LaxPower Back-Up Stick

So Wombat, how’s the anger therapy going?
I’m watching Cornell vs Yale. Why is Morrill in Yale blue with his pedigree?

Hawkeye wrote: ↑Fri May 03, 2019 9:09 pm I will expand more on this tomorrow morning when I post my penultimate bracketology post, but that's the gist of it. In this scenario, Towson's metrics are more in line with Denver and UNC's than they are with the last at-large teams to make the field.

Roughly where these two teams would fall in this scenario
Hopkins: 8 RPI, 1 SOS, 15 QW, 5 RPIW, 3 RPIL
Towson: 10 RPI, 13 SOS, 18 QW, 10 RPIW, 21 RPIL

Not close enough for the head-to-head to flip it in my mind. Maryland is the only team that I could possibly see them getting in over, and it's a long shot in my mind.

Thanks, I appreciate you explaining your reasoning on that. Here would be my response:

RPI: okay, we know Hopkins comes out marginally better in this scenario. Not a dramatic difference, but not zero; if everything else is tied, this will be decisive.

SOS means nothing in a vacuum; it's a way of contextualizing W-L record. So really the second item should be written as 8-7 vs #1 SOS, 10-5 vs #12 SOS. Intuitively these seem about even, i.e, +/-2 wins is roughly the discount you'd expect for moving 10 spots up/down in SOS.

I already listed what Hopkins and Towson's QWs are. Towson's are better. If there's a metric called "QW" that has Hopkins ahead of Towson, that simply discredits that metric rather than changing my opinion of the two teams.

RPI/W and RPI/L: yeah, technically these are criteria on the menu. It's unclear to me if they've ever been ordered. Seriously, I've never given them the slightest bit of attention, and there's never been a time when the committee did something weird and it became explicable if you looked at RPI/W and RPI/L. I don't think there's much evidence these numbers really matter.

...

I just feel like if you're pitching this to an AD who walked into the process cold, and you say to him:

Team A beat Team B

Team A beat a top-10 team that beat Team B

Team A is 10-5 and Team B is 8-7

he'd say, "OK, Team A is obviously better, right?" And then you start to get into "no, no, RPI/W etc." He says, "that's cool, but it doesn't beat out my initial impression of these teams." Are you certain that if you and he are on the committee your voice wins out?

Homer wrote: ↑Fri May 03, 2019 11:11 pm I just feel like if you're pitching this to an AD who walked into the process cold, and you say to him:

Team A beat Team B

Team A beat a top-10 team that beat Team B

Team A is 10-5 and Team B is 8-7

he'd say, "OK, Team A is obviously better, right?" And then you start to get into "no, no, RPI/W etc." He says, "that's cool, but it doesn't beat out my initial impression of these teams." Are you certain that if you and he are on the committee your voice wins out?

It's going to seem like I'm deflecting what you're saying here, but I'm really not trying to do that.

In this argument, where I can't persuade the rest of the committee that Hopkins isn't ahead of Towson for some reason... the argument then becomes Hopkins vs. Maryland for the last spot. And using that same logic that we just put Towson over Hopkins, I think it's safe to say that Hopkins>Maryland. Hopkins is still in.

The debate - what little there is - will center around Maryland, not Hopkins. Cornell/Towson*/Ohio State/pick your flavor against Maryland for the last spot. I think it will fall Maryland's way.

On another note, if Towson wins tomorrow, I think they're actually going to be right on the cusp of hosting a first round game. To me, they really are in an all or nothing situation.

Homer wrote: ↑Fri May 03, 2019 11:11 pm
I just feel like if you're pitching this to an AD who walked into the process cold, and you say to him:

Team A beat Team B

Team A beat a top-10 team that beat Team B

Team A is 10-5 and Team B is 8-7

he'd say, "OK, Team A is obviously better, right?" ....?

IMO, as much as we fans love to [trash] talk about the Transitive Property, it don't work in reality, so point two is out. That leaves, head-to-head and total records. I'm a big fan of head-to-head, but hopefully, as AD I would ask a couple of questions before making a vote, specifically, I'd ask the following for context:

1) Good wins (top 10 or at least against tourney teams)
2) Bad losses (however defined)
3) SOS/RPI to put those records into context. (Did Team A feast on sub-30 teams?)

And finally, I'd ask when that head-to-head was. (Early season, or neutral site in league tourney?)

Granted, not all of these are part of the selection criteria, but that's what I'd ask anyway.

the cornell loss helps hop.
towson and loyola wins will help more.

that a bunch of these guys think they're in ahead of umd, unequivally, is funny. they should be cheering hard for towson and the hounds.

wgdsr wrote: ↑Sat May 04, 2019 12:46 am the cornell loss helps hop.
towson and loyola wins will help more.

that a bunch of these guys think they're in ahead of umd, unequivally, is funny. they should be cheering hard for towson and the hounds.

I got some bad news for you. It also harms the credibility of this post.

Loyola died today.

P.S. Hopkins is in unequivocally over Maryland. But both are most likely going to be in.

DALaxDad wrote: ↑Fri May 03, 2019 10:12 pm So Wombat, how’s the anger therapy going?
I’m watching Cornell vs Yale. Why is Morrill in Yale blue with his pedigree?

I’d give interesting locker room speeches.

Hawkeye wrote: ↑Sat May 04, 2019 1:19 am

wgdsr wrote: ↑Sat May 04, 2019 12:46 am the cornell loss helps hop.
towson and loyola wins will help more.

that a bunch of these guys think they're in ahead of umd, unequivally, is funny. they should be cheering hard for towson and the hounds.

I got some bad news for you. It also harms the credibility of this post.

Loyola died today.

P.S. Hopkins is in unequivocally over Maryland. But both are most likely going to be in.

I’m still concerned by a potential 0-6 record against top-ten RPI teams.

Blue Jays will be fine in terms of RPI, SOS, Avg RPI wins, and Avg RPI losses. However, in some years, the committee seemed to consider record against the top20 RPI above all else, and Hopkins is not particularly strong there.

Hopefully, there won’t be any more ridiculous upsets this weekend in conference tournaments (I’m talking to you, Towson).

Blue Jays should forget all else and just take care of business and defeat Penn State. Tall order, but Hopkins can pull the upset if they play their best game.

DocBarrister (concerned but optimistic)

Hawkeye wrote: ↑Fri May 03, 2019 11:27 pm It's going to seem like I'm deflecting what you're saying here, but I'm really not trying to do that.

In this argument, where I can't persuade the rest of the committee that Hopkins isn't ahead of Towson for some reason... the argument then becomes Hopkins vs. Maryland for the last spot. And using that same logic that we just put Towson over Hopkins, I think it's safe to say that Hopkins>Maryland. Hopkins is still in.

The debate - what little there is - will center around Maryland, not Hopkins. Cornell/Towson*/Ohio State/pick your flavor against Maryland for the last spot. I think it will fall Maryland's way.

On another note, if Towson wins tomorrow, I think they're actually going to be right on the cusp of hosting a first round game. To me, they really are in an all or nothing situation.

Thanks, I appreciate the discussion. I guess I'm still struck by the disparity between the confident tone of the posts asserting that Hopkins is definitively ahead, Towson needs the AQ to qualify, etc., and the way this position seems to dissolve when subjected to even modest scrutiny. At which point you shift ground to making equally unequivocal claims about Hopkins and Maryland. I'm still not sure I'm hearing an actual argument for why one should obviously choose Hopkins over Towson, as opposed to a behavioristic prediction of "this is what the committee" does based on inputs that wouldn't be very persuasive were they offered as explicit justifications.

The key discriminators you cited were:

-- A small difference in RPI

-- A statistic (SOS) that's meaningless unless contextualized, at which point it ceases to be a clear-cut differentiator in this case

-- A pretend statistic (QW) that's obviously wrong

-- A pair of stats (RPI/W and RPI/L) of which I suspect the chief appeal is that they're easy to operationalize in a mathematical model and not that they correspond to anything interesting in reality, except when everything else is held equal

On that last point, here are the RPI's of Hopkins' 8 wins: 11, 11, 17, 21, 24, 32, 34, 35. And here are the RPI's of Towson's top 8 wins: 7, 9, 18, 26, 32, 32, 35, 41. Those average out to 23.1 for Hopkins vs. 25.0 for Towson. The Tigers' average is dragged down by also having beaten Hofstra (45) and Fairfield (52). Poor Tigers, if they'd had the good sense to get blown out by Penn State and Virginia instead of scheduling games they actually won, they'd have a better RPI/W and RPI/L, like Hopkins! Do you see why this is kind of a dumb metric?

The averages for each team's top 4 wins are identical, and both also beat MSM and Delaware, so if we boil this RPI/W argument down to its actual point, it's basically that Towson beat Delaware again and Jacksonville, whereas Hopkins beat Princeton and Michigan, plus maybe let's penalize Towson for the extra games they won.

The case for Hopkins seems to come down to that, plus a couple spots in RPI, plus the fairly subjective notion that Drexel would be a "terrible loss." Does that add up to a convincing justification for moving Hopkins ahead of a team with a better record that they lost to? Maybe for some people it does, and maybe that's exactly how the committee would see it. I just find the confidence with which this (and similar claims about other teams) is being put forward as a quasi-objective fact with which nobody can disagree rather strange.

Homer wrote: ↑Sat May 04, 2019 9:15 am At which point you shift ground to making equally unequivocal claims about Hopkins and Maryland.

The point of the Maryland discussion was this:

If we, as a hypothetical committee, have decided to emphasize a set of parameters where Towson (with a loss today) is evaluated ahead of Hopkins, those same parameters still have Hopkins unequivocally ahead of Maryland. Perhaps even more so than what those of us who are operating under parameters that yield Hopkins>Towson think. We, as a committee, must apply consistent criteria/parameters when evaluating all teams, right? We can't have Towson>Hopkins (largely) on the strength of head-to-head and them completely ignore that Hopkins beat Maryland twice to also come up with Maryland>Hopkins.

I think Towson>Hopkins is already going to require some gymnastics that I do not see the committee performing. And I definitely do not see any reasonable parameters under which Maryland is evaluated ahead of Hopkins. Towson ahead of Hopkins is more reasonable than that.

We wouldn’t be having this life or death discussion if Hopkins had managed to win a few more fricking games.

Going through the season and getting six losses is no way to chart and guarantee your path into the playoffs.

And recent euphoria and koolaid not withstanding, the fact remains that this was a very lousy season with the exception of dominating the meatheads twice at the end.

Anything that happens now is all gravy and deserved, one way or the other.

WOMBAT, Mod Emeritus wrote: ↑Sat May 04, 2019 9:31 am Going through the season and getting six losses is no way to chart and guarantee your path into the playoffs.

Hopkins clearly should apply RRR's idea that only W-L record matters and schedule the entirety of the MAAC and NEC each year. (not serious)

Hawkeye wrote: ↑Sat May 04, 2019 9:32 am

WOMBAT, Mod Emeritus wrote: ↑Sat May 04, 2019 9:31 am Going through the season and getting six losses is no way to chart and guarantee your path into the playoffs.

Hopkins clearly should apply RRR's idea that only W-L record matters and schedule the entirety of the MAAC and NEC each year. (not serious)

Best to ignore R3. Your brain will thank you.

WOMBAT, Mod Emeritus wrote: ↑Sat May 04, 2019 9:42 am Best to ignore R3. Your brain will thank you.

You said it, not me.

WOMBAT, Mod Emeritus wrote: ↑Sat May 04, 2019 9:31 am We wouldn’t be having this life or death discussion if Hopkins had managed to win a few more fricking games.

Going through the season and getting six losses is no way to chart and guarantee your path into the playoffs.

And recent euphoria and koolaid not withstanding, the fact remains that this was a very lousy season with the exception of dominating the meatheads twice at the end.

Anything that happens now is all gravy and deserved, one way or the other.

Maybe if some of the kids getting more run in the last couple of games had gotten more playing time earlier, Hop wouldn't have had those second half collapses. Even two more wins would have eliminated this pins and needles situation Hop is now in. Maybe Hop just matches up well against Maryland and tonight's results against the Lions will be similar to what happened in Happy Valley. I think better "win to be in", but I am no expert. If Hop gets thumped tonight, not a good image for the committee to consider. (error of recency) If Hop get handled tonight and/or dismantled in the first round of the tournament, I bet the discussion on here goes back to what was being said before the two "Terp Takedowns".

Despite all that, still a fan - Go Hop, We Want More

DougELax wrote: ↑Sat May 04, 2019 9:54 am If Hop gets thumped tonight, not a good image for the committee to consider. (error of recency)

Would it be any worse of an image than the one Maryland left the committee with? Or Cornell? Or if Towson loses to a middling Drexel team? Or Cuse losing in the first round of its conference's fake tournament? Most at-large teams leave the committee with a bad aftertaste—that's why they're at-large teams and not AQ conference winners.

I agree with Hawkeye. I like Towson's chances better than he does but I still don't understand why this hypothetical scenario is pitting Towson vs. Hopkins for the last at-large spot. If the committee is valuing head-to-head then Maryland is not getting in over Hopkins. If the committee values RPI and average RPI win/loss then none of them are getting in over Hopkins, but especially not Cornell. The Jays have more "outs" than any of these teams on the bubble, IMO. Is Hopkins GUARANTEED, with 100% absolute certainty to be in? Of course not. But I have a very hard time believing they do not select the #8 or 9 RPI team with two wins over Maryland, who will finish outside the top 10 RPI. There's a reason the Pat Stevens of the world, who are usually pretty good at predicting the committee's thinking, has Hopkins safely in the field.

In 2016 when Hopkins got in over Rutgers despite Rutgers' two wins, it was because Rutgers was like #20 in RPI (I think Hop might have been #10?) They don't consider teams outside the top 15. In fact even #12-13 is pushing it. You can argue with that all you want. Maybe it's unfair. But it's how they have always operated. So this is not remotely the same situation as 2016. If Hop was 15+ or so in RPI then there'd be a bigger discussion. But not only do they have the H2H wins but they're also ahead in RPI. Bad comparison all around.

Hawkeye wrote: ↑Sat May 04, 2019 9:30 am

Homer wrote: ↑Sat May 04, 2019 9:15 am At which point you shift ground to making equally unequivocal claims about Hopkins and Maryland.

The point of the Maryland discussion was this:

If we, as a hypothetical committee, have decided to emphasize a set of parameters where Towson (with a loss today) is evaluated ahead of Hopkins, those same parameters still have Hopkins unequivocally ahead of Maryland. Perhaps even more so than what those of us who are operating under parameters that yield Hopkins>Towson think. We, as a committee, must apply consistent criteria/parameters when evaluating all teams, right? We can't have Towson>Hopkins (largely) on the strength of head-to-head and them completely ignore that Hopkins beat Maryland twice to also come up with Maryland>Hopkins.

I think Towson>Hopkins is already going to require some gymnastics that I do not see the committee performing. And I definitely do not see any reasonable parameters under which Maryland is evaluated ahead of Hopkins. Towson ahead of Hopkins is more reasonable than that.

Great moments in SOS uber alles dogmatism:

Team A 11-4
Team B 10-5
Team C 8-7

"There is no reasonable parameter under which B>C and A>C. None!" I think I just saw steam coming out of Runrussellrun's ears.

More realistically: if your primary criterion is quantity and quality of a team's high-end wins, with everything else secondary to that, then I notice

Maryland has 4 top-20 wins, including a top-5
Towson has 3 top-20 wins, including two top-10
Hopkins has 3 top-20 wins, none of them top-10

I don't think this is an idle hypothetical, either. Number of QWs and where they fall in the top 20 generally seems to be a very big consideration, more important IMO than RPI except in extreme cases. Maybe you disagree, but it's simply wrong to say that there's no such thing as a decision strategy that would rate both Towson and Maryland ahead of Hopkins.

Someone please explain to me why Syracuse isn't getting more scrutiny?

https://extra.ncaa.org/solutions/rpi/St ... ty0503.pdf

Is this a serious question?

Wins against:

#15 Army
#9 Hopkins
#4 Duke
#12 Cornell
#17 North Carolina

Not only are they a lock, they're looking at a likely home game.

WOMBAT, Mod Emeritus wrote: ↑Sat May 04, 2019 9:31 am We wouldn’t be having this life or death discussion if Hopkins had managed to win a few more fricking games.

Going through the season and getting six losses is no way to chart and guarantee your path into the playoffs.

And recent euphoria and koolaid not withstanding, the fact remains that this was a very lousy season with the exception of dominating the meatheads twice at the end.

Anything that happens now is all gravy and deserved, one way or the other.

All the Blue Jays really needed was one more win (any win) and they would have been a lock for the tournament.

They aren’t a lock right now.

Play to win ... and they’ll get in.

Beat Penn State!

DocBarrister