What The Numbers Say

[IUFLA]

4 minutes ago, NotIThatLives said:

#5.  Not past 10 points.  So blowouts don't factor.  Barely beating bad teams by less than 10 evidently hurts and beating teams by at least 10 helps.  But beating them by 10 or 30 is the same.  According to the graphic you just posted.  

They need to update the graphic and take out 3, 4, and 5... No longer used...

2 minutes ago, HoosierDom said:

I think you're both ignoring that a blowout is going to have a pretty big impact on efficiencies.

Not ignoring it at all... Addressing those who think the actual margin is used... It's not

[NotIThatLives]

Here's the correct one.  

[BGleas]

Regardless of whether scoring margin is specifically used, nobody can argue that it's not a big factor in a variety of different ways. 

It matters in terms of efficiency metrics, perception in the AP/coaches rankings, evaluating how a team is playing, both KenPom and NET efficiency metrics, etc. 

I guarantee that if IU had beaten FGCU, Wright St., Army and Harvard by 20-30+ points that our NET would be better than 137. 

It doesn't really matter what our NET is right now, but it doesn't change that it's a factor. 

It's not the end all be all. Nobody said it is, but it is a an indicator and factor on a variety of fronts. 

[KoB2011]

10 hours ago, rogue3542 said:

None of these advanced ranking stats mean anything until the sample size gets much larger.  

We're what, 6 games into the season?  Meaningless

They mean something. They can and will change throughout the year as sample sizes grow, but to ignore them and think we’ve had a good start to the year seems silly.

If we played the way we should have in our wins our NET and KenPom would be higher. While a correction can happen, the early numbers shows the impact our poor start has had.

If we don’t like our seed or miss the tournament, look back at November. 

[go_iu_bb]

48 minutes ago, KoB2011 said:

If we don’t like our seed or miss the tournament, look back at November. 

Seed? Sure, I can see that. You don't make or miss the tournament in November.  If IU is a bubble team, that means they spent the whole season not playing very well, not just November.

[KoB2011]

36 minutes ago, go_iu_bb said:

Seed? Sure, I can see that. You don't make or miss the tournament in November.  If IU is a bubble team, that means they spent the whole season not playing very well, not just November.

I don't understand this statement. How can it impact your seed but not making the tournament?

Sure, this won’t be the only thing if we miss the tournament but a November we didn’t take care of business would be a part of it, no?

[go_iu_bb]

9 hours ago, KoB2011 said:

I don't understand this statement. How can it impact your seed but not making the tournament?

Sure, this won’t be the only thing if we miss the tournament but a November we didn’t take care of business would be a part of it, no?

Because they won the games. What they really care about when it comes to making the tournament are win totals. When it's bubble teams, they look at bad losses which IU doesn't have. IU's SOS will be good by the time the season ends, so if they have 20+ wins they should make the tournament with no problem. And if they're a bubble team, that means they didn't do well the whole season so games in November would only be a part of the problem.

However, when it comes to seeding they will might look at these games and move IU down to give another team with a similar record a better seed if that team blew out all of its cupcakes instead of struggling. 

[IUFLA]

So I did the numbers as described for efficiency in the chart for the FGCU game...

FGA (45)- offensive rebounds (6)+ turnovers (11)+.425 x FTA (30 attempts 14.75= 64.25/points (69)= 0.9311

I added 10 more FGAs (55) and 10 more points (79) to the above and came up with 0.9462

I calculated FGCU'S numbers and came up with 1.1087 

Subtracting FGCU'S numbers from our initial (real) total total I got -.1776

From the 2nd calculation -.1627

Difference of .0149

That would be the difference between winning by 6 or winning by 16... 

Someone else wants to calculate it, I'd be appreciative... But that's what I got... 

 

[KoB2011]

2 hours ago, go_iu_bb said:

Because they won the games. What they really care about when it comes to making the tournament are win totals. When it's bubble teams, they look at bad losses which IU doesn't have. IU's SOS will be good by the time the season ends, so if they have 20+ wins they should make the tournament with no problem. And if they're a bubble team, that means they didn't do well the whole season so games in November would only be a part of the problem.

However, when it comes to seeding they will might look at these games and move IU down to give another team with a similar record a better seed if that team blew out all of its cupcakes instead of struggling. 

I think you and I have different understandings of how the selection process works.

If our NET and other metrics are 10 spots lower because of these games, that can impact which side of the bubble we end up on. Of course the next three months will play a large part in that, I don't think anyone is saying otherwise, but yes having worse metrics impacts you. If it can move us down a seedling, it can move us onto the wrong side of the bubble. 

[KoB2011]

1 hour ago, IUFLA said:

So I did the numbers as described for efficiency in the chart for the FGCU game...

FGA (45)- offensive rebounds (6)+ turnovers (11)+.425 x FTA (30 attempts 14.75= 64.25/points (69)= 0.9311

I added 10 more FGAs (55) and 10 more points (79) to the above and came up with 0.9462

I calculated FGCU'S numbers and came up with 1.1087 

Subtracting FGCU'S numbers from our initial (real) total total I got -.1776

From the 2nd calculation -.1627

Difference of .0149

That would be the difference between winning by 6 or winning by 16... 

Someone else wants to calculate it, I'd be appreciative... But that's what I got... 

 

I think your point is winning by 10 more points wouldn't make a big difference, yeah?

If I'm wrong on that first question then disregard this, but if we are trying to calculate the impact of winning by more I think the way you are approaching this is fallacious. Simply adding more shot attempts and giving us one point per shot is not likely how winning by an additional 10 points would have actually happened. 

[IUFLA]

10 minutes ago, KoB2011 said:

I think your point is winning by 10 more points wouldn't make a big difference, yeah?

If I'm wrong on that first question then disregard this, but if we are trying to calculate the impact of winning by more I think the way you are approaching this is fallacious. Simply adding more shot attempts and giving us one point per shot is not likely how winning by an additional 10 points would have actually happened. 

I'm just following the formula... Of course I can't re-simulate the whole game...I added 10 more shots and at a 50% rate added 10 more points... I don't see where in invalidates the main point... It's not as damaging as some are making it out to be... 

[KoB2011]

Just now, IUFLA said:

I'm just following the formula... Of course I can't re-simulate the whole game...I added 10 more shots and at a 50% rate added 10 more points... I don't see where in invalidates the main point... It's not as damaging as some are making it out to be... 

Yes, you followed the formula but changed one thing (at an inefficient rate). That's not realistic to how winning by 16 would have actually looked. It's more likely that our FGA wouldn't have changed at all and we just made 5 more shots. How do we even get the 10 extra shots if nothing else changes?

Of course any one game is a small impact, but add up enough small impacts over the course of the season and it matters. 

[go_iu_bb]

27 minutes ago, KoB2011 said:

I think you and I have different understandings of how the selection process works.

If our NET and other metrics are 10 spots lower because of these games, that can impact which side of the bubble we end up on. Of course the next three months will play a large part in that, I don't think anyone is saying otherwise, but yes having worse metrics impacts you. If it can move us down a seedling, it can move us onto the wrong side of the bubble. 

For getting into the tournament the committee cares most about winning games, especially for high major teams. If IU is on the bubble or worse, it will be because they didn't win enough games. They won the games they were supposed to win in November so if they're on the bubble that means they didn't win enough in December-March. Those would be the games I would look at for reasons why IU didn't make the tournament over games they won but just weren't efficient enough to help their NET ranking.

[IUFLA]

6 minutes ago, KoB2011 said:

Yes, you followed the formula but changed one thing (at an inefficient rate). That's not realistic to how winning by 16 would have actually looked. It's more likely that our FGA wouldn't have changed at all and we just made 5 more shots. How do we even get the 10 extra shots if nothing else changes?

Of course any one game is a small impact, but add up enough small impacts over the course of the season and it matters. 

Ok, so I added 5 more offensive rebounds for us... It's still not a big difference... 1.00 to .93...

Don't hate on the results just because they don't support your narrative...

Last words yours... I'm out

 

 

 

[NCHoosier32]

[KoB2011]

54 minutes ago, go_iu_bb said:

For getting into the tournament the committee cares most about winning games, especially for high major teams. If IU is on the bubble or worse, it will be because they didn't win enough games. They won the games they were supposed to win in November so if they're on the bubble that means they didn't win enough in December-March. Those would be the games I would look at for reasons why IU didn't make the tournament over games they won but just weren't efficient enough to help their NET ranking.

Okay, to each their own. I'll look at the total picture, and while yes not winning enough games to safely be in would be a reason, the metrics would also be a reason we end up on the wrong side of the bubble. It's both/and, not either/or. 

[ColtsIUDomer]

Is this March 5th?  lol...Man I wish!..I'd be heading to Vegas in a couple weeks if it was. 

[KoB2011]

52 minutes ago, IUFLA said:

Ok, so I added 5 more offensive rebounds for us... It's still not a big difference... 1.00 to .93...

Don't hate on the results just because they don't support your narrative...

Last words yours... I'm out

 

 

 

And that still isn't an accurate way to look at it 🤷‍♂️

If we won by 10 more points, we were some blend of ~10-15% better on both ends if you adjusted everything.

The simple math would be 10 more points / 60 possessions is 16.666% better points per posession. Slightly more than 60 possessions so I rounded down to 10-15%, but you can't just adjust one or two factors in an equation, that isn't how basketball games unfold. 

 

[Kdug]

3 hours ago, IUFLA said:

So I did the numbers as described for efficiency in the chart for the FGCU game...

FGA (45)- offensive rebounds (6)+ turnovers (11)+.425 x FTA (30 attempts 14.75= 64.25/points (69)= 0.9311

I added 10 more FGAs (55) and 10 more points (79) to the above and came up with 0.9462

I calculated FGCU'S numbers and came up with 1.1087 

Subtracting FGCU'S numbers from our initial (real) total total I got -.1776

From the 2nd calculation -.1627

Difference of .0149

That would be the difference between winning by 6 or winning by 16... 

Someone else wants to calculate it, I'd be appreciative... But that's what I got... 

 

Raw efficiency is simply points divided by possessions. There were 64 possessions against FGCU and IU won 69-63. So IU’s efficiency was 1.08 and FGCU’s was 0.98.

If IU won 79-63 instead with the same number of possessions, IU would’ve had a 1.23 efficiency vs the same 0.98.

So the actual result of the game was a net efficiency of 0.1 points per possession. Adding 10 points for IU resulted in a net efficiency of 0.25 points per possession.

In your example you added 10 FGA and 10 points. So that added 10 possessions at an efficiency of 1 point per possession to IU. We’d also have to add 10 possessions to FGCU. If we want to see what a 16 point win looks like, we have to give FGCU 10 more possessions at 0 points per possession. In this example IU has an efficiency of 1.08 (79/74) while FGCU is at 0.85 (63/74). So adding 10 to the scoring margin with 10 extra possessions gives a 0.23 net efficiency per possession.

These are raw efficiency numbers. Basically every advanced metric then weighs that efficiency based on the quality of the opponent.

[KoB2011]

18 minutes ago, Kdug said:

Raw efficiency is simply points divided by possessions. There were 64 possessions against FGCU and IU won 69-63. So IU’s efficiency was 1.08 and FGCU’s was 0.98.

If IU won 79-63 instead with the same number of possessions, IU would’ve had a 1.23 efficiency vs the same 0.98.

So the actual result of the game was a net efficiency of 0.1 points per possession. Adding 10 points for IU resulted in a net efficiency of 0.25 points per possession.

In your example you added 10 FGA and 10 points. So that added 10 possessions at an efficiency of 1 point per possession to IU. We’d also have to add 10 possessions to FGCU. If we want to see what a 16 point win looks like, we have to give FGCU 10 more possessions at 0 points per possession. In this example IU has an efficiency of 1.08 (79/74) while FGCU is at 0.85 (63/74). So adding 10 to the scoring margin with 10 extra possessions gives a 0.23 net efficiency per possession.

These are raw efficiency numbers. Basically every advanced metric then weighs that efficiency based on the quality of the opponent.

Thank you for doing the math where I was too lazy.

A 10 point difference would have been a sizable difference in our efficiency in this game. 

[Kdug]

3 minutes ago, KoB2011 said:

Thank you for doing the math where I was too lazy.

A 10 point difference would have been a sizable difference in our efficiency in this game. 

Yup, big difference. We’d need to weight the efficiency vs an average D1 opponent, and FGCU is below average. But if we assumed those were the opponent adjusted values, a 0.1 vs 0.23 net efficiency per possession is the difference between being ranked around 75th in kenpom vs ranked around 10th.

The army game is what’s really killing our adjusted efficiency.

[Kdug]

The talk about efficiency numbers had me curious, so I looked to see what our game by game efficiency was using T-Rank. Torvik has a few unique aspects to his algorithm, but conceptually it still should be similar to kenpom and NET. But mainly I'm using his numbers because they're free and easily available. I'm including what the overall rank would be for each game if that were our season long adjusted net efficiency. The efficiency numbers are all per 100 possessions:

• H vs FGCU: +2.3 (140th)
• H vs Army: -14.3 (332nd)
• H vs Wright State: +17.1 (29th)
• N vs UCONN: -0.5 (177th)
• N vs Louisville: +6.7 (105th)
• H vs Harvard: +9.8 (73rd) - I know this was in Indy, but Torvik has it listed as home
• H vs Maryland: +33.1 (1st)

[KoB2011]

1 hour ago, Kdug said:

The talk about efficiency numbers had me curious, so I looked to see what our game by game efficiency was using T-Rank. Torvik has a few unique aspects to his algorithm, but conceptually it still should be similar to kenpom and NET. But mainly I'm using his numbers because they're free and easily available. I'm including what the overall rank would be for each game if that were our season long adjusted net efficiency. The efficiency numbers are all per 100 possessions:

• H vs FGCU: +2.3 (140th)
• H vs Army: -14.3 (332nd)
• H vs Wright State: +17.1 (29th)
• N vs UCONN: -0.5 (177th)
• N vs Louisville: +6.7 (105th)
• H vs Harvard: +9.8 (73rd) - I know this was in Indy, but Torvik has it listed as home
• H vs Maryland: +33.1 (1st)

That's for this leg work. I think there is a lot that is telling here, but to focus on the positive my eyes have told me something started to click in the second half against Louisville and has continued since there. These number back that up as they've been getting better game over game.

Having said that, still need to continue to improve and build on this. The three games we have coming up are tough enough that we can play some of the best basketball we've played all year and lose all three. That's certainly not ideal, but if we are continuing to improve and that just is the end result then so be it. 

[5fouls]

A lot of positive talk about Gunn's defense during the Michigan game  The numbers bear that out.  CJ now has the best Defensive Rating on the team.  

[kyhoosier29]

MM at 6 and Gabe at 7 along TG last make these hard to take serious (no offense). We know MM struggles on D and Gabe / TG are much better (by no means perfect… had their struggles at times).