Good wins versus bad losses
- Thread starter NUThump
- Start date
As long as Purdue holds onto a top ten ranking, I'll take that victory.
Always better to have great wins with bad losses vs a mediocre resume with nothing standing out.
This. "A mediocre resume with nothing standing out" is exactly why a 20-win NU team didn't even make the NIT in 2016.
Depends what you mean by "better."
From the computer ratings perspective, its all about the difference between expected outcome and actual outcome. So they penalize us more for losing to Chicago State than credit us for beating Purdue. But thats because we were supposed to crush Chicago State.
From the "Net Ratings" methodology, I'm not sure.
From a human perspective, most would be inclined to credit NU for beating a really good team and forgive NU for losing to a really bad team, treating the latter as a fluke.
Screw the "ratings". A conference win is better than a non-conference win. I want a Big Ten title more than a high seed in the tournament.
I remember reading that it's not our NET ranking that matters, but the NET rankings of our opponents. They compile a list of good wins and bad losses and compare teams on that basis.
The great win is better. The bad loss hurts and can be a tiebreaker against another bubble team who doesn't have such a loss, especially a home loss. A decent number of teams at that level will also have a gross loss in there somewhere. An elite win against a possible top seed is something an entire resume gets built around and few other teams in the bubble or in the mid seeds will have anything comparable.
Or, you know, you could just play focused against a horrible opponent at home and take care of business and not worry about it, but I digress.
Or, you know, you could just play focused against a horrible opponent at home and take care of business and not worry about it, but I digress.
A win over #1 gets the court stormed, a top 25 ranking and national coverage. A win over Chicago State gets you nothing
Seems more important to me to show that you can beat the number one team. When you do that it is a statement that you belong in the conversation. Beating lessers teams is important if you are building toward the future. Beating the best teams is the future you are trying to realize.
The NET ratings are not explained very well. I've been searching around for a decent explanation - but encountering secret formulas.
They definitely use Ken Pom's efficiency metrics.
But, "efficiency" is just a cool-sounding word for (points scored / possessions).
So you score 80 points in 60 possessions and you are very efficient on offense and your opponent is very inefficient on defense in that game. Very simple. Not nearly as complicated as most people think. Combine the results of each of the games.
What is unclear to me is how a team gets into a "Quadrant" for the purposes of the NET's Quad 1-4 wins and losses. My suspicion is that Ken Pom's ratings determine the initial ranking of the teams.
What seems to be "secret" is how the NCAA combines the Quad 1-4 wins and losses to get a NET rating for each team. And thats what matters.
Quad 1-4 is a largely arbitrary distinction determined after the ratings are finalized. Quadrant wins and losses are used by the committee to distinguish between teams for selection purposes so they're not just taking the top 68 teams (minus auto bids) off the list.
So it helps our rating, for example, to have a win over Dayton, but as far as I know we don't get any sort of extra bonus just because Dayton is rated 25 (quad 1) and not 31 (quad 2).
When you write "Quad 1-4 is a largely arbitrary distinction determined after the ratings are finalized" that reflects the lack of transparency the NCAA has opted for.
I think what they might be doing is taking Ken Pom's ratings "as is" and applying the quadrants to those.
So if KenPom says NU is #55, then Illinois is playing a Quad 2 game tonight. NU is playing a Quad 1 game.
They probably do the whole schedule and apply some number of points for each Quad 1 win/loss, Quad 2 win/loss, Quad 3 win/loss and Quad 4 win/loss.
Maybe they give +10 for a Quad 1 win, +5 for a Quad 2 win, +2 for a Quad 3 win and +1 for a Quad 4 win, -10 for a Quad 4 loss, -5 for a Quad 3 loss ,-2 for a Quad 2 loss and -1 for a Quad 1 loss.
Regardless, they should say what they're doing, but my guess is they don't because they want to be able to override the ratings in favor of big schools with big alumni bases, without public scrutiny.
I have no idea how they calculate the actual NET rating because, as you say, they don't publish their "Team Value Index" formula. Their adjusted efficiency formula does indeed seem similar to KenPom. So you could be right about how it's determined.
My point is that when we talk about a "Quad 1 win," we're talking about a team's current ranking. At season's end, the committee gets the final NET rankings and are tasked with sorting those teams into the field of 68. One of the ways they do this is by evaluating how many Quad 1 wins a given team has and/or how many Quad 3 or Quad 4 losses. Those quadrant records aren't determinative of the NET ratings because the NET ratings are already done. It's just an arbitrary distinction determined by putting brackets around different parts of the rankings.
Similar threads
