Another Win. Another Drop in the NET? | Page 3 | Syracusefan.com

Another Win. Another Drop in the NET?

Ok, if you guys think the ACC is better than the Big 10 & 12, I don't even know what to tell you. The ACC sucks
Do you even watch college basketball? The Big 12 and Big 10 aren’t that good. Both vastly overrated by the NET. There’s not much difference between the conferences.
 
Blowout losses.
And we have greater luck - called out in Kenpom but comes through in NET.
But we’ve played very few close games so how can our “luck” metric be high? It makes no sense.
 
Regarding margin of victory and the NET rankings. Should you ever put walk ons in the game when you are up 20 with 1 minute left? Should you ever take your starters out? Should you run up the score if up by 15 with a minute left? Cuse had the ball and ran out the last 20 seconds last night and for the first time, I was thinking, shoot a 3. Increase that margin of victory. You shouldn't have to coach that way to build a tourney resume.
 
But we’ve played very few close games so how can our “luck” metric be high? It makes no sense.
It means we won games we weren't expected to win based on our bad rating (i.e. beat teams ranked better in KP). Another explanation for what KP calls "luck" or lucky wins, is that we're underrated based on efficiency/margin sinking the ship and it's not actually luck, it's that we're better than the rankings suggest. "Luck, which is the deviation in winning percentage between a team’s actual record and their expected record using the correlated gaussian method. The luck factor has nothing to do with the rating calculation, but a team that is very lucky (positive numbers) will tend to be rated lower by my system than their record would suggest."
 
But we’ve played very few close games so how can our “luck” metric be high? It makes no sense.
We've won a lot of close games, only had a very few blowout wins. But every time we lose we get blown out. The bad overall margin of victory shows up in our key metrics that are used to predict win-loss records.

Our actual record is significantly better than our predicted record says we should be.

They call this effect "luck"
 
It means we won games we weren't expected to win based on our bad rating (i.e. beat teams ranked better in KP). Another explanation for what KP calls "luck" or lucky wins, is that we're underrated based on efficiency/margin sinking the ship and it's not actually luck, it's that we're better than the rankings suggest. "Luck, which is the deviation in winning percentage between a team’s actual record and their expected record using the correlated gaussian method. The luck factor has nothing to do with the rating calculation, but a team that is very lucky (positive numbers) will tend to be rated lower by my system than their record would suggest."
You're exactly right. Kenpom wants to protect his algorithm so he uses luck to manage why Cuse is going to win 20 games instead of the 11 that he predicted. Things like first year head coach, prior year finish all go into this.
 
You're exactly right. Kenpom wants to protect his algorithm so he uses luck to manage why Cuse is going to win 20 games instead of the 11 that he predicted. Things like first year head coach, prior year finish all go into this.
I think "Luck" is just a bad name for it. Maybe deviation would be better. To your point, maybe it's because "luck" makes it seem like his rating is correct but sports are unpredictable, whereas deviation would indicate his rankings are wrong.
 
I think "Luck" is just a bad name for it. Maybe deviation would be better. To your point, maybe it's because "luck" makes it seem like his rating is correct but sports are unpredictable, whereas deviation would indicate his rankings are wrong.
You make your own luck.
 
You're exactly right. Kenpom wants to protect his algorithm so he uses luck to manage why Cuse is going to win 20 games instead of the 11 that he predicted. Things like first year head coach, prior year finish all go into this.

This is 100% untrue. It's purely a formulaic calculation taking our schedule strength against our efficiency (in actual results based on margin). I wouldn't call the stat "luck", but it has nothing to do with any of your claims above.

This is not to say there are not flaws in KP, there are.
 
Do you even watch college basketball? The Big 12 and Big 10 aren’t that good. Both vastly overrated by the NET. There’s not much difference between the conferences.
Lol
 
But we’ve played very few close games so how can our “luck” metric be high? It makes no sense.

There are two factors that could determine "positive" luck per KP.
1) Winning a bunch of close games, and not losing many. (as you noted). This is the more common one.
or
2) Getting blown out when you lose, and not beating bottom half teams by as much as expected (and the trend not being offset much the other way) This lowers your relative efficiency, which indicates you shouldn't win as much games overall.

We fall into 2) obviously.

As I said I wouldn't call the stat "luck".
 
There are two factors that could determine "positive" luck per KP.
1) Winning a bunch of close games, and not losing many. (as you noted). This is the more common one.
or
2) Getting blown out when you lose, and not beating bottom half teams by as much as expected (and the trend not being offset much the other way) This lowers your relative efficiency, which indicates you shouldn't win as much games overall.

We fall into 2) obviously.

As I said I wouldn't call the stat "luck".
#2 sounds like the NET lol
 
I'm much pissier (not a word?) about the Mountain West's NETs than Big 12 personally. At least I know Houston, Kansas, BYU, Baylor, etc are legit teams.

The MWest's best OOC wins are 1 win over Creighton and 2 wins over St Mary's...then they just beat each other...uh ok?
 
Regarding margin of victory and the NET rankings. Should you ever put walk ons in the game when you are up 20 with 1 minute left? Should you ever take your starters out? Should you run up the score if up by 15 with a minute left? Cuse had the ball and ran out the last 20 seconds last night and for the first time, I was thinking, shoot a 3. Increase that margin of victory. You shouldn't have to coach that way to build a tourney resume.

Last minute play isn't really going to make much of a difference. Remember if teams start doing it, they all do it, so it evens out. *** If everyhing breaks right you can maybe add 10-20 overall points over the course of 31 games.

We are about 170-180 points below expectation (for a KP / NET 40 type teams) in our Q3 and Q4 wins, and our losses. That is based on what happened over the full game for the most part. No matter how we managed our last 5 minutes of games that wasn't going to be overcome. We stayed close against bad teams.

*** That being said I tell all my member teams to keep the pedal to the metal as long as possible in all their OOC games, which I'm pretty sure the Big 12 did this year. If you were considering bringing in the walk-ons at the 4 minute mark... do it at the 1 or 2 minute mark.

In conference this strategy doesn't really work, because its and in and out as a group.
 
If we win the next 2 road games, we might crack the Top 70 in NET (probably just lower 70's though), even if we barely win both games. (it will be quite close I think) Optically that would be significant.

The NET, compared to KP, does appear to give more of a power boost to getting a road win. This is based on my tracking of big movers from week to week, where I see teams with a big road win moving up more in NET than KP. Otherwise they run fairly parallel.

For example:
- We are going to be 9.5 to 10.0 point underdogs vs Clemson (opening line), so simply by winning that game we are going to be 11 points above "margin expectation" which is a healthy bump under both ( KP would stop there in terms of positive impact)
- In NET there is clearly some extra juice in their secret sauce for road wins. So we would get extra bonus for that road win beyond the "margin" gain, but its hard to quantify.

Similarly we would get a bit of juice from beating Louisville on the road.
 
I'm much pissier (not a word?) about the Mountain West's NETs than Big 12 personally. At least I know Houston, Kansas, BYU, Baylor, etc are legit teams.

The MWest's best OOC wins are 1 win over Creighton and 2 wins over St Mary's...then they just beat each other...uh ok?

The MWC had a better Q1+Q2 OOC Win% as a conference, then the Big 12, SEC and the ACC. True they don't play much against the P6. They do play quite a number of games against fairly good schools from other conferences, including more road games - which the P6 avoids.

It's always been my thing to dry to dive into things (as the MWC was a talking point at the time(, which i did in early February,

 
This is 100% untrue. It's purely a formulaic calculation taking our schedule strength against our efficiency (in actual results based on margin). I wouldn't call the stat "luck", but it has nothing to do with any of your claims above.

This is not to say there are not flaws in KP, there are.
What does luck in Kenpom mean then ? We are 6th in the league out of 360 teams. +.136.
 
I think "Luck" is just a bad name for it. Maybe deviation would be better. To your point, maybe it's because "luck" makes it seem like his rating is correct but sports are unpredictable, whereas deviation would indicate his rankings are wrong.
Agreed. Calling it Luck removes the blame from the system. When really its faultiness with the algorithm. Calling it what it is, deviation, would undermine the NET system. Now they wouldn't want to do that...
 

Forum statistics

Threads
170,596
Messages
4,900,619
Members
6,004
Latest member
fsaracene

Online statistics

Members online
273
Guests online
1,463
Total visitors
1,736


...
Top Bottom