Columbia Women Rip Dartmouth 81-42

[lion600]

This was a one-sided affair with the Lions dominating Dartmouth in every category. Columbia has improved significantly in the last two months which is a tribute to Coach,Griffith, her coaching staff, and, of course, the players themselves. An offense which was ineffective at the start of the season has become a powerhouse with numerous players now contributing. Our two brightest stars, Riley and Perri, are shining brighter than ever, but so are virtually everyone else. Broom had a terrific game against Dartmouth, but so did Simmons who keeps expanding her game. Fliss, Maria and Blau also excelled. Despite all the superlatives, the competition is also improving so our players need to continue working hard if they want to win the Ivy League Title and more

[PeteD]

It’s great to see the confidence of this team increase as the season progresses. Broom and Simmons are becoming key contributors to a championship-level team, which bodes well for the next couple of seasons.

They clinched an Ivy Madness berth with today’s win, their 6th straight.

[Kreeeker]

Really impressed with how our guard play has developed. This version of Broom is already looking like the PG that will lead this team for the next two years.

I’m a bit more worried about the inside presence. Page is fundamental for how the team thrives in big games. Early fouls on Page and Rafiu could be a big issue in ILT. Noone from the bench has shown a good enough level to give me confidence. Too many mistakes

[cc66]

Broom seems to have suddenly hit her stride. Her offense has picked up to 9 ppg over the last 8 games, and I can easily see her averaging in the low teens by next season. It looks as if she is just about to displace Maria Avijlas in the starting lineup.

Unfortunately, Brown, the stronger team next weekend, is the back end of a back to back. I would expect Megan to cap the first team’ s minutes at 25 and assume that with a decent lead, the reserves can carry us to victory

One persistent curiosity: why has Shay Shippen, the all-time hs scorer in Idaho, gotten into no recent games, even ones where we were leading by 30+ pts.?

[Kreeeker]

I have no idea. I guess she lacks a lot of defense to be able to play.

On another note, it seems we are going to need the same or more from Broom. Sad news: Fliss was wearing a boot today, so I would assume she is out at least for next weeked, and who knows for how long.

[cc66]

Did you see her in a boot on campus?

I was wondering why she only played 19 minutes yesterday.

We will miss her rebounding and inside game.

[CC89]

I think Shay’s been in street clothes recently and maybe even a boot. I think she’s injured. The coaching staff has done a better job recently of getting more players into the games. I’m wondering why one of our spanish players Maria Arribola (spell?) is not getting any playing time of late. She’s done well when she’s been in there.

I wouldn’t assume Shay is a bad defender. Some have said that about Riley in the past and she is actually a good defender. Looks can be deceiving. Basketball IQ plays a big role in defense.

If Fliss is out for the ILT, we’re in trouble. She’s the heart and sole of this team along with Perri Page.

[DoubleDegree]

Does anyone know why coach Griffith insists on picking up full court on D while up 30+? I’ve seen it before and it always seems odd. Curious if she’s ever addressed it.

[CC89]

One of the things I like about watching Columbia woman’s basketball (and I think this is true for the entire IL) that there is no garbage time. Each possession is a learning experience for whoever’s is on the court and they play hard to the last possession. I find this refreshing. Even if there is a 30 point gap, this scrimmage is useful in player and team development. Don’t expect a kneel down.

[DoubleDegree]

Sure, I get all that. I guess having played the game myself I came from a time when if you were beating the sh*t out of a sorry team you’d at least pull out of full court pressure. Times have changed, apparently.

[LocalTiger]

Griffith is a great coach, and has built an enviable program.

Now, she needs to learn how to win (and lose) with class.

[Michael]

I think the fact that the NET rankings have margin-of-victory as one of the criteria might factor in, especially given how much the NCAA tournament selection committee apparently considers the NET with bubble teams and/or seeding. (Granted, an aggressive press in the last two minutes, say, of a blowout probably won’t radically change victory margin).

[DoubleDegree]

Ah, this makes sense.

[JDP]

The women’s NET really isn’t about raw margin of victory the way people think. It’s driven much more by efficiency margin — offensive efficiency minus defensive efficiency — calculated possession by possession.

That matters because defensive efficiency doesn’t stop counting when the game is effectively over. Every late possession still moves the math. If you ease up, empty the bench, or stop pressing, efficiency drops fast — and so does NET. From a pure numbers standpoint, keeping starters or regular rotation players in longer actually optimizes the formula.

Do I love that incentive structure? Not particularly. I believe the Men’s version has a cap of 10 by game to address the structure. The women’s model quietly rewards teams for playing clean, high-efficiency basketball for the full 40 minutes instead of managing a blowout traditionally. But NET is heavily used by the NCAA selection committee and it’s the third Ivy League tiebreaker, so coaches would be negligent to ignore it.

And while NET probably won’t ultimately decide a Brown–Penn tiebreaker, the reality is the two programs sit close enough in NET that both staffs have to be conscious of protecting and maximizing their number every night. Small efficiency swings add up over a season.

So when you see pressure defense or limited subs late, it’s usually not about running up the score. It’s coaches managing the evaluation system they’re judged by — even if the optics sometimes feel uncomfortable.

The old-school instinct says empty the bench. The modern reality says the spreadsheet is still watching.

[PeteD]

Understood, but there must be a better way. I’m sure someone could come up with a method or formula to “turn off the spreadsheet” when the score reaches a certain point differential (y) with x minutes remaining.

[JDP]

I suspect a better methodology could exist and that @recspecs730 has the talent to develop NET 2.0, but the underlying challenge is structural. The statistical tools needed to meaningfully differentiate the top 25–32 teams are not necessarily the same tools that separate teams ranked 50–150 or 150–363. One formula is being asked to solve very different problems simultaneously.

Ultimately, the NET’s primary purpose is practical: selecting and seeding at-large NCAA tournament teams. The model therefore has to be optimized for last 12 in last 12 out bubble evaluation, not for perfect fairness across every game state. As with any optimization problem, that inevitably creates unintended incentives at the margins.

In theory, modern technology makes more dynamic solutions possible. With real-time play-by-play data already feeding advanced scoring systems, a form of live NET logic could exist. For example, once a game reaches the fourth quarter with a scoring differential above a defined threshold (say 20+ points) and win probability exceeds ~95%, both benches could receive an in-game signal that additional possessions will no longer materially impact efficiency inputs.

Conceptually, that would “turn off the spreadsheet” without requiring coaches to manage analytics incentives themselves. The open question, however, is whether the additional complexity actually improves outcomes.

For those that care about the underpinnings of the NET … here is what Chat GPT told me …

The Men’s and Women’s NET systems are conceptually similar, but they are not identical models — the women’s version relies more heavily on efficiency-based performance metrics and applies different weighting and normalization choices reflecting structural differences in the two games.

Here’s the clean analytical breakdown.

---

1. Core Structure — Same Framework

Both the Men’s NET and Women’s NET are built from two primary pillars:

Component	Men’s NET	Women’s NET
Team Value Index (TVI)
Adjusted Net Efficiency
Opponent Adjustments
Game Location Weighting
Scoring Margin Cap

So philosophically, both systems try to answer:

How good are you + who did you beat + where did you play?

---

2. The Biggest Difference: Efficiency Construction

Men’s NET

Efficiency is important — but results drive more separation.

Men’s NET blends:

• Game results (wins/losses vs opponent quality)

• Efficiency margin (offense − defense)

• Schedule strength interactions

Historically, the men’s model evolved partly from RPI replacement logic, meaning:
résumé still matters heavily.

---

Women’s NET

The women’s system is much closer to a possession-efficiency model.

Efficiency differential plays a larger mathematical role:

Net Efficiency=Adjusted Offensive Efficiency−Adjusted Defensive EfficiencyNet Efficiency=Adjusted Offensive Efficiency−Adjusted Defensive Efficiency

Where efficiency is calculated per possession, not per game.

Key implication:

Every possession matters
Late-game execution still moves ratings
Lineup degradation shows up quickly

This is why women’s NET often tracks closely with analytics sites like HerHoopStats.

So what the difference?

Excellent question — because the difference is intentional, not accidental.
The NCAA built two related but distinct rating systems because the statistical environments of men’s and women’s Division I basketball are materially different.

Let’s walk through the theoretical logic first, then the margin-of-victory issue.

---

1. The Core Theoretical Problem the NCAA Had to Solve

A ranking system must answer:

How much information does a single game actually contain about team quality?

The answer differs between men’s and women’s basketball.

---

2. Distribution of Team Strength (The Key Justification)

Men’s Division I Basketball

Competitive distribution is dense.

• Talent spread is relatively compressed

• Upsets occur frequently

• Score margins cluster tightly

• Hundreds of teams are plausibly competitive

Statistically:
Game outcome itself carries strong signal

Winning vs losing already tells you a lot.

So the men’s model can lean more heavily on:

• wins/losses

• opponent quality

• résumé outcomes

Efficiency helps — but results remain informative.

---

Women’s Division I Basketball

Competitive distribution is top-heavy.

• Elite teams dramatically stronger than median teams

• Blowouts occur more often

• Conference imbalance is larger

• Fewer true upset environments

Statistically:
Binary outcomes lose informational value

Example:

Game	Information Content
#3 beats #180 by 35	Expected
#3 beats #180 by 12	Very informative
#3 beats #180 by 45	Also informative

A simple win/loss cannot distinguish performance quality.

So the women’s NET must rely more on:

possession efficiency
scoring control
defensive consistency

This extracts signal from games where outcome is predetermined.

---

3. Why Efficiency Gets More Weight in Women’s NET

From a modeling standpoint:

Observed Margin=True Team Strength+NoiseObserved Margin=True Team Strength+Noise

In WBB:

• Outcome variance is lower

• Margin variance contains usable signal

Therefore efficiency differential becomes a better estimator of underlying team strength.

In MBB:

• Upset probability is higher

• Late randomness larger

• Margin contains more noise

So over-weighting margin would mis-rank teams.

---

4. Margin of Victory — Yes, the Men Do Cap It

Both Men’s and Women’s NET cap margin of victory.

The NCAA does not publish the exact number, but multiple reverse-engineering studies show:

• Effective cap ≈ 10–15 points

• Additional scoring beyond the cap provides diminishing or zero benefit

Conceptually:

Adjusted Margin=min⁡(Actual Margin,Cap)Adjusted Margin=min(Actual Margin,Cap)

Purpose:

• Prevent incentive to run up scores

• Reduce sportsmanship concerns

• Limit garbage-time distortion

---

5. Why the Cap Exists (Game Theory Issue)

Without a cap:

Coaches maximize ranking by:

• pressing late

• fouling while ahead

• extending starters unnecessarily

The cap converts incentives from:

“Win by as much as possible”

to

“Play efficiently throughout regulation.”

---

6. Why Women’s NET Still Feels Margin Sensitive

Here’s the subtle point that causes confusion:

Even though MOV is capped…

Efficiency is calculated possession-by-possession.

So late-game defensive lapses can still hurt because:

• defensive efficiency worsens

• offensive efficiency may stagnate

• lineup substitutions change possession outcomes

Result:

Margin isn’t increasing
Efficiency differential is decreasing

That distinction matters enormously.

---

7. Theoretical Summary

Question	Men’s NET Answer	Women’s NET Answer
What carries more signal?	Winning	Performance quality
Talent distribution	Dense	Top-heavy
Upset frequency	Higher	Lower
Efficiency weighting	Moderate	Higher
MOV capped?	Yes	Yes
Late possessions matter	Less	More

---

8. The NCAA’s Underlying Philosophy

The systems diverge because the NCAA is trying to estimate:

True team strength under different competitive ecosystems.

Same sport family.
Different statistical universe.

[recspecs730]

Score differential cap is really quite a challenging statistical issue. In most games, UConn women would be favored by 10-15 points (or even 20), so they are effectively hurt in NET any time they win by more because they didn’t “win by as many as the expectation said they should”. I’d have to think more carefully about how to do this because you don’t really want to incentivize running up the score but caps can also punish good teams.

[WayneSzoke]

I see some IG clips of today’s practice at Yale and it appears as if Fliss is participating.

[PeteD]

Indeed. Here’s a photo posted on the team’s Twitter account a few hours ago. Looking good.

[WayneSzoke]

Let’s Go Fliss!!