FanLax Forum Poll

[Chousnake]

3-2 Yale two slots ahead of 5-1 Cornell after Saturday's beatdown is hilarious.

Yup. Yale has worse losses than Cornell but their better wins, puts them ahead of Cornell. Win percentage is a dull tool to evaluate teams. But, if you like to look at Win Percentages... Yale's 60% Win% is weaker than Cornell's 83% but Yale's current SOS (Strength of Schedule) is #1 while Cornell's is #15. SOS matters. Strength of wins matters.

Sometimes your algorithm spits out a result so ludicrous that it invalidates your entire algorithm. This is one of those times.

Back to the drawing board, I guess.

EDIT: You conveniently sidestep the elephant in the room (and the main point of the poster that you were responding to) that it isn't just about win percentage; it's the fact that Cornell went up to Yale and completely dismantled them 2 days ago. They won by 10 and it wasn't even that close. It was an annihilation.

This! My biggest issue with computer lax rankings, and some NCAA selections in the past, is the complete dismissal or de-emphasis of head to head results. Isn't that ultimately what decides everything? All post season tournaments in every sport are based on head to head wins. If not, we would be awarding titles based on algorithms. The FanLax computer is no better than HAL if it rates Yale over Cornell after Cornell doubled up Yale in Yale 2 days ago.

[Matnum PI]

You conveniently sidestep the elephant in the room... an annihilation.

This! My biggest issue with computer lax rankings, and some NCAA selections in the past, is the complete dismissal or de-emphasis of head to head results...

Head-to-head doesn't work. Superficially, it makes enormous sense. But in reality, it's just one game. A team's entire body of work is more meaningful than one game. Even if that game is head-to-head. If rankings were based on head-to-head, then Jacksonville should be ahead of Duke... and then JHU should be ahead of Jacksonville..,. and then UNC should be ahead of JHU and then etc. It never stops. Head-to-head doesn't work. Body of work works.

[HGK]

I’d say one of 13 or 14 games matters.

[ICGrad]

You conveniently sidestep the elephant in the room... an annihilation.

One game, head-to-head. That and only that. A team's entire body of work is more meaningful than one game. Even if that game is head-to-head.

Please. Stop being disingenuous. We're talking 1 game out of what, six games that they've played? And a game that took place 2 days ago. And Cornell has one fewer loss, the huge head-to-head win, and the game wasn't even close. And it was at Yale.

Sure, if we were looking back after 13 games, and way back in week 6 Cornell beat Yale but Yale had a better record, maybe. But come on; this is a weak response.

And why is Villanova ranked higher than Yale in your rankings? Yales (and Cornell) both have significantly higher SoS and Yale beat Villanova.

Again, back to the drawing board. Your crack is whack...

[rolldodge]

Here's the method with Massey rankings:

Top 5 win = 4 points
Top 10 win = 3 points
Top 20 win = 2 points
Top 30 win = 1 point
Top 40 win = 0.5 points
All other wins = 0.25 point

Top 10 loss = -0.5 points
Top 30 loss = -1 point
All other losses = -2 points

1.( 10.75 )Notre Dame 6-0 Marquette (0.5), Cleveland State (0.25), Georgetown (3), Maryland (4), Ohio State (2), Michigan (1)

2.(10) Duke 7-1 Bellarmine (0.5), High Point (0.5), Denver (2), Penn (2), Syracuse (1), Loyola (2), UNC (3) (Jacksonville -1)

3.(7.75) Virginia 6-1 Michigan (1), Harvard (1), Ohio State (2), Richmond (1), Hopkins(3), Towson (0.25) (Maryland -0.5)

4.(6.75) Maryland 5-2 Richmond (1), Syracuse (1), Princeton (2), Albany (0.25), Virginia (4) (Loyola -1, Notre Dame -0.5)

4.(6.75) Hopkins 7-3 Georgetown(3), Jacksonville (2), Utah (0.5), St Josephs (1), Syracuse (1), Delaware (1), Navy (0.25) (North Carolina -0.5, Loyola -1, UVA -0.5)

6.(6.25) Loyola 5-2 Maryland (4), Hopkins (3), Towson (0.25), Lafayette (0.25), Bucknell (0.25) (Rutgers -1, Duke -0.5)

7.(6) Penn State 5-2 Lafayette (0.25), Stony Brook(0.25), Yale (2), Penn (2), Cornell (4) (Villanova -0.5, Marquette -2)

8.(5.25) Rutgers 7-1 Marist (0.25), Stony Brook (0.25), St John’s (0.25), Loyola (2), Utah (0.5), Princeton (2), UMass (1) (Army -1)

9.(5) Cornell 5-1 Albany (0.25), Lehigh (1), Hobart (0.5), Ohio State (2), Yale (2) (Penn State -0.5)

9.(5) Villanova 6-1 Penn State (2), Delaware (1), Hofstra (0.25), Drexel (0.5), Penn (2), Monmouth (0.25) (Yale -1)

11.(4.75) Jacksonville 6-1 Marist (0.25), Duke (4), High Point (0.5), UMass Lowell (0.25), St Johns (0.25) (Hopkins -0.5)

12.(4.5) Yale 3-2 Villanova (3), UMass (1), Denver (2) (Penn State -1, Cornell -0.5)

13.(4) Army. 6-1 Rutgers (3), Wagner (0.25), Mercer (0.25), Holy Cross (0.25), Lafayette (0.25), Lehigh (1) (Umass -1)

13.(4) Penn 3-3 Georgetown (3), St Joe’s (1), Princeton (2) (Duke -0.5, Penn State -1, Villanova -0.5)

15.(3.25) Denver 4-3 Utah (0.5), Merrimack (0.25), UNC (3), Ohio State (2) (Air Force -1, Duke -0.5, Yale -1)

15.(3.25) North Carolina 5-3 Mercer (0.25), Hopkins (3), Syracuse (1), Brown (1), Dartmouth (0.5) (Ohio State -1, Denver -1, Duke -0.5)

17.(2) Ohio State 4-4 Air Force (1), Cleveland State (0.25), UNC (3), Detroit (0.25) (UVA -0.5, Cornell -0.5, ND -0.5, Denver -1)

18.(1.5) Georgetown 3-3 Princeton (2), Richmond (1). High Point (0.5) (Hopkins -0.5, Penn -1, ND -0.5)

19.(0.75) Dartmouth 4-1 Merrimack (0.25), Holy Cross (0.25), Sienna (0.25), Hobart (0.5) (UNC -0.5)

20.(0.5) UMass 4-3 Umass Lowell (0.25), Army (2), Vermont (0.5), Albany (0.25) (Boston U -1, Yale -1, Rutgers -0.5 )

21.(0.25) Air Force 5-2 Denver (2), Colgate (0.25), Mercer (0.25), Bryant (0.5), Queens (0.25) (Ohio State -1, Stony Brook -2)

21.(0.25) Harvard 3-2 Bucknell (0.25), Vermont (0.5), Brown (1) (Virginia -0.5, Michigan -1)

23.(0) Marquette 4-3 Lindwood (0.25), Michigan (1), Detroit (0.25), Penn State (2) (Utah -1, ND -0.5, Bellarmine -2)

23.(0) Boston U 5-1 Bryant (0.5), UMass (1), Colgate (0.25), Bucknell (0.25), Holy Cross (0.25) (Vermont -2)

23.(0) Lehigh 5-2 Fairfield (0.25), Hobart (0.5), Navy (0.25), Bing (0.25), HC (0.25) (Cornell -0.5, Army -1)

[Matnum PI]

...We're talking 1 game out of what, six games...

And why is Villanova ranked higher than Yale in your rankings? Yales (and Cornell) both have significantly higher SoS and Yale beat Villanova...

But why take this one game? Why take Cornell beat Yale? Why not take Marquette beat PSU? As a rule, if we're going to be fair, we can't pick and choose the individual games we want to emphasize. Yesterday or the day before or otherwise, every game counts.

Villanova is ranked higher than Yale because their losses are very similar but... They have better wins. PSU, Penn, and Delaware while Yale has V'nova, Denver, and UMass. Body of work, not one game. Consistent. Not disingenuous.

[CU77]

Why take Cornell beat Yale? Why not take Marquette beat PSU?

IMO, closeness of each game matters. Ignoring that data almost certainly yields suboptimal results. Cornell blew Yale off the field, Marquette v PSU went down to the wire.

Timing also matters. Does Maryland's loss to Loyola matter as much now?

This is all assuming that what is wanted is as accurate a ranking as possible of where teams sit today.

[Matnum PI]

IMO, closeness of each game matters. Ignoring that data almost certainly yields suboptimal results. Cornell blew Yale off the field, Marquette v PSU went down to the wire.

I disagree. Understand your perspective and disagree. Scores to me are irrelevant. Just Ws and Ls. The goal is to win the race. How much you win by, especially during a long season, is irrelevant.

Timing also matters. Does Maryland's loss to Loyola matter as much now?

For me, yes. 100% yes. Every regular season game counts equally. Predictively, trying to predict the future, I'd agree with you. But in terms of trying to provide a fair, merit-based system, every game counts equally. Come play-offs, I feel differently. Play-offs, by nature, are a very different format. But, for me, in the regular season, every game counts equally.

[Gobigred]

3-2 Yale two slots ahead of 5-1 Cornell after Saturday's beatdown is hilarious.

Yup. Yale has worse losses than Cornell but their better wins, puts them ahead of Cornell. Win percentage is a dull tool to evaluate teams. But, if you like to look at Win Percentages... Yale's 60% Win% is weaker than Cornell's 83% but Yale's current SOS (Strength of Schedule) is #1 while Cornell's is #15. SOS matters. Strength of wins matters.

20-10 matters, too. Your algorithm sucks.

[joewillie78]

3-2 Yale two slots ahead of 5-1 Cornell after Saturday's beatdown is hilarious.

Yup. Yale has worse losses than Cornell but their better wins, puts them ahead of Cornell. Win percentage is a dull tool to evaluate teams. But, if you like to look at Win Percentages... Yale's 60% Win% is weaker than Cornell's 83% but Yale's current SOS (Strength of Schedule) is #1 while Cornell's is #15. SOS matters. Strength of wins matters.

20-10 matters, too. Your algorithm sucks.

This gives me bad memories of a few years back when Cornell defeated Notre Dame AT NOTRE DAME, and when the brackets came out, ND was in and Cornell out, with very similar resumes. You would have thought that a head to head win on the ROAD , when 2 teams are very similar would tilt to the winning team.
Yes, Cornell demolished Yale but the Cornell/ND game if memory serves me correct was a 2 goal Cornell win.
Of course, if it came down to a Yale/Cornell 1 in 1 out scenario and the resumes are similar, one would hope that this emphatic Cornell win would be the difference, however given what happened a few years ago, nothing this committee does shocks me.
That is why I always hope and pray, they win the ILT and take it out of the hands of the committee.
Gobigred
Joewillie78

[Matnum PI]

... but the Cornell/ND game if memory serves me correct was a 2 goal Cornell win... however given what happened a few years ago...

And if this was a play-off game, I'd agree with you. But it wasn't. During the regular season, teams should be judged on their body of work during the 2023 regular season, the entirety of their 2023 regular season.

[ICGrad]

... but the Cornell/ND game if memory serves me correct was a 2 goal Cornell win... however given what happened a few years ago...

And if this was a play-off game, I'd agree with you. But it wasn't. During the regular season, teams should be judged on their body of work during the 2023 regular season, the entirety of their 2023 regular season.

You keep saying Body of Work as if Yale is 12-1 and Cornell is 5-6 and just happened to beat Yale. So let's take a look at the body of work:

Cornell: 5-1
Yale: 3-2

Both teams have lost to PSU. Cornell kicked the snot out of Yale. They have comparable SoS.

Cornell has the considerably better body of work. Either your ratings are deeply flawed, or the sample size is too small to be meaningful. My guess is the former, with a sprinkling of the later thrown in.

[LaxDownUnder]

Villanova is ranked higher than Yale because their losses are very similar but... They have better wins. PSU, Penn, and Delaware while Yale has V'nova, Denver, and UMass. Body of work, not one game. Consistent. Not disingenuous.

Do they though? Using the standings of these three wins for Vilanova (6, 12, and 27 on the Fanlax Computer Rankings) shouldn't their wins be considered below Yale's (4, 17, and 20)? According to the description of the "Win Ranking" category on the site, the smaller the sum of these wins, the better they are. Vilanova's 6, 12, and 27 add up to 45 while Yale's 4, 17, and 20 add up to 41. How then are their wins any better than Yale's?

[Matnum PI]

Vilanova's 6, 12, and 27 add up to 45 while Yale's 4, 17, and 20 add up to 41. How then are their wins any better than Yale's?

The rankings are based off of the FanLax Forum Rankings. (You need an "anchor ranking" or the rankings are like a dog chasing it's tail. Plus, one can argue with another's rankings. One cannot argue with their own rankings.) So V: 8+18+21=47 and Y: 7+15+29=51. V's win ranking is better.

[Matnum PI]

... So let's take a look at the body of work:
Cornell: 5-1
Yale: 3-2
Both teams have lost to PSU. Cornell kicked the snot out of Yale. They have comparable SoS.
Cornell has the considerably better body of work. Either your ratings are deeply flawed, or the sample size is too small to be meaningful. My guess is the former, with a sprinkling of the later thrown in.

You're looking at the body of work with a very broad brushstroke. i.e. You're speaking to a general SOS while I'm speaking to the specific strength of each opponent. 5 and 6 games into the season is very much enough games for a team to prove themselves with this system. (P'ton might disagree but I would argue that the fault of their current poor ranking isn't the few games they've played as much as their inability to win against better-ranked teams.) And, based on these 5-6 games, Cornell has beaten Yale, OSU, and Lehigh and lost to PSU while Yale has beaten V'nova, Denver, and UMass and lost to Cornell and PSU. Based off of the FanLax Forum Poll, Yale's losses are worse but... Yale's wins are better and these wins more than off-set the losses.

[Typical Lax Dad]

... So let's take a look at the body of work:
Cornell: 5-1
Yale: 3-2
Both teams have lost to PSU. Cornell kicked the snot out of Yale. They have comparable SoS.
Cornell has the considerably better body of work. Either your ratings are deeply flawed, or the sample size is too small to be meaningful. My guess is the former, with a sprinkling of the later thrown in.

You're looking at the body of work with a very broad brushstroke. i.e. You're speaking to a general SOS while I'm speaking to the specific strength of each opponent. 5 and 6 games into the season is very much enough games for a team to prove themselves with this system. (P'ton might disagree but I would argue that the fault of their current poor ranking isn't the few games they've played as much as their inability to win against better-ranked teams.) And, based on these 5-6 games, Cornell has beaten Yale, OSU, and Lehigh and lost to PSU while Yale has beaten V'nova, Denver, and UMass and lost to Cornell and PSU. Based off of the FanLax Forum Poll, Yale's losses are worse but... Yale's wins are better and these wins more than off-set the losses.

Yep. I made this point about a team last season…. But some may argue that if they manage to beat Cornell twice then they should play for the National Title.

[LaxDownUnder]

Vilanova's 6, 12, and 27 add up to 45 while Yale's 4, 17, and 20 add up to 41. How then are their wins any better than Yale's?

The rankings are based off of the FanLax Forum Rankings. (You need an "anchor ranking" or the rankings are like a dog chasing it's tail.) So V: 8+18+21=47 and Y: 7+15+29=51. V's win ranking is better.

Gotcha. I thought you were working off the Computer rankings, my mistake.

I would be curious about why there is this discrepancy in the computer rankings, however. Does the computer need an "anchor" for their ratings and thats why the Win Rank category doesn't always go in ascending order based on the sum of the top 3 wins? How does the computer determine what this anchor should be?

[Matnum PI]

I would be curious about why there is this discrepancy in the computer rankings, however. Does the computer need an "anchor" for their ratings and thats why the Win Rank category doesn't always go in ascending order based on the sum of the top 3 wins?

If I'm understanding your question correctly, Yes.

... How does the computer determine what this anchor should be?

We feed the computer the anchor ranking. The season starts with the Final FanLax Computer Ranking of the previous season being used as the anchor ranking. Then, as the season progresses, the anchor ranking is composed of the Forum Ranking (for the top teams) and the Computer Ranking (for the rest of the teams).

[runrussellrun]

Princeton v. Bryant: Who are you taking?
Princeton v. Dartmouth: Who are you taking.

If the answer to either is Princeton, then they aren't overrated (according to your own metrics).

It is Princeton... if I were a predictive algorithm. I'm not programmed to predict the future. I'm programmed to evaluate the past. Based on their current resume, Bryant is "better" than Princeton. i.e. Bryant has a better resume and, for evaluative purposes, a better resume matters.

Do they, though? They have a significantly better record, sure, but against a significantly weaker schedule. I'm not sure if, one paper, their resume is better. They've just racked up wins against inferior opponents.

Anyway, you said overrated, not over-ranked. Princeton is the better than either team. Ultimately, that is a factor I consider when I enter my rankings.

there it IS again......the "weaker teams" claim. Basing this, on what, exactly?



Perhaps the computers staring point, for the past, would be better served, it it just had numbers 1 thru 18......sorry, but MOST feel that making the playoffs....IS ....the end game.

18. Manhattan
17 Robby Mo
16 Vermont
15. Georgetown
14 Harvard (beat down by 10 )
13 Ohio St
12 Boston U
11 Brown
10 St. Joes
9 Richmond
8. UVA
7 Yale
6 Penn
5 Delaware
4 Rutgers
3 Princeton
2 Cornell
1. Maryland.
.
(ranked by N$aa margin of victory, since it IS this weeks shiny new toy )

see what happens, computer, when last year, hence this year, the REAL top teams (the ones that make the playoffs, unless THAT doesn't matter

[Matnum PI]

...inability to win against better-ranked teams...

Yep. I made this point about a team last season…. But some may argue that if they manage to beat Cornell twice then they should play for the National Title.

People die and the relevant people's families decide not to have funerals. In each case, the reason why is "s/he was not kind towards the end of her/his lives". Which I find odd. This may be true but... What about the 84 years they lived before those less-than-kind 1-2 years at the end of their life? A real-life recency bias. And, also, an inability to recognize the importance of a person's entire body of work. I'm not suggesting that someone's Great Aunt, cousin, or otherwise are wrong. To each their own. But I am suggesting that how people evaluate people, more relevant for us, how people evaluate teams, varies. In the play-offs, I get it. Win, move on. Lose, go home. That I get. But during the regular season, the entirety of each team's regular season is what is being evaluated. Ws, Ls, high-points, low-points. No conjecture about how they barely lost, could've won, so it should almost count as a W. No conjecture about how they barely won, could've lost, so it should almost count as an L. Whether this is after 6 or 16 games, just Ws and Ls for every game for the entirety of the season.