2022 D1 Selection Committee

[RopeUnit]

I can’t recall a Top 10 RPI ever being left out.
But wondering has a team with 3 bad losses ever been invited as an at-large?

[10stone5]

They’ll look at those losses, not just bad losses,

and especially this last game with Notre Dame
stands out like a sore thumb.

[keno in reno]

They’ll look at those losses, not just bad losses,

and especially this last game with Notre Dame
stands out like a sore thumb.

this comes back to the topic of the selection committee...how much can Breschi influence this and is he a friend of Dano or a foe of Duke?

[Farfromgeneva]

They’ll look at those losses, not just bad losses,

and especially this last game with Notre Dame
stands out like a sore thumb.

this comes back to the topic of the selection committee...how much can Breschi influence this and is he a friend of Dano or a foe of Duke?

If Breschi really were to be a scale tipping influence for Duke overcoming negative committee sentiment overall that would say a whole lot for the insular, clubby nature of the game and it’s future growth prospects. UNC pushing for Dukes inclusion and such.

[rolldodge]

LAF’s methodology:

“ Accuracy: 53/56 (99%) for last 7 years excluding 2021 (8/8)

The sum of a teams rpi + qw +1/2 sos is a key indicator of the at-large probability. The Division I Mens
Lacrosse Championship is a 17-team, single-elimination tournament. Nine conferences are eligible for
automatic qualification. However, the two lowest-ranked AQ teams, regardless of conference RPI, will
compete in play-in games to determine the final team in the 16-team bracket. The remaining teams are selected
at-large. If a team has an AQ bid then its at-large probabilility goes to 0.00.

Primary Criteria for Selecting At-Large Teams

[1] Strength of Schedule (SOS) Only Top 10 based on RPI*
[2] Results of Rating Pertcentage Index (RPI)
[3] Record against teams ranked 1-5, 6-10, 11-20, and 21+
[4] Average RPI win (average RPI of all wins)
[5] Average RPI loss (average RPI of all losses)
[-] Head-to-head competition
[-] Results against common opponents
[-] Significant wins (defeating teams ranked higher in RPI)
[-] Significant losses (defeated by teams ranked lower in RPI)
[-] Location of the contests

Note that the SOS on Laxbytes computes all games whereas the SOS appearing here considers only the
10 highest ranked teams. The NCAA has excluded Hampton from the RPI calculation.”

[a fan]

RPI, as everyone knows, is a failed methodology given the small sample sizes. On top of that, the RPI differences between the four or five bubble teams is statistically insignificant. If the committee leans more toward quality wins (vs 1-5, 6-10) vs poor loses, the final selection or two could be different than I think most people are anticipating.

I despise RPI. It's a stupid system that rewards losses to "good" teams.

That said, that's the number they use to determine both SOS and QW. Doesn't make a whole lot of sense to do that, an ignore such a high RPI. Duke is at #8 as of this morning.

In other words, you say that there's little statistical difference in RPI....yet the NCAA values a win against an RPI #8 much higher than a win against an RPI#14.

So it doesn't make a whole lot of sense to use RPI for both QW and SOS...and yet at the same time, ignore Duke's high RPI of #8 at the end of the season. It would basically be saying that a win over Duke is a high QW in the eyes of the committee...yet at the same time, Duke doesn't merit a tournament bid. That doesn't make sense to me.

Duke has wins over (as of this morning) #9 UVa, #12 Richmond, #18 UNC, #18 UNC, and #20 Denver. Five top 20 wins, and a #8 RPI? Tough to not give them a bid, if you ask me. But I'm not in charge.

[rolldodge]

He’s gotten 3 bids wrong out of 56 over the last 7 years. Not counting 2021.

LAF does nice work but is that the right way to think about it when 50-80% of those at larges were slam dunks over that period? Isn’t it really measuring like the last two ( or maybe 3) in? That would be more like 3/14 to 3/21 wouldn’t it?
[/quote]

Yeah, exactly this. I really admire all the contributions Laf has made and continues to make to this community, but touting that stat is sort of ridiculous. Most years at least 6 at-larges are completely noncontroversial, and often it's 7. Usually there's at most one spot that's in the "toss-up" range, and I bet if you looked at posters' opinions on here in the run-up to selection, probably it'd be about 70-30 in terms of guessing right on who gets that last spot. If we say maybe 10 actually challenging picks over a 7 year period, and Laf's missed on 3, that's the same 70% rate.

I also have no idea what the numbers listed as (apparently) selection probabilities on his site are actually supposed to mean. Even aside from the problem of not updating to reflect known facts -- e.g., Richmond's probability of selection is 100%, not 34%, and Hopkins's probability is 0%, not 5% -- am I really supposed to read this to mean Ohio State has a 98% chance of being selected?? That seems like it's wrong by, I don't know, a factor of about two? What exactly does it mean that Hofstra's chances of making the tournament are 4% and UNC's are 0%? That a team with Hofstra's resume gets selected about once every 25 years?
[/quote]


This is revisionist history. Before LAF started doing his calculations, peoples expectations were based on things like the coaches poll and controversy was more than one or two bubble teams. His record is indeed impressive in that context and has driven our discussion now to the point where we are only arguing about one or two teams. For example, No one is any longer under the false impression that Notre Dame is going to get a top seed like the polls might suggest

[Farfromgeneva]

I recall some of these conversations on Lp which would predate this. It was really that 07-09 selection stretch that had everyone focus. Especially 07. Then having ND get in instead of GT in 2010 and making it to the finals so I sort of disagree that the discussion didn’t exist. He may have illuminated a lot of folks but to say no one understood before then I recall differently with a very high degree of confidence.

Heck there was one person who migrated over who showed up at the tail end of LP, a Rutgers fanatic, who refused to listen for two straight years to how the methodology is applied and why they missed the cut. At least one year to Hop. Did not listen to numerous people explaining it. And that would be before 7yrs ago

[rolldodge]

I recall some of these conversations on Lp which would predate this. It was really that 07-09 selection stretch that had everyone focus. Especially 07. Then having ND get in instead of GT in 2010 and making it to the finals so I sort of disagree that the discussion didn’t exist. He may have illuminated a lot of folks but to say no one understood before then I recall differently with a very high degree of confidence.

Heck there was one person who migrated over who showed up at the tail end of LP, a Rutgers fanatic, who refused to listen for two straight years to how the methodology is applied and why they missed the cut. At least one year to Hop. Did not listen to numerous people explaining it. And that would be before 7yrs ago

Fair enough. I agree with your recollection But LAF took those discussions and made a rational system out of them that could be pointed to and evaluated year to year.

[Farfromgeneva]

The criticism of LAF her seems bizarre down to transposing logos so I’m all for defending his efforts. Like complaining about lacrossereferences analysis as if these guys are either idiots or biased when they’re both working hard in an agnostic, quantitative information in order to provide a bette experience for others.

“Yo my free steak wasn’t good I want another!”

“But you finished it”

“Yeah but it wasn’t perfect”

[runrussellrun]

RPI, as everyone knows, is a failed methodology given the small sample sizes. On top of that, the RPI differences between the four or five bubble teams is statistically insignificant. If the committee leans more toward quality wins (vs 1-5, 6-10) vs poor loses, the final selection or two could be different than I think most people are anticipating.

I despise RPI. It's a stupid system that rewards losses to "good" teams.

That said, that's the number they use to determine both SOS and QW. Doesn't make a whole lot of sense to do that, an ignore such a high RPI. Duke is at #8 as of this morning.

In other words, you say that there's little statistical difference in RPI....yet the NCAA values a win against an RPI #8 much higher than a win against an RPI#14.

So it doesn't make a whole lot of sense to use RPI for both QW and SOS...and yet at the same time, ignore Duke's high RPI of #8 at the end of the season. It would basically be saying that a win over Duke is a high QW in the eyes of the committee...yet at the same time, Duke doesn't merit a tournament bid. That doesn't make sense to me.

Duke has wins over (as of this morning) #9 UVa, #12 Richmond, #18 UNC, #18 UNC, and #20 Denver. Five top 20 wins, and a #8 RPI? Tough to not give them a bid, if you ask me. But I'm not in charge.

we're sorry......Maryland has beaten how many teams playing (possibly) next weekend ? Not smelling much taste, in the Terps OCC bland schedule.

Beat UAlbany.....a team trounced by AE winner Vermont. neato. Throw in a sprigg of Patriot league expired Loyola and a pinch of Princeton pepper....and BLAND.....all you got for out of conference flavoring ? CAA ? NEC ? Albert Stryrons Store? WHo won the Big East match up, Gtown or Maryland.

rpi is pathetic....especially with the bland seasoning and lack of playing , or NOT playing.....half of Div. 1 leagues....at all.

Patriot League championship game a great watch. Wonder how Hopkins would do against BU

[masondixonlax]

RPI, as everyone knows, is a failed methodology given the small sample sizes. On top of that, the RPI differences between the four or five bubble teams is statistically insignificant. If the committee leans more toward quality wins (vs 1-5, 6-10) vs poor loses, the final selection or two could be different than I think most people are anticipating.

I despise RPI. It's a stupid system that rewards losses to "good" teams.

That said, that's the number they use to determine both SOS and QW. Doesn't make a whole lot of sense to do that, an ignore such a high RPI. Duke is at #8 as of this morning.

In other words, you say that there's little statistical difference in RPI....yet the NCAA values a win against an RPI #8 much higher than a win against an RPI#14.

So it doesn't make a whole lot of sense to use RPI for both QW and SOS...and yet at the same time, ignore Duke's high RPI of #8 at the end of the season. It would basically be saying that a win over Duke is a high QW in the eyes of the committee...yet at the same time, Duke doesn't merit a tournament bid. That doesn't make sense to me.

Duke has wins over (as of this morning) #9 UVa, #12 Richmond, #18 UNC, #18 UNC, and #20 Denver. Five top 20 wins, and a #8 RPI? Tough to not give them a bid, if you ask me. But I'm not in charge.

we're sorry......Maryland has beaten how many teams playing (possibly) next weekend ? Not smelling much taste, in the Terps OCC bland schedule.

Beat UAlbany.....a team trounced by AE winner Vermont. neato. Throw in a sprigg of Patriot league expired Loyola and a pinch of Princeton pepper....and BLAND.....all you got for out of conference flavoring ? CAA ? NEC ? Albert Stryrons Store? WHo won the Big East match up, Gtown or Maryland.

think you're alone on this opinion

[CU77]

LAF’s methodology:

“ Accuracy: 53/56 (99%) for last 7 years excluding 2021 (8/8)

53/56 = 94.6%

[a fan]

I recall some of these conversations on Lp which would predate this. It was really that 07-09 selection stretch that had everyone focus. Especially 07. Then having ND get in instead of GT in 2010 and making it to the finals so I sort of disagree that the discussion didn’t exist. He may have illuminated a lot of folks but to say no one understood before then I recall differently with a very high degree of confidence.

Heck there was one person who migrated over who showed up at the tail end of LP, a Rutgers fanatic, who refused to listen for two straight years to how the methodology is applied and why they missed the cut. At least one year to Hop. Did not listen to numerous people explaining it. And that would be before 7yrs ago

Fair enough. I agree with your recollection But LAF took those discussions and made a rational system out of them that could be pointed to and evaluated year to year.

Yes, he did!

Any of you following the monkey wrench's today? Penn is about to take down Yale, and BU is up on Army late in the 4th.

Crazy year! I love it!

[wgdsr]

LAF’s methodology:

“ Accuracy: 53/56 (99%) for last 7 years excluding 2021 (8/8)

53/56 = 94.6%

rpi : 38/40 = 95%.
last 3 years 100%.

[runrussellrun]

RPI, as everyone knows, is a failed methodology given the small sample sizes. On top of that, the RPI differences between the four or five bubble teams is statistically insignificant. If the committee leans more toward quality wins (vs 1-5, 6-10) vs poor loses, the final selection or two could be different than I think most people are anticipating.

I despise RPI. It's a stupid system that rewards losses to "good" teams.

That said, that's the number they use to determine both SOS and QW. Doesn't make a whole lot of sense to do that, an ignore such a high RPI. Duke is at #8 as of this morning.

In other words, you say that there's little statistical difference in RPI....yet the NCAA values a win against an RPI #8 much higher than a win against an RPI#14.

So it doesn't make a whole lot of sense to use RPI for both QW and SOS...and yet at the same time, ignore Duke's high RPI of #8 at the end of the season. It would basically be saying that a win over Duke is a high QW in the eyes of the committee...yet at the same time, Duke doesn't merit a tournament bid. That doesn't make sense to me.

Duke has wins over (as of this morning) #9 UVa, #12 Richmond, #18 UNC, #18 UNC, and #20 Denver. Five top 20 wins, and a #8 RPI? Tough to not give them a bid, if you ask me. But I'm not in charge.

we're sorry......Maryland has beaten how many teams playing (possibly) next weekend ? Not smelling much taste, in the Terps OCC bland schedule.

Beat UAlbany.....a team trounced by AE winner Vermont. neato. Throw in a sprigg of Patriot league expired Loyola and a pinch of Princeton pepper....and BLAND.....all you got for out of conference flavoring ? CAA ? NEC ? Albert Stryrons Store? WHo won the Big East match up, Gtown or Maryland.

think you're alone on this opinion

What "opinion" would that be? Thought what was written was pretty factual. Maryland has beaten NOT one team that IS heading to the n$aa's. Not one.

Princeton couldn't even make it's own league tournament....yet are getting an invite seems logical


carry on

[CU77]

Final RPI after Penn & BU wins:

1. Maryland 0.6920
2. Georgetown 0.6536
3. Princeton 0.6519
4. Penn 0.6514
5. Rutgers 0.6294
6. Yale 0.6235
7. Duke 0.6121
8. Cornell 0.6103
9. Virginia 0.6051
10. Brown 0.6038
11. Notre Dame 0.6019
12. BU 0.5990
13. Richmond 0.5936
14. Harvard 0.5863
15. OSU 0.5863

[FlyEaglesFly]

I posted this in the Gtown thread, but I have the feeling either Gtown or MD is gonna get screwed in the quarters. Watching all these Ivy teams, I would have to imagine the top 2 seeds would much rather play them then the lower seeded ACC in the quarters.

[masondixonlax]

RPI, as everyone knows, is a failed methodology given the small sample sizes. On top of that, the RPI differences between the four or five bubble teams is statistically insignificant. If the committee leans more toward quality wins (vs 1-5, 6-10) vs poor loses, the final selection or two could be different than I think most people are anticipating.

I despise RPI. It's a stupid system that rewards losses to "good" teams.

That said, that's the number they use to determine both SOS and QW. Doesn't make a whole lot of sense to do that, an ignore such a high RPI. Duke is at #8 as of this morning.

In other words, you say that there's little statistical difference in RPI....yet the NCAA values a win against an RPI #8 much higher than a win against an RPI#14.

So it doesn't make a whole lot of sense to use RPI for both QW and SOS...and yet at the same time, ignore Duke's high RPI of #8 at the end of the season. It would basically be saying that a win over Duke is a high QW in the eyes of the committee...yet at the same time, Duke doesn't merit a tournament bid. That doesn't make sense to me.

Duke has wins over (as of this morning) #9 UVa, #12 Richmond, #18 UNC, #18 UNC, and #20 Denver. Five top 20 wins, and a #8 RPI? Tough to not give them a bid, if you ask me. But I'm not in charge.

we're sorry......Maryland has beaten how many teams playing (possibly) next weekend ? Not smelling much taste, in the Terps OCC bland schedule.

Beat UAlbany.....a team trounced by AE winner Vermont. neato. Throw in a sprigg of Patriot league expired Loyola and a pinch of Princeton pepper....and BLAND.....all you got for out of conference flavoring ? CAA ? NEC ? Albert Stryrons Store? WHo won the Big East match up, Gtown or Maryland.

think you're alone on this opinion

What "opinion" would that be? Thought what was written was pretty factual. Maryland has beaten NOT one team that IS heading to the n$aa's. Not one.

Princeton couldn't even make it's own league tournament....yet are getting an invite seems logical


carry on

...you're an idiot. "They has beat NOT one team that IS heading the the NCAA's"... Rutgers is going to the NCAA's, as is Notre Dame, UVA, and possibly Ohio state

[a fan]

Final RPI after Penn & BU wins:

1. Maryland 0.6920
2. Georgetown 0.6536
3. Princeton 0.6519
4. Penn 0.6514
5. Rutgers 0.6294
6. Yale 0.6235
7. Duke 0.6121
8. Cornell 0.6103
9. Virginia 0.6051
10. Brown 0.6038
11. Notre Dame 0.6019
12. BU 0.5990
13. Richmond 0.5936
14. Harvard 0.5863
15. OSU 0.5863

Can't see the Committee leaving off the #7 RPI....using a system that is ENTIRELY built on RPI. In other words, QW metric is based on RPI rankings. And the SOS metric is based on RPI metrics.

We'll find out soon enough!