2024 Bracketology

[HGK]

“All of this other eye test stuff is bias (recency). Do I think the Terps are playing well? No. Anyone can see that. But they've played really well through a couple of different parts of their schedule. No one talks about a team that started out like crap (hello, Georgetown) and penalizes them for that because "they're on a hot streak!"”


Georgetown had their worst loss, best loss and best two wins in their first four games - Loyola, Hopkins, Penn and ND. Terrible example for your thesis as they aren’t anywhere close to a hot streak and had best two RPI wins and best RPI loss before March!

[Stiffler]

Sadly, I am always reminded it isn't the 18 best teams that make the D1 lacrosse tourney. It is the 18 that qualify or picked.

Forget Ivy situation (cause same thing happened to ACC a few years ago when Carc and Quint about had stroke's, live on-air), the seeding is a complete disaster. Uva seeded after getting doors blown off by ND and losing 4 straight. UMD seeded after getting doors blown off by PSU (and PSU barely got in), Denver has higher seed after losing to freakin Villanova, over G'Town won the conference tourney again. Princeton, the winner of one of the top 3 conference tourney, no seed. 50% of the seeds are airballs by the "Committee".

The committee counts every game equally in theory and that is what is annoying fans. They see teams struggling down the stretch and forget about the other 80% of the season. Princeton didnt play a top 10 team in its conference and lost to the same Maryland team you are talkign about. Michigan beat 2 top 10 teams to finish the season (3 overall) and they are unseeded. This is RPI top 10:

1 Notre Dame- #1 seed

2 Duke- #2 seed

3 Johns Hopkins- #3 seed

4 Syracuse- #4 seed

5 Virginia- #5 seed

6 Denver- #6 seed

7 Maryland- #7 seed

8 Princeton

9 Penn State

10 Georgetown - #8 seed

Only one that is off is Princeton and Gtown. Maybe that is based on top 10 wins, which Princeton had none. Gtown had the best win in the country this year..

Until the reliance on RPI is ended, this is basically what people should expect every year. Its going to closely reflect the RPI.

If it is just RPI and that is what people should expect, then let's stop the farce of a committee. No context around wins/losses will be considered to any large degree, I guess. Head to head, who is playing well at the end of the season (like in hoops), conference tourney results are not in consideration. Maybe that is easiest, most fair and best, but it also gives you what happened to the Ivy's this year and the ACC a few years back.

[10stone5]

As far as Play In games go - that's a good one, Albany and Sacred Heart are both playing high level lacrosse at this time.

[laxfanatic]

Time is well beyond expanding the tournament.
D-3 plays weds & sat or sunday why not D1

[coda]

Sadly, I am always reminded it isn't the 18 best teams that make the D1 lacrosse tourney. It is the 18 that qualify or picked.

Forget Ivy situation (cause same thing happened to ACC a few years ago when Carc and Quint about had stroke's, live on-air), the seeding is a complete disaster. Uva seeded after getting doors blown off by ND and losing 4 straight. UMD seeded after getting doors blown off by PSU (and PSU barely got in), Denver has higher seed after losing to freakin Villanova, over G'Town won the conference tourney again. Princeton, the winner of one of the top 3 conference tourney, no seed. 50% of the seeds are airballs by the "Committee".

The committee counts every game equally in theory and that is what is annoying fans. They see teams struggling down the stretch and forget about the other 80% of the season. Princeton didnt play a top 10 team in its conference and lost to the same Maryland team you are talkign about. Michigan beat 2 top 10 teams to finish the season (3 overall) and they are unseeded. This is RPI top 10:

1 Notre Dame- #1 seed

2 Duke- #2 seed

3 Johns Hopkins- #3 seed

4 Syracuse- #4 seed

5 Virginia- #5 seed

6 Denver- #6 seed

7 Maryland- #7 seed

8 Princeton

9 Penn State

10 Georgetown - #8 seed

Only one that is off is Princeton and Gtown. Maybe that is based on top 10 wins, which Princeton had none. Gtown had the best win in the country this year..

Until the reliance on RPI is ended, this is basically what people should expect every year. Its going to closely reflect the RPI.

and everybody knows in advance what goes into rpi. the only time that's drastically changed is 2022 (and 2021). if you have a fluid system (in theory) teams know what's expected, and the teams do, too. there's a base to part of that, and penn as an example has played the rpi game pretty well over time.

the rest on late games meaning more is shouting at clouds, imo. as someone said, that's what conference tournies are for. so is goal differential (no offense, coda). the object is to win games. and cheapening earlier games makes no sense to me. these guys play for a lot of the year, to be prepared from the jump, and when late season pressure is on if you're bubble territory.

I don’t think you cheapen games, but teams either grow over the course of a season or fall back. Various things can contribute to that. It should be accounted for. The committee is using the most basic and out-dated model available (RPI). I think the sport deserves better. I would not be upset, if there was an off-season discussion on how to do a better job.

i read in my opinion to most posts when i can. so that is how i'll read yours, and in this case why i couched mine. i disagree, that's fanlax. they don't erase the games when it comes to who qualifies for a conference's tournaments bc who is playing better lately, either. the whole season matters. and again, most every team has been given an opportunity to turn it on late and be rewarded thru aqs. it's been great for the game, i am all for that change. i love march madness.

rpi is not the problem. it's a short season, you won't find a system that works awesome-sauce with it. the application of it is turrible. it's not supposed to consider straight rpi as even a primary, if anyone reads the actual rulebook. it's supposed to be what rpi's you win and lose to. the committee has gotten morphed into it being a driver.

it could be set up with actual hard based numbers, giving you a result. it's not. it's bunched in groups of 5 and discounting or nicking a win/loss vs #21 rpi, same as team 70. and different than #20. it's dumb.

rpi is not outdated. it's wins and losses. winning by how much is dumb also. imo.

I dont get the score differential hatred. I am not sure how you are supposed to judge the performance. This has proven wrong at every study of modeling sports, but each to their own. This would be like trying to rank 100m racers by simply whether or not they won. Imagine how ridiculous that would be.

[coda]

From a statistical perspective, the issue to me isn't the RPI or Strength of Schedule (SOS) being used as individual composite variables. To me, it's the use of both in conjunction. The RPI is a strength of schedule indicator (opponent's win%, opponent's opponents' win%). RPI and SOS share significant variance. It's the reason why playing Hampton drops a team's RPI even after a win. Is there a multicollinearity issue where they're literally measuring the same thing? Maybe not to that extreme, but I bet the correlation between the two is really strong.

All of this other eye test stuff is bias (recency). Do I think the Terps are playing well? No. Anyone can see that. But they've played really well through a couple of different parts of their schedule. No one talks about a team that started out like crap (hello, Georgetown) and penalizes them for that because "they're on a hot streak!"

Teams play between 12 and 16 games. All games count. The last 2 or 4 don't have any more significance than the first 2 or 4. After all, very few think Albany is going to knock off Notre Dame just because Albany comes in on a heater.

I think people read things as black and white and that is where the issues arise. We can take the Maryland. Their resume makes them a tournament team. That is just a fact. There is also consideration to teams limping in, which there are more than usual. Its not unfair to look at a team that is limping in and say RPI has them at 7, but they have looked lost in the last 3 games. Lets seed them at 9/10 to reflect the current play. The idea isnt to wipe out the season, but to have the ability to recognize that teams change over the course of the season. You can go to the extreme on this. Imagine if the ND team got caught in a drug ring. They lost 12 kids to a season suspension. Their RPI would #1. Should the committee factor that in? Sports are not linear. Teams and players grow at different speeds. Injuries occur that can dramatically change the outlook. I am saying that the committee should have the ability and desire to recognize these things. I generally root for the Big 10, so thins is not sour grapes from me.

[laxreference]

I dont get the score differential hatred. I am not sure how you are supposed to judge the performance. This has proven wrong at every study of modeling sports, but each to their own.

I agree that it is likely going to improve a model that is designed to rank teams. But I question whether the extra benefit of including scoring margin is worth it. Is a ranking that includes goal margin going to be that much better than one that only includes wins and losses?

This is a bit more philosophical, but I think college coaches have two goals with respect to on-field stuff.

1. Win games
2. Develop your players

If you are up 4 goals with 5 minutes left in a game, you have some leeway to rest your starters and get your next generation of contributors some experience.

If scoring margin is included in something that affects resumes/selections, that leeway pretty much goes away. If you are not guaranteed to have an at-large worthy resume, you can't afford to let that 4-goal lead turn into a 2-goal lead.

Again, I'm not suggesting that scoring margin may not produce a better ranking, but there is a trade-off here. And I think that putting an extra constraint on coaches' decision-making, especially one that has nothing to do with the two goals I mentioned above, is not worth the incremental improvement in a ranking system that it would produce.

[coda]

I dont get the score differential hatred. I am not sure how you are supposed to judge the performance. This has proven wrong at every study of modeling sports, but each to their own.

I agree that it is likely going to improve a model that is designed to rank teams. But I question whether the extra benefit of including scoring margin is worth it. Is a ranking that includes goal margin going to be that much better than one that only includes wins and losses?

This is a bit more philosophical, but I think college coaches have two goals with respect to on-field stuff.

1. Win games
2. Develop your players

If you are up 4 goals with 5 minutes left in a game, you have some leeway to rest your starters and get your next generation of contributors some experience.

If scoring margin is included in something that affects resumes/selections, that leeway pretty much goes away. If you are not guaranteed to have an at-large worthy resume, you can't afford to let that 4-goal lead turn into a 2-goal lead.

Again, I'm not suggesting that scoring margin may not produce a better ranking, but there is a trade-off here. And I think that putting an extra constraint on coaches' decision-making, especially one that has nothing to do with the two goals I mentioned above, is not worth the incremental improvement in a ranking system that it would produce.

As you know some models will throw out garbage time. That can avoid things like that. Like I said this is a bigger discussion on how to do it better. You can devise a model and put in parameters like up 12 at halftime, up 7 to start the 4Q, etc. and it turns off. I do think it is important to understand that a top 10 team going to overtime vs the 30th ranked team is a poor result and vice versa.

[Hooz123]

I dont get the score differential hatred. I am not sure how you are supposed to judge the performance. This has proven wrong at every study of modeling sports, but each to their own.

I agree that it is likely going to improve a model that is designed to rank teams. But I question whether the extra benefit of including scoring margin is worth it. Is a ranking that includes goal margin going to be that much better than one that only includes wins and losses?

This is a bit more philosophical, but I think college coaches have two goals with respect to on-field stuff.

1. Win games
2. Develop your players

If you are up 4 goals with 5 minutes left in a game, you have some leeway to rest your starters and get your next generation of contributors some experience.

If scoring margin is included in something that affects resumes/selections, that leeway pretty much goes away. If you are not guaranteed to have an at-large worthy resume, you can't afford to let that 4-goal lead turn into a 2-goal lead.

Again, I'm not suggesting that scoring margin may not produce a better ranking, but there is a trade-off here. And I think that putting an extra constraint on coaches' decision-making, especially one that has nothing to do with the two goals I mentioned above, is not worth the incremental improvement in a ranking system that it would produce.

Scoring margin is a poor measure for the obvious reasons already articulated above. Certain coaches have class and choose to run their reserves when the game is in control. Whereas, a$$holes like John Tillman get off by running his starters and running up the score against his rivals because he's a known di$Khead (see 2022).

[masondixonlax]

I dont get the score differential hatred. I am not sure how you are supposed to judge the performance. This has proven wrong at every study of modeling sports, but each to their own.

I agree that it is likely going to improve a model that is designed to rank teams. But I question whether the extra benefit of including scoring margin is worth it. Is a ranking that includes goal margin going to be that much better than one that only includes wins and losses?

This is a bit more philosophical, but I think college coaches have two goals with respect to on-field stuff.

1. Win games
2. Develop your players

If you are up 4 goals with 5 minutes left in a game, you have some leeway to rest your starters and get your next generation of contributors some experience.

If scoring margin is included in something that affects resumes/selections, that leeway pretty much goes away. If you are not guaranteed to have an at-large worthy resume, you can't afford to let that 4-goal lead turn into a 2-goal lead.

Again, I'm not suggesting that scoring margin may not produce a better ranking, but there is a trade-off here. And I think that putting an extra constraint on coaches' decision-making, especially one that has nothing to do with the two goals I mentioned above, is not worth the incremental improvement in a ranking system that it would produce.

Scoring margin is a poor measure for the obvious reasons already articulated above. Certain coaches have class and choose to run their reserves when the game is in control. Whereas, a$$holes like John Tillman get off by running his starters and running up the score against his rivals because he's a known di$Khead (see 2022).

[Dlaxva5]

In summary - teams "struggling" at the end of the season are primarily playing in conference games and conference tournament games. And if you are in a good conference you could lose 2, 3 or 4 in the last month. All games matter because they typically play out of conference first.

Teams that are on the bubble - play inconsistently; have some good wins, but also some head-scratching losses; did well in their conference, but typically lost one they should not have, and didn't win the AQ for their conference tournament and typically didn't make the conference final.

We need to expand the tournament or change the criteria because some bubble teams didn't make it? The ones "playing well" at the end of the season? The ones high up in the Coaches Poll (no bias in that poll). The teams with 4 and 5 star recruits who didn't play consistently all season??

No need to "improve" or change anything - really nothing to fix. Model is simple - all games matter; win out of conference as much as possible; do well in conference and get to championship game. DON'T have a loss that can hurt you.

Easiest way to "fix" (never going to happen) - Play the conference game first and the conference tournament - have that all done by end of March - you'll know then if you are the AQ or not. Then play all the OOC games in April - how fun would that be!

[NYlax222]

Appreciate the good comments about scoring margin incentivizing teams to run it up, but, to me scoring margin matters alot. Seperating teams via RPI , how many 'good wins', etc is helpful, but, not every win - or loss - is the same. 5 or 10 goal margins tell us something more so than an OT game - win or lose. Realize any single game can be an outlier, so i dont' believe in highlighting single games to prove a point. But, just as there is a difference in most people 's mind between a win vs. 11-20 , vs. a win vs. Top 5, so too there should be a difference between winning by 1 or 2, or by 5-10.

[NYlax222]

One other small point. Losses to teams outside the Top 20 are deemed 'bad'. But getting blown out by 10 to a Top 20 team is not bad?

[Hooz123]

Appreciate the good comments about scoring margin incentivizing teams to run it up, but, to me scoring margin matters alot. Seperating teams via RPI , how many 'good wins', etc is helpful, but, not every win - or loss - is the same. 5 or 10 goal margins tell us something more so than an OT game - win or lose. Realize any single game can be an outlier, so i dont' believe in highlighting single games to prove a point. But, just as there is a difference in most people 's mind between a win vs. 11-20 , vs. a win vs. Top 5, so too there should be a difference between winning by 1 or 2, or by 5-10.

Honestly this is just a bad take. You're trying to tell me that a 10-4 Cuse win over Duke in March is better than an 18-17 thriller OT win in April vs. Virginia? [insert any equivalency and team here]. I'm sorry that makes no sense and it's obvious that "analysis" is entirely subjective from person-to-person and impossible to quantify based on numerous variables. If anything it makes the current problem even more convoluted and FUBAR. How on earth is "scoring margin" going to streamline this?

[DocBarrister]

Appreciate the good comments about scoring margin incentivizing teams to run it up, but, to me scoring margin matters alot. Seperating teams via RPI , how many 'good wins', etc is helpful, but, not every win - or loss - is the same. 5 or 10 goal margins tell us something more so than an OT game - win or lose. Realize any single game can be an outlier, so i dont' believe in highlighting single games to prove a point. But, just as there is a difference in most people 's mind between a win vs. 11-20 , vs. a win vs. Top 5, so too there should be a difference between winning by 1 or 2, or by 5-10.

Honestly this is just a bad take. You're trying to tell me that a 10-4 Cuse win over Duke in March is better than an 18-17 thriller OT win in April vs. Virginia? [insert any equivalency and team here]. I'm sorry that makes no sense and it's obvious that "analysis" is entirely subjective from person-to-person and impossible to quantify based on numerous variables. If anything it makes the current problem even more convoluted and FUBAR. How on earth is "scoring margin" going to streamline this?

I agree that using scoring margin is a bad idea, for a variety of reasons.

Most obviously, running up scores against weak teams shouldn’t be rewarded.

Second, winning close games, I think, is a sign of quality, not a negative thing.

It could be argued that winning close or OT games against quality opponents helps to prepare a team for the NCAA tournament. Nothing tests a team more than a nailbiter decided late by a single goal.

This is a bit like the annual campaign to value pretty win-loss records over SOS. I recall one team in particular entering the NCAA tournament undefeated with a weak SOS. Got tossed in the first round. Running up scores against a weak schedule doesn’t prepare you for the NCAA tournament.

Now, is that comparable to the value of using score differential as a tie breaker among ACC teams? No … the entire ACC has a strong SOS. Then using a “capped” score differential may make some sense.

Frankly, what would have made even more sense is to have invited the entire ACC to the ACC tournament with teams 4 and 5 playing in. That way, UNC wouldn’t have had their season end because they beat Duke by 3 goals instead of 5.


But that’s another discussion.

DocBarrister

[coda]

Appreciate the good comments about scoring margin incentivizing teams to run it up, but, to me scoring margin matters alot. Seperating teams via RPI , how many 'good wins', etc is helpful, but, not every win - or loss - is the same. 5 or 10 goal margins tell us something more so than an OT game - win or lose. Realize any single game can be an outlier, so i dont' believe in highlighting single games to prove a point. But, just as there is a difference in most people 's mind between a win vs. 11-20 , vs. a win vs. Top 5, so too there should be a difference between winning by 1 or 2, or by 5-10.

Honestly this is just a bad take. You're trying to tell me that a 10-4 Cuse win over Duke in March is better than an 18-17 thriller OT win in April vs. Virginia? [insert any equivalency and team here]. I'm sorry that makes no sense and it's obvious that "analysis" is entirely subjective from person-to-person and impossible to quantify based on numerous variables. If anything it makes the current problem even more convoluted and FUBAR. How on earth is "scoring margin" going to streamline this?

Yes, it is.. Beating Virginia is 18-9 is more impressive (ND), than beating Virginia 18-17 (Cuse). Models are not there to gauge the entertainment value of a game. They are there to measure the relative performances of teams. Beating the piss out Virginia is more impressive than squeaking a win out.. Its completely illogical to argue that it isnt. A top 5 team struggling vs a non top 15 team is great television, but still a poor performance. Without using scoring margin it is much harder to measure relative performance.

The comments on running up the score are just ill-informed. Many models disregard "garbage time", so it is easy to limit the affect of running up the score.

[Hooz123]

Appreciate the good comments about scoring margin incentivizing teams to run it up, but, to me scoring margin matters alot. Seperating teams via RPI , how many 'good wins', etc is helpful, but, not every win - or loss - is the same. 5 or 10 goal margins tell us something more so than an OT game - win or lose. Realize any single game can be an outlier, so i dont' believe in highlighting single games to prove a point. But, just as there is a difference in most people 's mind between a win vs. 11-20 , vs. a win vs. Top 5, so too there should be a difference between winning by 1 or 2, or by 5-10.

Honestly this is just a bad take. You're trying to tell me that a 10-4 Cuse win over Duke in March is better than an 18-17 thriller OT win in April vs. Virginia? [insert any equivalency and team here]. I'm sorry that makes no sense and it's obvious that "analysis" is entirely subjective from person-to-person and impossible to quantify based on numerous variables. If anything it makes the current problem even more convoluted and FUBAR. How on earth is "scoring margin" going to streamline this?

Yes, it is.. Beating Virginia is 18-9 is more impressive (ND), than beating Virginia 18-17 (Cuse). Models are not there to gauge the entertainment value of a game. They are there to measure the relative performances of teams. Beating the piss out Virginia is more impressive than squeaking a win out.. Its completely illogical to argue that it isnt. A top 5 team struggling vs a non top 15 team is great television, but still a poor performance. Without using scoring margin it is much harder to measure relative performance.

The comments on running up the score are just ill-informed. Many models disregard "garbage time", so it is easy to limit the affect of running up the score.

Such bullocks. Scoring differential has a laundry list of subjective variables that don't correspond to anything useful beyond a soulless metric for a tiebreak. Let's look at an obvious, simple example: 2017-18 NBA season. Jazz were 19-28 on Jan 22, but they finished 29-6 and got the 5 seed. Are you trying to tell me that an 18+ win vs. the lowly Jazz in December 2017 would be worth more than a 6 point win against them in when they're clicking in April 2018? Stupid argument. See UMD v. Virginia 2022.

If you are beating a team, for example, 16-9 (+7) heading into the 4th and the other team essentially conceded, but you decide to play your starters to rattle off another 8 goals in the 4th quarter while the other team conceded defeat at the top of the quarter, and you win, say 23-12, (+11), does it make a frigging difference? Are you going to have some data nerds go through the context of each game? What the heck

[coda]

Appreciate the good comments about scoring margin incentivizing teams to run it up, but, to me scoring margin matters alot. Seperating teams via RPI , how many 'good wins', etc is helpful, but, not every win - or loss - is the same. 5 or 10 goal margins tell us something more so than an OT game - win or lose. Realize any single game can be an outlier, so i dont' believe in highlighting single games to prove a point. But, just as there is a difference in most people 's mind between a win vs. 11-20 , vs. a win vs. Top 5, so too there should be a difference between winning by 1 or 2, or by 5-10.

Honestly this is just a bad take. You're trying to tell me that a 10-4 Cuse win over Duke in March is better than an 18-17 thriller OT win in April vs. Virginia? [insert any equivalency and team here]. I'm sorry that makes no sense and it's obvious that "analysis" is entirely subjective from person-to-person and impossible to quantify based on numerous variables. If anything it makes the current problem even more convoluted and FUBAR. How on earth is "scoring margin" going to streamline this?

Yes, it is.. Beating Virginia is 18-9 is more impressive (ND), than beating Virginia 18-17 (Cuse). Models are not there to gauge the entertainment value of a game. They are there to measure the relative performances of teams. Beating the piss out Virginia is more impressive than squeaking a win out.. Its completely illogical to argue that it isnt. A top 5 team struggling vs a non top 15 team is great television, but still a poor performance. Without using scoring margin it is much harder to measure relative performance.

The comments on running up the score are just ill-informed. Many models disregard "garbage time", so it is easy to limit the affect of running up the score.

Such bullocks. Scoring differential has a laundry list of subjective variables that don't correspond to anything useful beyond a soulless metric for a tiebreak. Let's look at an obvious, simple example: 2017-18 NBA season. Jazz were 19-28 on Jan 22, but they finished 29-6 and got the 5 seed. Are you trying to tell me that an 18+ win vs. the lowly Jazz in December 2017 would be worth more than a 6 point win against them in when they're clicking in April 2018? Stupid argument.

I am not sure that that is supposed to even mean, just nonsensical rambling. It is pretty simple. You know only have the following information and you have to bet on the game between Team A and Team C..

Team A beat Team B by 9 goals
Team C beat beat Team B by 1 goal.

base on that limited information I can not imagine a person saying "put my money on Team C". Perhaps that would be you, if it is I would love to bet with you.

[Hooz123]

Appreciate the good comments about scoring margin incentivizing teams to run it up, but, to me scoring margin matters alot. Seperating teams via RPI , how many 'good wins', etc is helpful, but, not every win - or loss - is the same. 5 or 10 goal margins tell us something more so than an OT game - win or lose. Realize any single game can be an outlier, so i dont' believe in highlighting single games to prove a point. But, just as there is a difference in most people 's mind between a win vs. 11-20 , vs. a win vs. Top 5, so too there should be a difference between winning by 1 or 2, or by 5-10.

Honestly this is just a bad take. You're trying to tell me that a 10-4 Cuse win over Duke in March is better than an 18-17 thriller OT win in April vs. Virginia? [insert any equivalency and team here]. I'm sorry that makes no sense and it's obvious that "analysis" is entirely subjective from person-to-person and impossible to quantify based on numerous variables. If anything it makes the current problem even more convoluted and FUBAR. How on earth is "scoring margin" going to streamline this?

Yes, it is.. Beating Virginia is 18-9 is more impressive (ND), than beating Virginia 18-17 (Cuse). Models are not there to gauge the entertainment value of a game. They are there to measure the relative performances of teams. Beating the piss out Virginia is more impressive than squeaking a win out.. Its completely illogical to argue that it isnt. A top 5 team struggling vs a non top 15 team is great television, but still a poor performance. Without using scoring margin it is much harder to measure relative performance.

The comments on running up the score are just ill-informed. Many models disregard "garbage time", so it is easy to limit the affect of running up the score.

Such bullocks. Scoring differential has a laundry list of subjective variables that don't correspond to anything useful beyond a soulless metric for a tiebreak. Let's look at an obvious, simple example: 2017-18 NBA season. Jazz were 19-28 on Jan 22, but they finished 29-6 and got the 5 seed. Are you trying to tell me that an 18+ win vs. the lowly Jazz in December 2017 would be worth more than a 6 point win against them in when they're clicking in April 2018? Stupid argument.

I am not sure that that is supposed to even mean, just nonsensical rambling. It is pretty simple. You know only have the following information and you have to bet on the game between Team A and Team C..

Team A beat Team B by 9 goals
Team C beat beat Team B by 1 goal.

base on that limited information I can not imagine a person saying "put my money on Team C". Perhaps that would be you, if it is I would love to bet with you.

Oh wait wait, what if the top 2 players on Team B missed the game vs. Team A in February, but they were healed up and played against Team C in March? Stupid argument. Durrrr Like I said. So, the win against an injury riddled Team B 3 mos ago +9 is worth more than the +1 goal win against a fully healthy team now? Take a seat bro, L take.

Yea, let's do that ^,...that certainly won't create more anger and debate amongst the fans. I thought we were trying to figure out a way to streamline this not making it more f&cked up?

[coda]

Appreciate the good comments about scoring margin incentivizing teams to run it up, but, to me scoring margin matters alot. Seperating teams via RPI , how many 'good wins', etc is helpful, but, not every win - or loss - is the same. 5 or 10 goal margins tell us something more so than an OT game - win or lose. Realize any single game can be an outlier, so i dont' believe in highlighting single games to prove a point. But, just as there is a difference in most people 's mind between a win vs. 11-20 , vs. a win vs. Top 5, so too there should be a difference between winning by 1 or 2, or by 5-10.

Honestly this is just a bad take. You're trying to tell me that a 10-4 Cuse win over Duke in March is better than an 18-17 thriller OT win in April vs. Virginia? [insert any equivalency and team here]. I'm sorry that makes no sense and it's obvious that "analysis" is entirely subjective from person-to-person and impossible to quantify based on numerous variables. If anything it makes the current problem even more convoluted and FUBAR. How on earth is "scoring margin" going to streamline this?

Yes, it is.. Beating Virginia is 18-9 is more impressive (ND), than beating Virginia 18-17 (Cuse). Models are not there to gauge the entertainment value of a game. They are there to measure the relative performances of teams. Beating the piss out Virginia is more impressive than squeaking a win out.. Its completely illogical to argue that it isnt. A top 5 team struggling vs a non top 15 team is great television, but still a poor performance. Without using scoring margin it is much harder to measure relative performance.

The comments on running up the score are just ill-informed. Many models disregard "garbage time", so it is easy to limit the affect of running up the score.

Such bullocks. Scoring differential has a laundry list of subjective variables that don't correspond to anything useful beyond a soulless metric for a tiebreak. Let's look at an obvious, simple example: 2017-18 NBA season. Jazz were 19-28 on Jan 22, but they finished 29-6 and got the 5 seed. Are you trying to tell me that an 18+ win vs. the lowly Jazz in December 2017 would be worth more than a 6 point win against them in when they're clicking in April 2018? Stupid argument.

I am not sure that that is supposed to even mean, just nonsensical rambling. It is pretty simple. You know only have the following information and you have to bet on the game between Team A and Team C..

Team A beat Team B by 9 goals
Team C beat beat Team B by 1 goal.

base on that limited information I can not imagine a person saying "put my money on Team C". Perhaps that would be you, if it is I would love to bet with you.

Oh wait wait, what if the top 2 players on Team B missed the game vs. Team A in February, but they were healed up and played against Team C in March? Stupid argument. Durrrr Like I said. So, the win against an injury riddled Team B 3 mos ago is worth more than the +1 goal win against a fully healthy team now? Take a seat bro, L take.

Its the only information. I said that, because I knew someone would try to come up with scenarios to avoid the logical answer. Using RPI you would just have Team A beat team Team B and Team C beat team B.. That would be the end of the information, good luck making a decision that way.

by the way, your scenario would not be captured with RPI so I am not sure why you would even bring that up. Just a straw man argument.