LaxPower Back-Up Stick

rolldodge wrote: ↑Wed May 08, 2024 1:52 pm Apologies if this has been discussed above, but in taking margin of victory into account how do you incorporate different styles and pace of play? One can realistically imagine a team that is defense oriented and may only win their games by 2 or 3 goals but the outcome of the game is never in doubt. Do they get penalized for that?

coda wrote: ↑Wed May 08, 2024 1:44 pm

Big Dog wrote: ↑Wed May 08, 2024 1:18 pm

coda wrote: ↑Wed May 08, 2024 11:09 am

Big Dog wrote: ↑Wed May 08, 2024 11:00 am

coda wrote: ↑Wed May 08, 2024 9:43 am

NYlax222 wrote: ↑Wed May 08, 2024 9:14 am I'm not suggesting scoring margin alone solves the issue of ranking teams who ultimately don't play common schedule. However, truly am amazed by anyone who thinks that winning (or losing) by 10 is the same (i.e, provides same insight into respective rankings of teams) as winning (or losing) by one.

Imagine a world where two teams play same exact schedule. Both go 11-2. Firs team averages 8 goal margin in their win, and loses 2 in OT. Other team averages 2 goals marin in 11 wins, and gets blown out by 10 in losses. Per RPI, they are the same. Nobody would look at those two teams and conclude they are equivalent.

Folks love to call a 'bad loss' a loss (regardless of margin) to a non-Top 20 team, yet, losing by 10 isn't?

Can lacrosse do better? Yes. That is what this discussion is about. Nobody is saying there is a perfect model that will be 100% correct. Its just about improving. I just dont understand people arguing that less information is better.

The problem for lax is that there are too few games in a season to make "other information" statistically meaningful. And once league play starts, RPIs can't change much. (All your opponent's play each other.) The Net ranking in basketball (kinda) works bcos they play 30+ games.

That said, not a fan of scoring margin in such a short season.

You would obviously love more data-points, but college lacrosse season mirrors college football pretty well. I dont see a huge difference in the season in comparison. The less data-points you have the more information you need. Simpler models work better with more data pts.

And yet College football does not use margin of victory.

https://collegefootballplayoff.com/spor ... e-protocol

I dont even know how they calculate SOS. What metrics are they relying on? That we do not know. Likely that the models they look at uses pts and/or yards in their calculations, but without knowing what information they rely on it is impossible to say. More likely to use yards/play than pts per play, but that is unique to football.

Big Dog wrote: ↑Wed May 08, 2024 4:05 pm

coda wrote: ↑Wed May 08, 2024 1:44 pm

Big Dog wrote: ↑Wed May 08, 2024 1:18 pm

coda wrote: ↑Wed May 08, 2024 11:09 am

Big Dog wrote: ↑Wed May 08, 2024 11:00 am

coda wrote: ↑Wed May 08, 2024 9:43 am

NYlax222 wrote: ↑Wed May 08, 2024 9:14 am I'm not suggesting scoring margin alone solves the issue of ranking teams who ultimately don't play common schedule. However, truly am amazed by anyone who thinks that winning (or losing) by 10 is the same (i.e, provides same insight into respective rankings of teams) as winning (or losing) by one.

Imagine a world where two teams play same exact schedule. Both go 11-2. Firs team averages 8 goal margin in their win, and loses 2 in OT. Other team averages 2 goals marin in 11 wins, and gets blown out by 10 in losses. Per RPI, they are the same. Nobody would look at those two teams and conclude they are equivalent.

Folks love to call a 'bad loss' a loss (regardless of margin) to a non-Top 20 team, yet, losing by 10 isn't?

Can lacrosse do better? Yes. That is what this discussion is about. Nobody is saying there is a perfect model that will be 100% correct. Its just about improving. I just dont understand people arguing that less information is better.

The problem for lax is that there are too few games in a season to make "other information" statistically meaningful. And once league play starts, RPIs can't change much. (All your opponent's play each other.) The Net ranking in basketball (kinda) works bcos they play 30+ games.

That said, not a fan of scoring margin in such a short season.

You would obviously love more data-points, but college lacrosse season mirrors college football pretty well. I dont see a huge difference in the season in comparison. The less data-points you have the more information you need. Simpler models work better with more data pts.

And yet College football does not use margin of victory.

https://collegefootballplayoff.com/spor ... e-protocol

I dont even know how they calculate SOS. What metrics are they relying on? That we do not know. Likely that the models they look at uses pts and/or yards in their calculations, but without knowing what information they rely on it is impossible to say. More likely to use yards/play than pts per play, but that is unique to football.

pretty easy to google the formulas used, but No, it does not include margin of victory. (no bonus points for 'Bama beating on the Little Sisters of teh Poor by 70.)

For example, one that used to be popular:
Though the BCS is now defunct, its calculation for strength of schedule lives on. This system takes the sum of the team's opponent's records and multiplies by two. It then adds that number to the team's opponent's opponent's record and multiplies by one. Finally, that number is divided by three.

Big Dog wrote: ↑Wed May 08, 2024 4:05 pm
pretty easy to google the formulas used, but No, it does not include margin of victory. (no bonus points for 'Bama beating on the Little Sisters of teh Poor by 70.)

For example, one that used to be popular:
Though the BCS is now defunct, its calculation for strength of schedule lives on. This system takes the sum of the team's opponent's records and multiplies by two. It then adds that number to the team's opponent's opponent's record and multiplies by one. Finally, that number is divided by three.

Wheels wrote: ↑Wed May 08, 2024 11:13 pm

Big Dog wrote: ↑Wed May 08, 2024 4:05 pm
pretty easy to google the formulas used, but No, it does not include margin of victory. (no bonus points for 'Bama beating on the Little Sisters of teh Poor by 70.)

For example, one that used to be popular:
Though the BCS is now defunct, its calculation for strength of schedule lives on. This system takes the sum of the team's opponent's records and multiplies by two. It then adds that number to the team's opponent's opponent's record and multiplies by one. Finally, that number is divided by three.

Hence the possible multicollinearity issues with the RPI.

Finster wrote: ↑Thu May 09, 2024 9:10 am

Wheels wrote: ↑Wed May 08, 2024 11:13 pm

Big Dog wrote: ↑Wed May 08, 2024 4:05 pm
pretty easy to google the formulas used, but No, it does not include margin of victory. (no bonus points for 'Bama beating on the Little Sisters of teh Poor by 70.)

For example, one that used to be popular:
Though the BCS is now defunct, its calculation for strength of schedule lives on. This system takes the sum of the team's opponent's records and multiplies by two. It then adds that number to the team's opponent's opponent's record and multiplies by one. Finally, that number is divided by three.

Hence the possible multicollinearity issues with the RPI.

Wheels: in the laxreference bracket bets, it looks like a majority of betters have taken Princeton to take out UMD. Not me. Though I don't think Maryland will go far, I think they get by Princeton.

I like the terps' chances against any team where if your defense can shut off one player, the opponent's offense must pivot dramatically. Here, I think Zapitello can shut off Makesy, which means Princeton then must turn to their freshmen.Tough ask in their first D1 tourney. I'll go with MD by 2.

keno in reno wrote: ↑Sun May 12, 2024 10:24 pm So I'd say the selection and seeding protocol worked perfectly this year. Nobody ever wants to congratulate the system, but it worked. Maybe you can argue amongst the seeds?

wgdsr wrote: ↑Thu May 16, 2024 1:52 pm so with the acc's getting 4 in and put in different brackets this time, we have 4 acc vs. ..... given everyone's obsession with conferences and related sports radio-speak, between fanlax, lax media and lax twitter, how many heads explode if we get acc tourney redux?

don't consider it very likely, but it's still there rn. would be pretty surprised if 2 acc's don't advance, and more than that and we're going to be hearing a whole lot about acc vs the world.