Yes I didn't realize the 1.1 or .9 could get adjusted so that definitely makes more difficult for us novices to track. Thanks guys

Quote from: USee on November 01, 2024, 05:52:12 PMThat's the exact number I got. When I modeled out potential wins for the next 3 games I got a final NPI for Wheaton of 64.19. Of course all the current teams NPI's will change each week so that's inexact at best. I don't understand how a home loss gets a 1.1 multiplier because that helps the losing team's NPI but I think that's how it's supposed to work.

I agree but that's how they had it in the D3 datacast dataset although right now can't get to dataset it seems to just link to the NCAA stats so not sure if guys are working on it. I did the Carroll calculations too using same parameters and got correct NPI.

Quote from: USee on November 01, 2024, 05:52:12 PMI don't understand how a home loss gets a 1.1 multiplier because that helps the losing team's NPI but I think that's how it's supposed to work.

Think of the multiplier as providing more or less gravity toward the the Game NPI for each matchup, not so much as scaling the NPI up or down. For the 1.1 multiplier, that Game NPI gets the x1.1 treatment, but the 1.1 games are also added to the net adjusted games played (so in a micro sense, that game gets multiplied by 1.1 then divided by 1.1).

Hypothetical example:

Home win with Game NPI (opponent win value) of 70: [70 x 0.9 = 63]
Home loss with a Game NPI (opponent loss value) of 50: [50 x 1.1 = 55]
(2.0 adjusted net games played)

(63 + 55) / 2.0 = 59.0 NPI

If the multiplier was 0.9 for all home games:

Home win with Game NPI (opponent win value) of 70: [70 x 0.9 = 63]
Home loss with a Game NPI (opponent loss value) of 50: [50 x 0.9 = 45]
(1.8 adjusted net games played)

(63 + 45) / 1.8 = 60.0 NPI

So really we should be looking at those Game NPI's (of 70 and 50) and see the 1.1 or 0.9 multiplier as pulling us more or less toward those games' component NPI values. So, in the example above, the 1.1 multiplier on the loss decreases the final Season NPI more than a 0.9 multiplier would. Hope that helps!

Quote from: ziggy on November 01, 2024, 12:43:09 PM

Quote from: Gregory Sager on November 01, 2024, 12:05:44 PM

Quote from: ziggy on November 01, 2024, 08:49:49 AM

Quote from: CarollFan on November 01, 2024, 08:14:23 AMNPU's current NPI is 57.157.
So using that NPI a win is 40+(.6x57.157)+((57.157-54)x.25)= 75.08 might be a little more because on the road. This is win%+sos+gwb.

That will get included in all the other game NPI's used for the overall calculation.

If NPU's NPI drops to 50 let's say, now it's 40+(.6x50)=70. Lose the gwb because NPU's NPI now less then 54.

You have no idea how much it warms my heart to see this kind of post popping up on the boards!

Just to be clear, The D3 Datacast perspective has always only been to educate about, not advocate for, NPI. Seeing the "back of the napkin" type NPI calculations showing up here is good feedback that our education efforts have been somewhat effective with people willing to learn.

I was only taking the opportunity to gratuitously rattle the cage of D3's most prominent numbers-nerds siblings. You guys know how much I actually appreciate your content. But I can't just keep posting the same tired meme:



... every time you guys get into the analytical weeds. I've got a reputation to keep up as well, so I have to maintain my d3boards gif-posting standards.

Quote from: Gregory Sager on November 02, 2024, 10:58:51 AM

Quote from: ziggy on November 01, 2024, 12:43:09 PM

Quote from: Gregory Sager on November 01, 2024, 12:05:44 PM

Quote from: ziggy on November 01, 2024, 08:49:49 AM

Quote from: CarollFan on November 01, 2024, 08:14:23 AMNPU's current NPI is 57.157.
So using that NPI a win is 40+(.6x57.157)+((57.157-54)x.25)= 75.08 might be a little more because on the road. This is win%+sos+gwb.

That will get included in all the other game NPI's used for the overall calculation.

If NPU's NPI drops to 50 let's say, now it's 40+(.6x50)=70. Lose the gwb because NPU's NPI now less then 54.

You have no idea how much it warms my heart to see this kind of post popping up on the boards!

Just to be clear, The D3 Datacast perspective has always only been to educate about, not advocate for, NPI. Seeing the "back of the napkin" type NPI calculations showing up here is good feedback that our education efforts have been somewhat effective with people willing to learn.

I was only taking the opportunity to gratuitously rattle the cage of D3's most prominent numbers-nerds siblings. You guys know how much I actually appreciate your content. But I can't just keep posting the same tired meme:



... every time you guys get into the analytical weeds. I've got a reputation to keep up as well, so I have to maintain my d3boards gif-posting standards.

With NPU having their best year in a long time that may help put Wheaton in the playoffs. Usee may owe you a big thank you.. 
Oops...

Quote from: KnightSlappy on November 02, 2024, 09:01:37 AM

Quote from: USee on November 01, 2024, 05:52:12 PMI don't understand how a home loss gets a 1.1 multiplier because that helps the losing team's NPI but I think that's how it's supposed to work.

Think of the multiplier as providing more or less gravity toward the the Game NPI for each matchup, not so much as scaling the NPI up or down. For the 1.1 multiplier, that Game NPI gets the x1.1 treatment, but the 1.1 games are also added to the net adjusted games played (so in a micro sense, that game gets multiplied by 1.1 then divided by 1.1).

Hypothetical example:

Home win with Game NPI (opponent win value) of 70: [70 x 0.9 = 63]
Home loss with a Game NPI (opponent loss value) of 50: [50 x 1.1 = 55]
(2.0 adjusted net games played)

(63 + 55) / 2.0 = 59.0 NPI

If the multiplier was 0.9 for all home games:

Home win with Game NPI (opponent win value) of 70: [70 x 0.9 = 63]
Home loss with a Game NPI (opponent loss value) of 50: [50 x 0.9 = 45]
(1.8 adjusted net games played)

(63 + 45) / 1.8 = 60.0 NPI

So really we should be looking at those Game NPI's (of 70 and 50) and see the 1.1 or 0.9 multiplier as pulling us more or less toward those games' component NPI values. So, in the example above, the 1.1 multiplier on the loss decreases the final Season NPI more than a 0.9 multiplier would. Hope that helps!

Thanks KnightSlappy. This is very helpful.

CarrollFan- Greg knows how thankful I am for him! And if there is one team in the league I'd like to have succeed more than my alma it's NPU. I loved their sweet 16 basketball season and watched almost every game, I love their men's soccer success (even this year when my team beat theirs!) and I have even watched a couple NPU women's basketball games.

Overall, I think it would be great to have a league like the WIAC where every week, every game was a battle and the outcome in question. Every team in our league has been to the NCAA playoffs since 2000 except NPU and Carroll. I'd welcome that to change.

Aw, man, you're making me blush!

Whenever people unfamiliar with North Park ask me which school is our biggest rival, I have to add two points of clarification:

1) Even though we've been in the same league since the 1960s, our biggest rival barely even knows that we exist (aside from men's soccer, regarding which they are just as invested in the rivalry as we are), and that one-sided animosity is a very frustrating thing for North Parkers; and

2) The students and alumni of our archrival make it even harder on us by generally being nice people who are impossible to dislike on a personal level. Why can't Wheaton produce a Norm Eash or two? 

I should add that after that last post I fully expect the North Park University Board of Trustees to show up at my door and demand that I hand over my diploma so that it can be burned in front of the steps of Old Main.

Carthage 22
North Park 21

Sigh.

Carthage deserved to win this game. The Firebirds executed only sporadically, but they dominated both sides of the line of scrimmage. When you do that, you don't have to execute perfectly every time, and they were able to do enough to eke out the kind of fourth-quarter comeback win that, ironically, NPU has excelled at all season long.

And so the best chance in 30 years to break The Streak goes by the boards, as Lucy pulls the football away from Charlie Brown yet again.

Wait 'til 2025, I guess.

Finals

NCC 56 Carroll 3
WashU 70 Millikin 21
Wheaton 65  Elmhurst 17
Carthage 22 NPU 21
IWU 29 Augie 21

Remaining conference games:
NCC (8-0) : Elmhurst, @IWU
Wheaton (6-2) : @Millikin, NPU
WashU (6-2) : @NPU, Augie
NPU (5-3) : WashU, @Wheaton
Carroll (4-4) : IWU, Carthage
IWU (4-4) : @ Carroll, NCC
Augie (3-5) : Carthage, @WashU
Elmhurst (2-6) : @NCC, Millikin
Carthage (2-6) : @Augie, @Carroll
Millikin (0-8) : Wheaton, @Elmhurst

Quote from: CarollFan on November 02, 2024, 03:55:15 PMFinals

IWU 29 Augie 21

Did Norm shake hands with the Augie coaches after the game?

Quote from: Gregory Sager on November 02, 2024, 03:53:35 PMCarthage 22
North Park 21

Sigh.

Carthage deserved to win this game. The Firebirds executed only sporadically, but they dominated both sides of the line of scrimmage. When you do that, you don't have to execute perfectly every time, and they were able to do enough to eke out the kind of fourth-quarter comeback win that, ironically, NPU has excelled at all season long.

And so the best chance in 30 years to break The Streak goes by the boards, as Lucy pulls the football away from Charlie Brown yet again.

Wait 'til 2025, I guess.

Quite a surprising outcome. 😧

Updated NPI #'s 11/2

https://stats.ncaa.org/selection_rankings/nitty_gritties/39931

Rank    Teams      NPI            Rank Movement from last week
#3      NCC             77.375        Same                     
#38    Wheaton        63.476      Better 41 to 38                         
#48    WashU          60.409      Better 58 to 48                             
#100    NPU           53.565      Worse 69 to 100                 
#106    Carroll       52.211      Worse 100 to 106                     
#119    IWU              49.475    Better 147 to 119                                                 
#149    Augie          45.643      Worse 132 to 149                 
#175    Carthage      42.424    Better 204 to 175                     
#181    Elmhurst       41.799    Worse 175 to 181                     
#226    Millikin      25.454      Better 232 to 226

Biggest drivers for Wheaton's increase are these NPI's increasing Carthage, IWU and WashU.  NPU's NPI for now still better then Carthage. It's possible they may both get used in Wheaton's calculations. They are also this week counting all 6 wins in the Wheaton NPI so they included the Elmhurst game.

Watched entire IWU-Augie game. Great one to watch, literally came down to last drive.

Beautiful day to play.

I think IWU ability to run the ball and stop the run was the difference in the game. Titans looked more physically developed in both lines.

What happened to original Augie QB that started last year?  Bhardwaj?