gas we won by 15.

In the Wagner game Sha emptied the bench very early and the end of the bench played major minutes while Copeland pressed that entire time keeping the score somewhat respectable.

Have we reached the point where keeping your best players on the court nearly the entire game to win by 30 is more important than respecting your opponent when the outcome is no longer in doubt and you want to reward your bench? I'm looking at you Jay Wright, Greg McDermott and others.

Unfortunately that is what's happening in the metrics era.

I completely agree. And what about developing players? It’s not even about emptying the bench. It’s about giving valuable time to kids that have the potential and need to learn in game conditions.

 

[Fishjam]

6 BE teams saw less than 6 spot change between their Preseason KP and end of regular season KP ranking. 4 teams moved up or down by 20-35 spots and then there is DePaul.

Uconn 4 - 2 (+2)
Creighton 12 - 10 (-2)
Marquette 11 - 14 (-3)
Seton Hall 56 - 57 (-1)
Villanova 23 - 29 (-6)
Providence 54 - 59 (-5)

St Johns 59 - 27 (+32)
Xavier 34 - 54 (-20)
Butler 96 - 66 (+30)
Georgetown 163 - 198 (-35)

DePaul 113 - 315 (-203)

 

[Halldan1]

Here's another. Started #13. 13 losses later dropped 6 spots.

Michigan St. (18-13)​

19

Michigan St.

B10

18-13

+19.79

 

[PirateEmotion]

I am the most anti-metrics poster on this board, and think they have no place in selection/seeding for the tournament. It is insane that in college basketball computer based algorithms trump the actual results.

With that said, predictive metrics are exactly that. If two teams were to play on a neutral court, what would the betting line be? Seton Hall would be a neutral court underdog against probably all other “bubble teams”. (NET, Kenpom, BPI)

Seton Hall’s on court record is better than all other bubble teams. (Strength of Record)

Based upon on court results which is at the core of basketball and athletics in general, SHU should be safely in the field. Their results are definitively in the top 36 for an at-large bid. Results are all that should matter in my opinion; that is the beauty of sports.

I do see myself continuing to struggle as the years go by with an ever intensifying case of metrics fatigue. I’m sure the Florida State football playoff incident is just the tip of this devastating iceberg.

Go Hall.

 

[walshtrips]

Never once have I said we don’t deserve a bid. Those quality wins will put us over the top.

I have never once said the NET should be used as a ranking tool. And for the record it is NOT.

I get what you're saying. I just don't get how Villanova with three horrible losses in their own backyard, and fewer Quad 1 wins than SHU, could be ranked so far ahead of us. The NCAA admits that Quad 3 losses are difficult to overcome, yet Nova with 3 sits at a NET 31.

 

[hallwins]

I have read that SOR may be the metric the Committee most relies upon. Our SOR is better than all the BE bubble teams.

 

[Halldan1]

And this should be interesting. Maryland getting destroyed by 14-16 Penn State and will drop to below .500.

As of right now

53

Maryland

B10

15-15

+14.03

57

Seton Hall

BE

20-11

+13.35

 

[Halldan1]

Final

• 
  Maryland
  69
• 
  Penn State
  85

Box

Penn State 85-69 Maryland (Mar 10, 2024) Box Score - ESPN

Box score for the Penn State Nittany Lions vs. Maryland Terrapins NCAAM game from March 10, 2024 on ESPN. Includes all points, rebounds and steals stats.

www.espn.com

 

[shupat08]

Final

• 
  Maryland
  69
• 
  Penn State
  85

Box

Penn State 85-69 Maryland (Mar 10, 2024) Box Score - ESPN

Box score for the Penn State Nittany Lions vs. Maryland Terrapins NCAAM game from March 10, 2024 on ESPN. Includes all points, rebounds and steals stats.

www.espn.com

UMD, RU, UMich are the last 3 teams in the Big14 standings.

 

[Halldan1]

53 Maryland	15-15	7-13		+14.03
97 Rutgers	15-16	7-13		+8.03
127 Michigan	8-23	3-17		+4.00

 

[garyshu71]

A lot of hot seats out there!

 

[TheHall87]

Going by memory....Thinking of the first UConn game. Wasn't the Hall up by nearly 30 points in the second half and then foot off the gas we won by 15.

Our biggest lead was 19 which occurred with 86 seconds to play.

We didn't save much gas.

 

[shupat08]

Hey, at least we separated from the other BE teams in the NET. Moved up from 63 to 61 jumping ahead of Providence and Xavier.

61 Seton Hall
64 Providence
65 Xavier
66 Butler

Xavier also lost in the OOC to Delaware and Oakland. And they finished under .500 and only finished 4 spots behind us? Something doesn't seem to add up based on the metrics bros "SeToN HaLL hAd HoRRiD OOC LoSeS" argument.....

I guess a 4 game BE difference and 5 game overall difference only results in a 4 spot analytics difference 😂