Louisville Game Combinations

[BBallF]

The following are the combinations of players on the court in tonight's game in chronological order:

1) Ash, Azzi, KK, Sarah, Serah
2) K9, Jana, Azzi, Sarah, Ash
3) Serah, KK, K9, Ash, Jana
4) Sarah, Ash, KK, Azzi, K9
5) KK, K9, Serah, Sarah, Ash
6) Ash, Azzi, Kk, Sears, Sarah
7) Azzi, Sarah, KK, Ice, Ash
8) Allie, Sarah, Ice, Azzi, KK
9) K9, Allie, Ice, Azzi, Sarah
10) Ash, Sarah, Allie, Azzi,K9
11) Ash, Azzi, Sarah, Serah, KK
12) Ash, Azzi, Sarah, Jana, KK
13) K9, Azzi, Sarah, Jana, Ash
14) Ash, Azzi, K9,Serah, KK
15) Sarah, K9, Serah, KK, Azzi
16) Ash, Serah, Sarah, KK, K9
17) Azzi, KK, Ash, Sarah, Serah
18) Azzi, KK, Ash, Sarah, Ice
19) Azzi, K9, KK, Sarah, Ice

We are definitely looking for the right combinations including this year's versions of small ball. See #4 and #14 above. Only 9 players played but the amount of combinations would indicate that Geno is searching for chemistry among the various groupings. What do you conclude when looking at this list?

 

[Doctor G]

Lots of experiments but Geno definitely prefers having Sarah on the court. Will be interesting to see the Blanca combos.

 

[Aluminny69]

Lots of experiments but Geno definitely prefers having Sarah on the court. Will be interesting to see the Blanca combos.

Both Strong and Fudd played 38 minutes each. UConn looks like a very "average" team without those two.

 

[Centerstream]

Both Strong and Fudd played 38 minutes each. UConn looks like a very "average" team without those two.

Of course they would look average. UConn has multiple players that are still adjusting to playing on the team, one of whom was injured prior the game.
Plus most teams would look very average without their two leading scorers.

 

[oldude]

Only 2,984 possible combinations to go… 😁

 

[BBallF]

In rewatching the first two quarters, Geno used a rotation off his starting team to substitute while maintaining the press. It was our best look last night. K9 and KK are an intriguing backcourt and with Sarah and another larger rebounder and scorer with speed, like Blanca, and a shooter like Azzi, you have a very fast, explosive team that has a lot of firepower. It's a combo we couldn't see last night but one that you will. I think you will see the press every first half at least because you almost have to with the personnel we have.

 

[Gael]

The following are the combinations of players on the court in tonight's game in chronological order:

1) Ash, Azzi, KK, Sarah, Serah
2) K9, Jana, Azzi, Sarah, Ash
3) Serah, KK, K9, Ash, Jana
4) Sarah, Ash, KK, Azzi, K9
5) KK, K9, Serah, Sarah, Ash
6) Ash, Azzi, Kk, Sears, Sarah
7) Azzi, Sarah, KK, Ice, Ash
8) Allie, Sarah, Ice, Azzi, KK
9) K9, Allie, Ice, Azzi, Sarah
10) Ash, Sarah, Allie, Azzi,K9
11) Ash, Azzi, Sarah, Serah, KK
12) Ash, Azzi, Sarah, Jana, KK
13) K9, Azzi, Sarah, Jana, Ash
14) Ash, Azzi, K9,Serah, KK
15) Sarah, K9, Serah, KK, Azzi
16) Ash, Serah, Sarah, KK, K9
17) Azzi, KK, Ash, Sarah, Serah
18) Azzi, KK, Ash, Sarah, Ice
19) Azzi, K9, KK, Sarah, Ice

We are definitely looking for the right combinations including this year's versions of small ball. See #4 and #14 above. Only 9 players played but the amount of combinations would indicate that Geno is searching for chemistry among the various groupings. What do you conclude when looking at this list?

BBallF, thanks for taking the time to do that work, which I assume you do manually.

It would be interesting if each lineup could be correlated with some sort of informative cumulative season stat such as PTS, +/-, or PER. That correlation may not "really" be reflective of a lineup's actual effectiveness, but it's something for stat-ophiles to think about and discuss.

I have no experience using Excel or some other algorithmic way to keep track of such correlations, but I can access and post a seasonal stat such as PER, which I suggest as a reasonable correlate for all-around player effectiveness. PER is a complicated formula developed by John Hollinger, who defined it thusly: "The PER sums up all a player's positive accomplishments, subtracts the negative accomplishments, and returns a per-minute rating of a player's performance."

Here are the cumulative 2025-26 PERs for the UConn players after one game, sorted by highest to lowest PER. One could argue that the lineup with the five highest PERs was the most effective.

Rk	Player	Pos	G	GS	MP	PER ▼
1	Sarah Strong	F	1	1	38	31.1
2	KK Arnold	G	1	1	30	26.2
3	Kayleigh Heckel	G	1	0	23	24.6
4	Azzi Fudd	G	1	1	38	17.2
5	Serah Williams	F	1	1	21	14.9
6	Jana El Alfy	C	1	0	8	10.2
7	Ashlynn Shade	G	1	1	28	6.5
8	Ice Brady	F	1	0	9	4.4
9	Allie Ziebell	G	1	0	6	-3.0

The All-PER lineup would be your lineup #15, which has a total PER of 114. Geno's starting lineup (#1) has a total PER of 95.9.

 

[oldhusky]

Still in a somewhat experimental mode. Until Blanca is available, and/or Morgan we don't have any combination with a more natural 3 in the mix. For all of Paige's talents and versatility, the most difficult to replace are the rebounds and blocked shots she provided from the 3 position last year. Paige wasn't out of position there or at the 1 or 2.

She was good on rebounds and blocked shots for a 3. It was really only small ball when she had to be a 4. That is not the case with our other natural guards. Azzi and Kayleigh have been relatively weak rebounders for guards, compared to most 3's severely so. Ash and KK are pretty good but still below average compared to a 3, so with one of those at the 3 it truly is small ball, but with Paige it wasn't.

Blanca or Morgan presumably solves that issue, and of course if we ever played Sarah at the 3, perhaps with Serah at the 4 and Jana at the 5, we could be as good on the boards as anybody.

I don't know what lineups will be common in the long run, but we could go small or big. My main point is Paige didn't have to give up anything as a guard playing the 3. The true guards that may be used there this year would be. The plusses may offset the minuses, and with Sarah and Serah upfront, the bigs can do their part on the boards, but rebounding with any of our natural guards at the 3 will suffer in a way that was not true with Paige there.

 

[CL82]

The following are the combinations of players on the court in tonight's game in chronological order:

1) Ash, Azzi, KK, Sarah, Serah
2) K9, Jana, Azzi, Sarah, Ash
3) Serah, KK, K9, Ash, Jana
4) Sarah, Ash, KK, Azzi, K9
5) KK, K9, Serah, Sarah, Ash
6) Ash, Azzi, Kk, Sears, Sarah
7) Azzi, Sarah, KK, Ice, Ash
8) Allie, Sarah, Ice, Azzi, KK
9) K9, Allie, Ice, Azzi, Sarah
10) Ash, Sarah, Allie, Azzi,K9
11) Ash, Azzi, Sarah, Serah, KK
12) Ash, Azzi, Sarah, Jana, KK
13) K9, Azzi, Sarah, Jana, Ash
14) Ash, Azzi, K9,Serah, KK
15) Sarah, K9, Serah, KK, Azzi
16) Ash, Serah, Sarah, KK, K9
17) Azzi, KK, Ash, Sarah, Serah
18) Azzi, KK, Ash, Sarah, Ice
19) Azzi, K9, KK, Sarah, Ice

We are definitely looking for the right combinations including this year's versions of small ball. See #4 and #14 above. Only 9 players played but the amount of combinations would indicate that Geno is searching for chemistry among the various groupings. What do you conclude when looking at this list?

Thanks for taking the time to put this together. There is an awful lot of information here. One thing I thought that was interesting was that Shade didn't leave the floor until the 8th sub. I think she was only absent from four of 19 iterations of players. (That's just a quick count, I could be wrong.)

 

[Doctor G]

Still in a somewhat experimental mode. Until Blanca is available, and/or Morgan we don't have any combination with a more natural 3 in the mix. For all of Paige's talents and versatility, the most difficult to replace are the rebounds and blocked shots she provided from the 3 position last year. Paige wasn't out of position there or at the 1 or 2.

She was good on rebounds and blocked shots for a 3. It was really only small ball when she had to be a 4. That is not the case with our other natural guards. Azzi and Kayleigh have been relatively weak rebounders for guards, compared to most 3's severely so. Ash and KK are pretty good but still below average compared to a 3, so with one of those at the 3 it truly is small ball, but with Paige it wasn't.

Blanca or Morgan presumably solves that issue, and of course if we ever played Sarah at the 3, perhaps with Serah at the 4 and Jana at the 5, we could be as good on the boards as anybody.

I don't know what lineups will be common in the long run, but we could go small or big. My main point is Paige didn't have to give up anything as a guard playing the 3. The true guards that may be used there this year would be. The plusses may offset the minuses, and with Sarah and Serah upfront, the bigs can do their part on the boards, but rebounding with any of our natural guards at the 3 will suffer in a way that was not true with Paige there.

Spot on analysis. Paige and Strong’s positional flexibility and unstoppable two-man game made small ball a lethal matchup problem on offense and defense last year. I, too, think Geno is counting on Blanca to step up and be the true 3 the team needs. I was bummed she couldn’t play against the cardinals and hope she is good to go on Sunday. Her length, strength, explosiveness, and aggression will be a difficult assignment for most other teams and will make a huge difference in numerous potential Husky lineups. She is obviously not the pick and pop three-level terror that Paige is but I believe she, Azzi, and Sarah will create their own magic domination chemistry along with the other players on the floor.

 

[my2cents]

The following are the combinations of players on the court in tonight's game in chronological order:

1) Ash, Azzi, KK, Sarah, Serah
2) K9, Jana, Azzi, Sarah, Ash
3) Serah, KK, K9, Ash, Jana
4) Sarah, Ash, KK, Azzi, K9
5) KK, K9, Serah, Sarah, Ash
6) Ash, Azzi, Kk, Sears, Sarah
7) Azzi, Sarah, KK, Ice, Ash
8) Allie, Sarah, Ice, Azzi, KK
9) K9, Allie, Ice, Azzi, Sarah
10) Ash, Sarah, Allie, Azzi,K9
11) Ash, Azzi, Sarah, Serah, KK
12) Ash, Azzi, Sarah, Jana, KK
13) K9, Azzi, Sarah, Jana, Ash
14) Ash, Azzi, K9,Serah, KK
15) Sarah, K9, Serah, KK, Azzi
16) Ash, Serah, Sarah, KK, K9
17) Azzi, KK, Ash, Sarah, Serah
18) Azzi, KK, Ash, Sarah, Ice
19) Azzi, K9, KK, Sarah, Ice

We are definitely looking for the right combinations including this year's versions of small ball. See #4 and #14 above. Only 9 players played but the amount of combinations would indicate that Geno is searching for chemistry among the various groupings. What do you conclude when looking at this list?

We need a nickname for Serah. Too hard for my old brain to keep track of Sarahs.

 

[BBallF]

Still in a somewhat experimental mode. Until Blanca is available, and/or Morgan we don't have any combination with a more natural 3 in the mix. For all of Paige's talents and versatility, the most difficult to replace are the rebounds and blocked shots she provided from the 3 position last year. Paige wasn't out of position there or at the 1 or 2.

She was good on rebounds and blocked shots for a 3. It was really only small ball when she had to be a 4. That is not the case with our other natural guards. Azzi and Kayleigh have been relatively weak rebounders for guards, compared to most 3's severely so. Ash and KK are pretty good but still below average compared to a 3, so with one of those at the 3 it truly is small ball, but with Paige it wasn't.

Blanca or Morgan presumably solves that issue, and of course if we ever played Sarah at the 3, perhaps with Serah at the 4 and Jana at the 5, we could be as good on the boards as anybody.

I don't know what lineups will be common in the long run, but we could go small or big. My main point is Paige didn't have to give up anything as a guard playing the 3. The true guards that may be used there this year would be. The plusses may offset the minuses, and with Sarah and Serah upfront, the bigs can do their part on the boards, but rebounding with any of our natural guards at the 3 will suffer in a way that was not true with Paige there.

True but Serah is a better rebounder than Jana and with a high PER team as pointed out by Gael with K9, KK, Azzi, Serah and Sarah on the floor, are you really losing any rebounds that Paige would have got at the 3 position? The #15 combination above started the 4th quarter and played 4 minutes. The rebounding was 3 defensive rebounds for Louisville and 2 for us. Louisville only missed 2 shots and we missed 3. All the rebounds were easy going right to the defender. The score was a margin of 21 when the fourth quarter began and was the same when the 4 minutes had elapsed.

So your point is well taken that Blanca and probably Morgan and Caroline are our natural 3s but I really liked combo #15 and I hope we see more of it.

I didn't see the drop off in the rebounding without a true 3 though. I did see it in the "in the lane" defense in the sense that it was open for a drive or to penetrate without any size to stop that 3 position offensive player. That I think is the weakness of combo #15 and even #16 when Ash came in to replace Azzi. It allows drives. Of course, you offset that with an explosive offensive combo and if one of our 3s dropped or Serah didn't walk on an easy lay up it would have been plus 5 at a minimum.

I can see why analytics become addictive. I still place great value on the "eye test" and that #15 combo is really very pleasing to the view.

 

[BBallF]

Thanks for taking the time to put this together. There is an awful lot of information here. One thing I thought that was interesting was that Shade didn't leave the floor until the 8th sub. I think she was only absent from four of 19 iterations of players. (that's just a quick counter I could be wrong.)

I have it at 5 combos for her not being on the court. She is a better rebounder than Azzi, K9 or KK, and Geno has said that he feels "comfortable" when she is on the floor. K9, Azzi and KK are all better drivers than Ash, but Ash is the second best shooter in the group behind Azzi. She is the 4th best dribbler in the group however so there are pluses and minuses all over the place. Again, what combos work the best in a given circumstance against a given team? That's what is being experimented with.

 

[NycUcWbbFan]

BBallF, thanks for taking the time to do that work, which I assume you do manually.

It would be interesting if each lineup could be correlated with some sort of informative cumulative season stat such as PTS, +/-, or PER. That correlation may not "really" be reflective of a lineup's actual effectiveness, but it's something for stat-ophiles to think about and discuss.

I have no experience using Excel or some other algorithmic way to keep track of such correlations, but I can access and post a seasonal stat such as PER, which I suggest as a reasonable correlate for all-around player effectiveness. PER is a complicated formula developed by John Hollinger, who defined it thusly: "The PER sums up all a player's positive accomplishments, subtracts the negative accomplishments, and returns a per-minute rating of a player's performance."

Here are the cumulative 2025-26 PERs for the UConn players after one game, sorted by highest to lowest PER. One could argue that the lineup with the five highest PERs was the most effective.

Rk	Player	Pos	G	GS	MP	PER ▼
1	Sarah Strong	F	1	1	38	31.1
2	KK Arnold	G	1	1	30	26.2
3	Kayleigh Heckel	G	1	0	23	24.6
4	Azzi Fudd	G	1	1	38	17.2
5	Serah Williams	F	1	1	21	14.9
6	Jana El Alfy	C	1	0	8	10.2
7	Ashlynn Shade	G	1	1	28	6.5
8	Ice Brady	F	1	0	9	4.4
9	Allie Ziebell	G	1	0	6	-3.0

The All-PER lineup would be your lineup #15, which has a total PER of 114. Geno's starting lineup (#1) has a total PER of 95.9.

PER is not tractable to be spliced into a rotations analysis (a cohort statistics exercise). One can use others, such as EFFiciency, PIE, NETRTG, +/-, PTS/p, etc.

I did a rotations analysis last year using play-by-play info (example here — I’ll do one when Blanca is healthy). It became clear that maximizing winning NCAAT games is not simply rank ordering rotations by some filter over a given evaluation period. It is all about rank ordering different rotations sequences that maximize winning time combinations.

Small ball last year was Geno’s winning time combination. Yet he did not make it his starting lineup. He described it as cat-mouse in the UCLA game. It’s effective when he uses it after he has taken measure of what the opponent has game-planned for that game after already tiring them out.

Last year, the winning time combinations were: the (Big 3 + Ash) + one of {Kaitlyn, KK}.

This year, because of the potency of various lineups and that this team is a returning champ with a potential trajectory better than last year’s zero-to-South Carolina ramp (after losing their secret scrimmage), winning time combinations are potentially more varied. I am hypothesizing that winning time combinations will be something like:

(Azzi, Sarah) + 2 or 3 or 1 of (Serah, Ash, Blanca) + 1 or 0 or 2 of (KK, Kayleigh).

Anything else is a bridge rotation (e.g. to give Azzi and Sarah rest for a long season and for farm system development). Aside: I think the rotation size — players that Geno trusts to play in competitive minutes of games — is 10/11.

I know it’s one game. But the 1Q of the Louisville game demonstrated spurtability from the starting lineup (it already is a winning time combination from the jump).

 

[BBallF]

We need a nickname for Serah. Too hard for my old brain to keep track of Sarahs.

Serah, Serah? Pronounced like the song? Your idea, think of one.

 

[BBallF]

PER is not tractable to be spliced into a rotations analysis (a cohort statistics exercise). One can use others, such as , PIE, NETRTG, +/-, PTS/p, etc.

I did a rotations analysis last year using play-by-play info (example here — I’ll do one when Blanca is healthy). It became clear that maximizing winning NCAAT games is not simply rank ordering rotations by some filter over a given evaluation period. It is all about rank ordering different rotations sequences that maximize winning time combinations.

Small ball last year was Geno’s winning time combination. Yet he did not make it his starting lineup. He described it as cat-mouse in the UCLA game. It’s effective when he uses it after he has taken measure of what the opponent has game-planned for that game after already tiring them out.

Last year, the winning time combinations were: the (Big 3 + Ash) + one of {Kaitlyn, KK}.

This year, because of the potency of various lineups and that this team is a returning champ with a potential trajectory better than last year’s zero-to-South Carolina ramp (after losing their secret scrimmage), winning time combinations are potentially more varied. I am hypothesizing that winning time combinations will be something like:

(Azzi, Sarah) + 2 or 3 or 1 of (Serah, Ash, Blanca) + 1 or 0 or 2 of (KK, Kayleigh).

Anything else is a bridge rotation (e.g. to give Azzi and Sarah rest for a long season and for farm system development). Aside: I think the rotation size — players that Geno trusts to play in competitive minutes of games — is 10/11.

I know it’s one game. But the 1Q of the Louisville game demonstrated spurtability from the starting lineup (it already is a winning time combination from the jump).

I have a headache now. I do wholeheartedly agree with spurtability which is now a UCONN neologism. I think combo #15 has a high level of spurtability also.

 

[my2cents]

Serah, Serah? Pronounced like the song? Your idea, think of one.

We should stay with the Hall/Oates theme. How about "SMILES", Serah Smiles. " Singing, oh won't you smile for me, Serah".lol

 

[Bone Dog]

Thanks for doing all this work. One thing that leaps off your combo list is that Geno has yet to “feel comfortable” when both Azzi and Sarah sit. 38 minutes is not sustainable for those two, or at least not a great idea. This may be alleviated once Blanca returns, a prospect that seems more important now.

I think some sort of statistical assessment of the combos would be great, or even just minutes as this would show where Geno’s comfort level is with each combo. But even this sounds like a huge task.

 

[msf22b]

Gael:
My problem with the chart you (?) created (supplied, proposed), is that l suspect that there are subjective elements that skew the results.

From just watching the game in its entirety (once) and some highlights
It is absolutely impossible for me to accept an analysis that ranks
Ashlynn’s contribution as half as valuable as Jana’s and
Less than half as Serah’s

Doesn’t pass the eye and smell test.
As well as Geno’s minutes …a form of evaluation…
I do agree with ranking K9 high, though

Numerical analysis is great…but it has to correlate to what one casually observes.

 

[and one]

Stats are fine and useful and a coach ignores them at their peril. However, no coach cedes decision making to the team statistician.

In the case at hand, based on one game, any best combination includes Sarah, Azzi and Serah, plus two of Ash, KK and K9. Jana, Ice and Allie will be useful for resting players and for specific match-ups. Not a knock on them, just right now they aren't as good as the top six. Where Blanca fits into this picture remains to be seen.

 

[BBallF]

We should stay with the Hall/Oates theme. How about "SMILES", Serah Smiles. " Singing, oh won't you smile for me, Serah".lol

Haha. Not unless her smile becomes her trademark. She wasn't smiling very much Tuesday night.

 

[The CDR-Ret]

The following are the combinations of players on the court in tonight's game in chronological order:

1) Ash, Azzi, KK, Sarah, Serah
2) K9, Jana, Azzi, Sarah, Ash
3) Serah, KK, K9, Ash, Jana
4) Sarah, Ash, KK, Azzi, K9
5) KK, K9, Serah, Sarah, Ash
6) Ash, Azzi, Kk, Sears, Sarah
7) Azzi, Sarah, KK, Ice, Ash
8) Allie, Sarah, Ice, Azzi, KK
9) K9, Allie, Ice, Azzi, Sarah
10) Ash, Sarah, Allie, Azzi,K9
11) Ash, Azzi, Sarah, Serah, KK
12) Ash, Azzi, Sarah, Jana, KK
13) K9, Azzi, Sarah, Jana, Ash
14) Ash, Azzi, K9,Serah, KK
15) Sarah, K9, Serah, KK, Azzi
16) Ash, Serah, Sarah, KK, K9
17) Azzi, KK, Ash, Sarah, Serah
18) Azzi, KK, Ash, Sarah, Ice
19) Azzi, K9, KK, Sarah, Ice

We are definitely looking for the right combinations including this year's versions of small ball. See #4 and #14 above. Only 9 players played but the amount of combinations would indicate that Geno is searching for chemistry among the various groupings. What do you conclude when looking at this list?

Only one sub at a time. Therefore, looking at nuance - NOT wholesale change against a ranked opponent. Louisville will be really good this year, as they effectively went toe-to-toe after surrendering a 14-1 onslaught at the start of the first quarter with effective "runs" of their own, and no "give up" in them as they closed to a 10 point margin in the 4th (but no closer).

The UConn 3-pt shooting will improve over the next few games.

Go Huskies!!!

 

[BBallF]

Thanks for doing all this work. One thing that leaps off your combo list is that Geno has yet to “feel comfortable” when both Azzi and Sarah sit. 38 minutes is not sustainable for those two, or at least not a great idea. This may be alleviated once Blanca returns, a prospect that seems more important now.

I think some sort of statistical assessment of the combos would be great, or even just minutes as this would show where Geno’s comfort level is with each combo. But even this sounds like a huge task.

It is for sure. What you can do is just see what the score differential is when different combos are out there. There are several variables including missed shots but they are part of the consideration. This combo shoots better than that one for example. I think part of the 38 minutes for Sarah and Azzi has to do with finding the right combos. There are a lot more possibilities this year as compared to the teams in the last 9 years for example. You also need that third consistent scorer and how that player blends with the team, and with which combo.

 

[Bone Dog]

Numerical analysis is great…but it has to correlate to what one casually observes.

I hear you. Quantitative analysis is limited by its sampling rate and the size of its data set. You can't capture every contribution a player makes without a really unwieldy sample of things she does, and this is especially true of Ash. But a well-chosen sampling of a few things -- say more than just points rebounds assists and turnovers -- can capture enough to give a pretty good picture, if the the data set is really large, like way larger than 35 games. Ash rebounds well and shoots well. But she does a ton more than that just through sheer energy and persistence, and these aspects of her game are really difficult to capture quantitatively. And sometimes, what she does yields a steal or a rebound for a teammate. The eye test is the only way to overcome this, though you have to be a really good observer, and watching video (as I have to, sigh) really limits how much you can see.

 

[The CDR-Ret]

Serah, Serah? Pronounced like the song? Your idea, think of one.

How about "S.W." (Serah Williams, surface warfare qualified) and SS (Sarah Strong, submarine warfare qualified) - both can be nuclear powered in the US Navy... (AI wouldn't let me put just the initials for Serah Williams, so I had to put "S.W.")

Go Navy!

Go Huskies!!!