LaxPower Back-Up Stick

Hoxwurth wrote: ↑Mon Mar 14, 2022 3:42 pm

Gobigred wrote: ↑Mon Mar 14, 2022 3:07 pm

laxreference wrote: ↑Mon Mar 14, 2022 10:48 am

Gobigred wrote: ↑Mon Mar 14, 2022 9:35 am Princeton's lost only to Maryland and has beaten Georgetown. Should be ranked ahead of the Hoyas.

Fair. Georgetown's resume is strong though. 6th nationally in the SOR ratings, although it is behind where Princeton sits in 4th. Princeton ends up where they are because they are still working their way up the Elo ratings. At 8th, they are several spots behind Georgetown (2nd). When the model averages it all out, Hoyas come out ahead. We'll see if that holds true over the coming weeks. If they both keep winning, the Tigers will eventually eclipse the Hoyas because their SOS from here on out is so much better.

Modify your ELO algorithm to reflect reality on the field. Results matter. Else no one but you will care about ELO.

The ELO calculation is likely good, but any model that uses prior games will struggle where so many teams like the Ivies missed so much time. One could quibble that the assumptions regarding the ELO start are wrong where Virginia (winner of the last two natties) was ranked behind Georgetown and Cornell to start the season. COVID missed games aside, changing the model could be effected by making different assumptions such as weighting later (i.e., tournament) games more heavily.

I remember reading that the ELO model does a pretty good job predicting future games, so tinkering may not make the model more accurate.

There are probably ways to tinker with Elo and make the predictions more accurate, but I've been happy with how well-calibrated it has been over the years. When Elo says 80% chance; it's been pretty accurate in that 2 out of 10 times, the underdog will win. As it should be. And making it more complex to match prior results more accurately is no guarantee that it would be more accurate going forward too. Overfitting is just as bad.

Also, the problem with adding to any model is that you make it more complex and complexity means it's harder to interpret, which is a definite negative. If there were issues where Elo was getting too many games wrong, then I'd consider it, but for the time being, it performs well enough that the downside of adding complexity is too high.

The Ivies did present a challenge. But I would argue that the objective Elo model, which ended up having the Ivies much higher-ranked than most rankings, actually was a lot closer to the mark. People assumed that the Ivies would be down because of the layoff, but that was without evidence; my model, for lack of a better option, assumed they'd be just as good as they were in 2020.

laxreference wrote: ↑Mon Mar 14, 2022 8:11 pm There are probably ways to tinker with Elo and make the predictions more accurate, but I've been happy with how well-calibrated it has been over the years. When Elo says 80% chance; it's been pretty accurate in that 2 out of 10 times, the underdog will win. As it should be. And making it more complex to match prior results more accurately is no guarantee that it would be more accurate going forward too. Overfitting is just as bad.

Also, the problem with adding to any model is that you make it more complex and complexity means it's harder to interpret, which is a definite negative. If there were issues where Elo was getting too many games wrong, then I'd consider it, but for the time being, it performs well enough that the downside of adding complexity is too high.

Well said.

I am not surprised to see the Ivies continue to perform well after 2020. Too much player improvement is attributed to game time instead of training and physical growth.

Hoxwurth wrote: ↑Tue Mar 15, 2022 9:08 am
I am not surprised to see the Ivies continue to perform well after 2020. Too much player improvement is attributed to game time instead of training and physical growth.

And so much of the prognostication about the Ivies were about losing Sowers and Teat. 1 player does not make a team. Those coaches recruited and developed those players into stars; turns out they have been able to do that again.

laxreference wrote: ↑Tue Mar 15, 2022 9:15 am

Hoxwurth wrote: ↑Tue Mar 15, 2022 9:08 am
I am not surprised to see the Ivies continue to perform well after 2020. Too much player improvement is attributed to game time instead of training and physical growth.

And so much of the prognostication about the Ivies were about losing Sowers and Teat. 1 player does not make a team. Those coaches recruited and developed those players into stars; turns out they have been able to do that again.

Lets go easy with the Ivies they are 5 and 4 against teams currently in the top 20 and that's including a win over Michigan who played the softest schedule in D1 so far. And Penn lucky to get wins over Penn State and Villanova not ranked. To me Princeton is the real deal the rest I have to wait and see.

lorin wrote: ↑Tue Mar 15, 2022 9:50 am Lets go easy with the Ivies they are 5 and 4 against teams currently in the top 20 and that's including a win over Michigan who played the softest schedule in D1 so far. And Penn lucky to get wins over Penn State and Villanova not ranked. To me Princeton is the real deal the rest I have to wait and see.

I'm always a fan of wait-and-see. Goes for any team, especially a team like Yale with a start that maybe is less impressive than their Elo rating suggests. But the top teams in the conference have banked the wins, and they are going to be there in their RPI come Selection Sunday. The gap between the Ivy top-3 and the ACC/Big Ten top-3 is smaller than the gap when you look top-to-bottom. So it does depend on how you are evaluating them.

But, in my most recent simulation runs, here's the probability that each of the Ivies ended up making the tournament:

Cornell: 97.2%
Princeton: 95.0%
Yale: 81.5%
Penn: 73.4%
Brown: 57.5%
Harvard: 2.5%
Dartmouth: 0.5%

lorin wrote: ↑Tue Mar 15, 2022 9:50 am

laxreference wrote: ↑Tue Mar 15, 2022 9:15 am

Hoxwurth wrote: ↑Tue Mar 15, 2022 9:08 am
I am not surprised to see the Ivies continue to perform well after 2020. Too much player improvement is attributed to game time instead of training and physical growth.

And so much of the prognostication about the Ivies were about losing Sowers and Teat. 1 player does not make a team. Those coaches recruited and developed those players into stars; turns out they have been able to do that again.

Lets go easy with the Ivies they are 5 and 4 against teams currently in the top 20 and that's including a win over Michigan who played the softest schedule in D1 so far. And Penn lucky to get wins over Penn State and Villanova not ranked. To me Princeton is the real deal the rest I have to wait and see.

6 of 7 ivies are now 1, 3, 9, 12, 13 and 17 rpi. rpi is now just starting to get meaningful. just as they head into ivy and in-conference season for many.

that looks like an acc distribution. and the lone outlier is 39, not the wasteland it's been in the past. if the teams they've played to date mostly hold up, other conferences may be scraping for spots.

laxreference wrote: ↑Tue Mar 15, 2022 10:10 am

lorin wrote: ↑Tue Mar 15, 2022 9:50 am Lets go easy with the Ivies they are 5 and 4 against teams currently in the top 20 and that's including a win over Michigan who played the softest schedule in D1 so far. And Penn lucky to get wins over Penn State and Villanova not ranked. To me Princeton is the real deal the rest I have to wait and see.

I'm always a fan of wait-and-see. Goes for any team, especially a team like Yale with a start that maybe is less impressive than their Elo rating suggests. But the top teams in the conference have banked the wins, and they are going to be there in their RPI come Selection Sunday. The gap between the Ivy top-3 and the ACC/Big Ten top-3 is smaller than the gap when you look top-to-bottom. So it does depend on how you are evaluating them.

But, in my most recent simulation runs, here's the probability that each of the Ivies ended up making the tournament:

Cornell: 97.2%
Princeton: 95.0%
Yale: 81.5%
Penn: 73.4%
Brown: 57.5%
Harvard: 2.5%
Dartmouth: 0.5%

When you say tournament, I assume you mean the NCAA and not the ILT? Right?
GOBIGRED
Joewillie78

joewillie78 wrote: ↑Tue Mar 15, 2022 10:18 am

laxreference wrote: ↑Tue Mar 15, 2022 10:10 am

lorin wrote: ↑Tue Mar 15, 2022 9:50 am Lets go easy with the Ivies they are 5 and 4 against teams currently in the top 20 and that's including a win over Michigan who played the softest schedule in D1 so far. And Penn lucky to get wins over Penn State and Villanova not ranked. To me Princeton is the real deal the rest I have to wait and see.

I'm always a fan of wait-and-see. Goes for any team, especially a team like Yale with a start that maybe is less impressive than their Elo rating suggests. But the top teams in the conference have banked the wins, and they are going to be there in their RPI come Selection Sunday. The gap between the Ivy top-3 and the ACC/Big Ten top-3 is smaller than the gap when you look top-to-bottom. So it does depend on how you are evaluating them.

But, in my most recent simulation runs, here's the probability that each of the Ivies ended up making the tournament:

Cornell: 97.2%
Princeton: 95.0%
Yale: 81.5%
Penn: 73.4%
Brown: 57.5%
Harvard: 2.5%
Dartmouth: 0.5%

When you say tournament, I assume you mean the NCAA and not the ILT? Right?
GOBIGRED
Joewillie78

Correct

joewillie78 wrote: ↑Tue Mar 15, 2022 10:18 am

laxreference wrote: ↑Tue Mar 15, 2022 10:10 am

lorin wrote: ↑Tue Mar 15, 2022 9:50 am Lets go easy with the Ivies they are 5 and 4 against teams currently in the top 20 and that's including a win over Michigan who played the softest schedule in D1 so far. And Penn lucky to get wins over Penn State and Villanova not ranked. To me Princeton is the real deal the rest I have to wait and see.

I'm always a fan of wait-and-see. Goes for any team, especially a team like Yale with a start that maybe is less impressive than their Elo rating suggests. But the top teams in the conference have banked the wins, and they are going to be there in their RPI come Selection Sunday. The gap between the Ivy top-3 and the ACC/Big Ten top-3 is smaller than the gap when you look top-to-bottom. So it does depend on how you are evaluating them.

But, in my most recent simulation runs, here's the probability that each of the Ivies ended up making the tournament:

Cornell: 97.2%
Princeton: 95.0%
Yale: 81.5%
Penn: 73.4%
Brown: 57.5%
Harvard: 2.5%
Dartmouth: 0.5%

When you say tournament, I assume you mean the NCAA and not the ILT? Right?
GOBIGRED
Joewillie78

Yes I should of had Cornell up top

Here's a piece I wrote for USA Lax Magazine that ran through the 5 conferences that still earned at least one at-large bid when I went through the simulations after last weekend's games.

If you are curious how the model spit out a (seemingly low) 86% chance that UVA makes the NCAA tournament, the Cavs are the "unlocked" team on the PRO site right now. Just click on the Projections tab.

lorin wrote: ↑Tue Mar 15, 2022 9:50 am

laxreference wrote: ↑Tue Mar 15, 2022 9:15 am

Hoxwurth wrote: ↑Tue Mar 15, 2022 9:08 am
I am not surprised to see the Ivies continue to perform well after 2020. Too much player improvement is attributed to game time instead of training and physical growth.

And so much of the prognostication about the Ivies were about losing Sowers and Teat. 1 player does not make a team. Those coaches recruited and developed those players into stars; turns out they have been able to do that again.

Lets go easy with the Ivies they are 5 and 4 against teams currently in the top 20 and that's including a win over Michigan who played the softest schedule in D1 so far. And Penn lucky to get wins over Penn State and Villanova not ranked. To me Princeton is the real deal the rest I have to wait and see.

FWIW, ACC is 5-12 against Top 20 opponents this year!!! But I largely agree that Yale/Brown/Harvard have a lot left to prove. I'd say same for Penn on the eye test, but at the end of the day when teams are selected for the NCAA's a W is a W, lucky or not.

RopeUnit wrote: ↑Thu Mar 17, 2022 10:39 am

lorin wrote: ↑Tue Mar 15, 2022 9:50 am

laxreference wrote: ↑Tue Mar 15, 2022 9:15 am

Hoxwurth wrote: ↑Tue Mar 15, 2022 9:08 am
I am not surprised to see the Ivies continue to perform well after 2020. Too much player improvement is attributed to game time instead of training and physical growth.

And so much of the prognostication about the Ivies were about losing Sowers and Teat. 1 player does not make a team. Those coaches recruited and developed those players into stars; turns out they have been able to do that again.

Lets go easy with the Ivies they are 5 and 4 against teams currently in the top 20 and that's including a win over Michigan who played the softest schedule in D1 so far. And Penn lucky to get wins over Penn State and Villanova not ranked. To me Princeton is the real deal the rest I have to wait and see.

FWIW, ACC is 5-12 against Top 20 opponents this year!!! But I largely agree that Yale/Brown/Harvard have a lot left to prove. I'd say same for Penn on the eye test, but at the end of the day when teams are selected for the NCAA's a W is a W, lucky or not.

this is the lacrosse analytics thread, so the acc is actually 9 and 9 against top 20. and 8 and 7 when it's out of conference only.

i agree the ivy is at present top dog. and b1g and acc little separation. things are getting late early for the acc, or specifically duke and especially cuse and nd.

As a result of the Navy win over Hopkins, here are how at-large chances changed over night:

BU: +3.0%
BRWN: +2.4%
YALE: +2.2%
PENN: +2.2%
DUKE: +1.1%
HARV: +1.0%
OSU: +1.0%
RU: +0.7%
UVA: +0.6%
COR: +0.3%
GTWN: +0.2%
ND: +0.1%

UNC: -0.8%
MD: -0.9%
PRIN: -1.8%
JHU: -12.6%

Here are the Live Win Probability Links for today's games.

- Air Force vs Mercer
- Utah vs Georgetown
- Jacksonville vs UMass-Lowell
- Yale vs Cornell
- Lehigh vs Army
- Towson vs Duke
- Lafayette vs Colgate
- Michigan vs Notre Dame
- Saint Joseph's vs Mount St Marys
- Merrimack vs Bryant
- Marquette vs Robert Morris
- Hobart vs Sacred Heart
- Providence vs Fairfield
- Quinnipiac vs Manhattan
- Cleveland State vs Hampton
- Wagner vs LIU
- NJIT vs Vermont
- Bellarmine vs VMI
- Boston U vs Holy Cross
- St. John's vs Dartmouth
- Penn vs Princeton
- Bucknell vs Loyola
- Brown vs Harvard
- Rutgers vs Hofstra
- Binghamton vs Albany
- St. Bonaventure vs Marist
- Maryland vs Virginia
- Syracuse vs Stony Brook
- Drexel vs Villanova

Here are the Live Win Probability Links for the 5 games today

- Denver vs Ohio State
- North Carolina vs High Point
- Siena vs Canisius
- UMBC vs Richmond
- Johns Hopkins vs Delaware

Here's my Top-20 for this week...

As always, here is the top individual EGA performances of the past week. The full post is here, but here are the highlights...

Brennan O'Neill (DUKE) and Duke had the kind of 2nd half that makes you wonder if they are putting it together a bit. They finished the Towson game on a 12-3 run that was one of their better halves of the year. Time will tell if they can carry that into ACC play.


Here is how the rest of the top 10 shook out:

Tucker Dordevic (SYR) - 7.71 EGA
Luke Wierman (MD) - 7.18 EGA
Chris Brown (PRIN) - 6.25 EGA
Sam Handley (PENN) - 6.35 EGA
Matthew Gunty (BRWN) - 6.70 EGA
Nicholas Ramsey (YALE) - 6.39 EGA
Zach Cole (SJOE) - 6.05 EGA
Brenden Crouse (AFA) - 5.81 EGA
Liam McDonough (UML) - 5.93 EGA

laxref,
does your rpi calculation include games/opponents that are on the schedule for the future, or just the ones played to date?

wgdsr wrote: ↑Fri Mar 25, 2022 1:02 pm laxref,
does your rpi calculation include games/opponents that are on the schedule for the future, or just the ones played to date?

Just those played to date.

thanks!