2022 D1 Selection Committee

[crazyhorse]

Never has a team with a top 10 rpi been left out of the tourney, much less the 7 rpi. and left out for the 15 rpi?

The fact that Duke had a #7 RPI invalidates the very notion of RPI.

Their best win was against the #9 RPI team. They lost to 3 teams with sub-20 RPIs - including to RPI 30 Syracuse - and lost a total of 6 games. A majority of their wins were against teams with RPIs in the mid-20 to 50-ish range.

The fact that that translates to a #7 RPI tells me that RPI is something that can be wholly disregarded.

Duke (and others) have figured out how to game the RPI system - lots of games against mediocre teams from mid-level conferences likely to end up with good records. Ironically, it doesn't even matter if you win or lose these games. This year the committee looked at their "accomplishments" and said "nyet." Good for them.

But it doesn't appear to be disregarded when it comes to seeding teams. Here's the top 8 seeds, with their RPI in parentheses:
1. Maryland (1)
2. Georgetown (2)
3. Penn (4)
4. Yale (6)
5. Princeton (3)
6. Rutgers (5)
7. Cornell (8)
8. Brown (10).

Looks pretty tight to me, and definitely not being disregarded. Duke gets knocked out, as they should, because they had three significant losses (which is part of the selection criteria) to Jacksonville, Loyola, and Syracuse. You can't respect the RPI when it comes to seeding teams, but then selectively disrespect Duke's RPI when it comes to giving ND credit for two top 10 wins.

[molo]

Do I think Duke and ND are better than some teams that got in? Probably, but whether or not you like the committee, they gave effectively said that records matter. The unprecedentedly bad performance by most of the ACC teams diluted the quality of wins against historically good teams who had bad years. Like fatty or not, he has long said that winning matters. Duke’s habit of losing to underdogs but cruising into the tournament finally caught up to them.
As an alum of the one remaining ACC team, I’m not thrilled about having to fly to a first round game, but I’m not crying about it. They lost three games, one a blowout against the anticipated NC, one a loss to a mid major that wasn’t close, and another blowout to a team that did not make the tournament. Rather than complain, it’s time to pull an Al Davis. Win or go home.

[suitcase10]

Never has a team with a top 10 rpi been left out of the tourney, much less the 7 rpi. and left out for the 15 rpi?

The fact that Duke had a #7 RPI invalidates the very notion of RPI.

Their best win was against the #9 RPI team. They lost to 3 teams with sub-20 RPIs - including to RPI 30 Syracuse - and lost a total of 6 games. A majority of their wins were against teams with RPIs in the mid-20 to 50-ish range.

The fact that that translates to a #7 RPI tells me that RPI is something that can be wholly disregarded.

Duke (and others) have figured out how to game the RPI system - lots of games against mediocre teams from mid-level conferences likely to end up with good records. Ironically, it doesn't even matter if you win or lose these games. This year the committee looked at their "accomplishments" and said "nyet." Good for them.

Do u really think Duke has figured out gaming the system and do , and other schools haven’t and don’t ? Come on. U think Dino is in a smoky room of quants
When he makes
His schedule? ( funny visual though ).

And this site has been debating for years the validity of RPI with regards to the small lax schedule, and there always head scratchers in the RPI .

Thank u crazy horse for posting criteria on site , again . It’s been up many times.

[rolldodge]

lots of opinions there.

youve seen the history of the comittee and its adherence to rpi. like it or not. Never has a team with a top 10 rpi been left out of the tourney, much less the 7 rpi. and left out for the 15 rpi? come on . a week ago, when someone said cornell has had a meh second half of the season, or that brown start OOC was average to below average, or that uva has gotten smoked agaisnt 2 of the top 3 teams they played, or rutgers- what did they do?-the answer was , and had to be " look at their rpi" . which was fine, right up until the comittee decided to historically off script.

You might look less uninformed if you'd read the NCAA's criteria for tournament selection. Or maybe you wouldn't.

These are the metrics that the selection committee is supposed to use, straight from the manual:
Strength-of-schedule index.
Results of the RPI:
- Record against ranked teams 1-5; 6-10; 11-15; 16-20; 21+
- Average RPI win (average RPI of all wins)
- Average RPI loss (average RPI of all losses)
Head-to-head competition:
- Results versus common opponents
- Significant wins and losses (wins against teams ranked higher in the RPI and losses against teams ranked lower in the RPI)
- Locations of contests

Where in the criteria does it say "record against tournament teams"?

In my opinion, that's where they went wrong - introducing their own new criteria which devalued the Duke wins. They should have just stuck to the stated criteria if they wanted to avoid all this second guessing.

I agree. "record against tournament teams" introduces circular logic. However, the metrics are technically just guidelines. They can make up whatever they want as they go along. Which should not be the case.

Important to note that "straight RPI" is not part of the criteria. So while it may be an outlier that Duke had such a high RPI, the only RPI that is taken into account is their opponents RPI.

[Typical Lax Dad]

lots of opinions there.

youve seen the history of the comittee and its adherence to rpi. like it or not. Never has a team with a top 10 rpi been left out of the tourney, much less the 7 rpi. and left out for the 15 rpi? come on . a week ago, when someone said cornell has had a meh second half of the season, or that brown start OOC was average to below average, or that uva has gotten smoked agaisnt 2 of the top 3 teams they played, or rutgers- what did they do?-the answer was , and had to be " look at their rpi" . which was fine, right up until the comittee decided to historically off script.

I had Brown on the bubble and I didn’t think Harvard would get an invite but those teams along with ND and Duke were bubble teams. ND lost to good teams and beat teams that were down this year. Duke beat teams in the tournament. I don’t care about them losing to ND head to head. As bad as their season was, my eyeball test says they took on more competition and were more successful at it than ND was.

[HooDat]

and STILL all anyone wants to talk about is ND.... they should be no better than 3rd in the "complaints line"

Jacksonville and Army deserved it FAR more than ND - but only the ND folks are whining....

[HooDat]

must be following the lead of their coach...

[CU77]

I sure miss hockey, where these arguments do not happen. Because hockey uses a strict mathematical formula for tournament selection, and nobody can claim that it wasn't applied.

[Homer]

Duke (and others) have figured out how to game the RPI system - lots of games against mediocre teams from mid-level conferences likely to end up with good records.

Noooooo. That's just about the worst conceivable scheduling strategy under the current system.

Those are difficult games to win -- even a very good team is likely to lose one now and again -- but you gain very little from winning them and take a huge hit to your resume if you lose. "Schedule heavy on teams in the 21-35 RPI range" is usually a recipe for disaster come selection time.

A few conferences -- the Patriot League most notably -- effectively lock their members into playing that type of schedule, which is the main reason getting an AL out of the Patriot is such an uphill slog and rarely happens.

That's just looking at the QW/bad loss part of the equation. Even if you're fixated on maximizing your own RPI, "mediocre teams from mid-level conferences likely to end up with good records" is still exactly what you don't want. They help you in the Opponents' Records column, but hurt you in the Opponents' Opponents' Records column.

What you want are teams with good records that all play each other, so that the teams inside the loop's +.500 record against the teams outside the loop propagates throughout the formula and benefits everybody inside the loop at the others' expense.

That's the standard criticism of RPI, that it gives a subsidy to individually mediocre teams in collectively strong conferences. If that's gaming the system, the classic profile of an offender would be like Brown this year. Looking at Brown's resume and where the committee slotted them actually strikes me (and I know some will find this quite ironic) as weirdly reminiscent of the late-2010s Hopkins teams that everybody used to groan about.

[jrn19]

lots of opinions there.

youve seen the history of the comittee and its adherence to rpi. like it or not. Never has a team with a top 10 rpi been left out of the tourney, much less the 7 rpi. and left out for the 15 rpi? come on . a week ago, when someone said cornell has had a meh second half of the season, or that brown start OOC was average to below average, or that uva has gotten smoked agaisnt 2 of the top 3 teams they played, or rutgers- what did they do?-the answer was , and had to be " look at their rpi" . which was fine, right up until the comittee decided to historically off script.

Georgetown had a Top 10 RPI and was left out in 2010

[nyjay]

Duke (and others) have figured out how to game the RPI system - lots of games against mediocre teams from mid-level conferences likely to end up with good records.

Noooooo. That's just about the worst conceivable scheduling strategy under the current system.

Those are difficult games to win -- even a very good team is likely to lose one now and again -- but you gain very little from winning them and take a huge hit to your resume if you lose. "Schedule heavy on teams in the 21-35 RPI range" is usually a recipe for disaster come selection time.

A few conferences -- the Patriot League most notably -- effectively lock their members into playing that type of schedule, which is the main reason getting an AL out of the Patriot is such an uphill slog and rarely happens.

That's just looking at the QW/bad loss part of the equation. Even if you're fixated on maximizing your own RPI, "mediocre teams from mid-level conferences likely to end up with good records" is still exactly what you don't want. They help you in the Opponents' Records column, but hurt you in the Opponents' Opponents' Records column.

What you want are teams with good records that all play each other, so that the teams inside the loop's +.500 record against the teams outside the loop propagates throughout the formula and benefits everybody inside the loop at the others' expense.

That's the standard criticism of RPI, that it gives a subsidy to individually mediocre teams in collectively strong conferences. If that's gaming the system, the classic profile of an offender would be like Brown this year. Looking at Brown's resume and where the committee slotted them actually strikes me (and I know some will find this quite ironic) as weirdly reminiscent of the late-2010s Hopkins teams that everybody used to groan about.

And have one of more of those teams in the loop get a couple of really good out of the loop wins so that everyone benefits from those big wins (Princeton over Georgetown and Rutgers, as a prime example).

[Gobigred]

In my opinion, that's where they went wrong - introducing their own new criteria which devalued the Duke wins. They should have just stuck to the stated criteria if they wanted to avoid all this second guessing.

Notre Dame's wins over Duke were devalued when the committee drilled down to see what underpinned Duke's seemingly too-high RPI. They found Duke had three losses against RPI 21+ teams. Look closely. That's part of an important selection criterion. The committee didn't make that criterion up. As a result, buh bye Duke. Therefore, buh bye ND's two best wins.

RPI does not look at a team's individual wins and losses, only at how many of each. It provides a rough structure from which to begin ranking teams. A starting point. Debatable how good in lacrosse, but that's what it is. It requires the committee to look deeper into a team's wins and losses to get the rankings right. This committee did that. An example: If a team beats Maryland and loses to Michigan, it would have the same RPI if it had lost to Maryland and beaten Michigan. You need to look at the won-lost detail to see the difference in those two sets of outcomes. They're identical to RPI.

[Homer]

Notre Dame's wins over Duke were devalued when the committee drilled down to see what underpinned Duke's seemingly too-high RPI. They found Duke had three losses against RPI 21+ teams. Look closely. That's part of an important selection criterion. The committee didn't make that criterion up. As a result, buh bye Duke. Therefore, buh bye ND's two best wins.

But I think the concern (at least, the reasonable, non-whiny version) is that that's exactly what the committee did with Notre Dame. Which would not be following the criteria.

The criteria say "when considering Team X for selection, value their win over Team Y according to Team Y's RPI ranking." That's not at all the same thing as "value their win over Team Y according to your independent assessment of Team Y applying all the selection criteria, as you would when considering Team Y for selection."

More broadly, I think something that's maybe getting lost with people splitting into opposing camps on this is that it's entirely possible to think the decision to leave Duke out was perfectly reasonable and consistent with the criteria, while finding the decision on Notre Dame iffy at best. That's basically my view: the committee got it right with Duke, then seemingly got sucked into some sort of weird circular reasoning where they treated their decision on Duke as dictating what to do with ND, and came out with a decision that doesn't make a ton of logical sense.

[crazyhorse]

In my opinion, that's where they went wrong - introducing their own new criteria which devalued the Duke wins. They should have just stuck to the stated criteria if they wanted to avoid all this second guessing.

Notre Dame's wins over Duke were devalued when the committee drilled down to see what underpinned Duke's seemingly too-high RPI. They found Duke had three losses against RPI 21+ teams. Look closely. That's part of an important selection criterion. The committee didn't make that criterion up. As a result, buh bye Duke. Therefore, buh bye ND's two best wins.

RPI does not look at a team's individual wins and losses, only at how many of each. It provides a rough structure from which to begin ranking teams. A starting point. Debatable how good in lacrosse, but that's what it is. It requires the committee to look deeper into a team's wins and losses to get the rankings right. This committee did that. An example: If a team beats Maryland and loses to Michigan, it would have the same RPI if it had lost to Maryland and beaten Michigan. You need to look at the won-lost detail to see the difference in those two sets of outcomes. They're identical to RPI.

Great points, and very well stated. Buh - bye Duke, no doubt. After that, we can agree to disagree on whether they should be changing RPI numbers. I think that's a slippery slope. For instance, if you're looking at significant losses, shouldn't you then circle back and look at Brown's RPI? Losses to two non-tourney teams in UMass and UNC, and if things had broken differently a 3rd in Harvard. Do you knock them back a few spots? And if so, how does that effect the number of top 10 wins for the schools that beat them? Gets really circular.
But I definitely see your point about how the committee applied the criteria - thank you. No need to respond - I'm satisfied and accept defeat on behalf of ND!

[middleAgedBear]

Speaking to some early season Ivy bad losses, can we remind people that the Ivy season starts way behind the rest of d1? Brown's loss to UNC was their 2nd game of the season (and third game in 2 years?), while UNC was on game 6. At home. That matters.

[ICGrad]

Noooooo. That's just about the worst conceivable scheduling strategy under the current system.

Those are difficult games to win -- even a very good team is likely to lose one now and again -- but you gain very little from winning them and take a huge hit to your resume if you lose. "Schedule heavy on teams in the 21-35 RPI range" is usually a recipe for disaster come selection time.

A few conferences -- the Patriot League most notably -- effectively lock their members into playing that type of schedule, which is the main reason getting an AL out of the Patriot is such an uphill slog and rarely happens.

Duke rode exactly this schedule, and lost a good number of those games in doing so, to a #7 RPI, and 9 times out of 10 - hell, 98 times out of 100 - a #7 RPI will get in into the tournament. The committee this year took a very close look @ Duke's resume, mostly because some of those losses stood out, but my guess is most years, with most committees, Duke is in. (And heck, look at all the hell the committee is taking for leaving out a team with the #7 - the #7! - RPI this year. "It's never been done before! Cats and Dogs living together! Oh My Lordy!!!").

I'll be interested to see if, going forward, the committee continues to scrutinize the inflated RPIs of some of these teams. If so, fantastic, and you are 100% correct; this is a bad scheduling strategy. But the fact that Duke rode that exact schedule and those results to a #7 RPI and being on the verge of an at-large bid tells me this is not a bad scheduling strategy at all, and that it's not the last time we see it.

[HGK]

And apologize if this has been answered but given Duke lost to three non top 20 teams how exactly was their RPI 7?

[DocBarrister]

I’ve been on the committee and you cannot just deviate from the selection criteria. The “eye test” is just a manifestation of the old boys network. As I have thought about this overnight, I am coming down on the side of “this is an exciting and novel moment for men’s lacrosse”.

—Coach Dom Starsia on Twitter

DocBarrister

[Hoya]

I’ve been on the committee and you cannot just deviate from the selection criteria. The “eye test” is just a manifestation of the old boys network. As I have thought about this overnight, I am coming down on the side of “this is an exciting and novel moment for men’s lacrosse”.

—Coach Dom Starsia on Twitter

DocBarrister

Dom - thank you.

[ICGrad]

And apologize if this has been answered but given Duke lost to three non top 20 teams how exactly was their RPI 7?

And their best win was over the #9 RPI team, and they lost 6 games. It's a formula, dude...what more can you say?

I mean, nothing screams #7 more than 0 wins - hell, 0 games - against the top 5, a single top 10 win, and a bunch of loses.