Toggle sidebar

Colgate #11 NET

D

Doctor Worm

Heisman
Feb 7, 2002
30,398
22,335
113
SIAP - but this is kind of bizarre.

Colgate is playing a 14 game schedule this year. 8 home, 6 away. They are 9-1 so far, with four more to play.

Here's the crazy part. The 14 games are played against only THREE OPPONENTS. They play Holy Cross (NET 247) six times: four home and two away. They play Army (NET 96) four times: two home and two away. Same with Boston University (NET 236). Their only loss was to Army, at home. Their remaining schedule is all at home: two against HC and two against BU.

By Quadrant, their record is:
QUAD 1: 0-0
QUAD 2: 2-0
QUAD 3: 3-1
QUAD 4: 4-0.

Somehow that translates to a NET ranking of 11! Go figure. I can't.
 
  • Haha
Reactions: DHajekRC84
P

PSAL_Hoops

Heisman
Feb 18, 2008
13,350
12,657
78
I can’t even begin to guess how the math could make that work. 6 of the 10 games they’ve played have been decided by less than 12 points. It’s not like they are blowing the snot off their repeat opponents every time. Go figure.
 
F

fluoxetine

Heisman
Nov 11, 2012
23,529
16,898
0
My system has Colgate #23. The math does work, somehow. It's probably because Army did well OOC?

I'm not much of a NET fan, but I don't think it's fair to judge it based on this season. The thing is designed for normal seasons where teams play normal schedules, not teams that played 10 games against 3 teams all in their own conference.
 
R

RUChoppin

Heisman
Dec 1, 2006
19,270
13,695
0
11 Colgate .... 0-0 Q1, 2-0 Q2, 3-1 Q3, 4-0 Q4
23 San Diego St .... 0-3 Q1, 4-1 Q2, 4-0 Q3, 6-0 Q4
30 Rutgers ... 4-6 Q1, 3-1 Q2, 3-0 Q3, 2-0 Q4
37 Missouri .... 5-4 Q1, 1-1 Q2, 6-0 Q3, 1-0 Q4

Missouri is 6-5 in Q1/Q2 and undefeated 7-0 Q3/Q4.... and is 26 spots behind Colgate?

The NET rankings make zero sense this year.
 
F

fluoxetine

Heisman
Nov 11, 2012
23,529
16,898
0
I mean the whole Patriot League played 8 total OOC games. That's going to break computer rankings, especially ones that don't use prior season expectations.
 
R

RUChoppin

Heisman
Dec 1, 2006
19,270
13,695
0
Colgate's marquee wins are two away tilts vs. #96 Army.... their other wins are against #236 Boston U, and #247 Holy Cross. No non-broken formula should be putting them in the Top 25.
 
  • Like
Reactions: mugrat86
K

Knight Owl

All-Conference
Jul 27, 2001
3,536
2,580
0
The NET now has only two factors and net efficiency is likely weighted as 50% of the calculation now. Colgate does shoot the ball very well. As long as the NET is used mostly to identify tournament-worthy teams (to compare mid-majors to other mid-majors or compare teams in the same conference or similar conferences) and not to seed teams then it’s reasonable.
 
R

RUChoppin

Heisman
Dec 1, 2006
19,270
13,695
0
Knight Owl said:
The NET now has only two factors and net efficiency is likely weighted as 50% of the calculation now. Colgate does shoot the ball very well. As long as the NET is used mostly to identify tournament-worthy teams (to compare mid-majors to other mid-majors or compare teams in the same conference or similar conferences) and not to seed teams then it’s reasonable.
Click to expand...

I'm sure our efficiency would look a lot better if 60% of our games were played against teams rated worse than 275 in kenpom, too (and the other 40% were played against a team rated worse than 180). Have to take quality of opponent into account - efficiency isn't calculated by playing in a vacuum.
 
  • Like
Reactions: DirtyRU
ScarletDave

ScarletDave

Heisman
Oct 7, 2010
34,599
15,353
85
Pretty easy to understand - some flaw in the algorithm where based on early season under normal circumstance (you only played 3 unique opponents so far) the math relies heavily on efficiency metrics, until you play more teams and it can do a true determination of who you beat and who they beat. Colgate playing the same teams over and over again has broken their code.
 
RU#1fan

RU#1fan

Heisman
Mar 7, 2003
23,597
12,308
113
RUChoppin said:
Colgate's marquee wins are two away tilts vs. #96 Army.... their other wins are against #236 Boston U, and #247 Holy Cross. No non-broken formula should be putting them in the Top 25.
Click to expand...

They shouldn’t be in the Top 50.
Some teams are going to get screwed by the NCAA Tournament
 
  • Like
Reactions: mugrat86
P

PSAL_Hoops

Heisman
Feb 18, 2008
13,350
12,657
78
I forgot about the part of these formulas that excludes the stats from your games for the opponents record. If you and your opponents all play only a few of the same teams someone always has to win. The whole formula blows up.
 
R

RU-ROCS

All-American
Feb 5, 2003
12,439
7,647
113
Colgate’s high NET rating seriously damages the credibility of the ranking system. In any system, there are statistical outliers, but No. 11 is ridiculous.
 
  • Like
Reactions: Night Man
G

Greene Rice FIG

Heisman
Dec 30, 2005
40,437
23,613
0
Look at the scoring margin. That is why they are rated so high.

this is why I think scoring margin should have no part in these equations
 
S

Scarlet Blind_rivals

All-Conference
Aug 5, 2001
4,621
4,681
62
Patriot League has 8 OOC games total, Army and Navy went 3-1.
Home 2-0, Neutral 3-1, Road 1-1
They were #14th/32 in 2019-20 and #14/31 in 2020-21

Colgate 9-1, NET 11, Ken Pom 96 (69%min,66%scoring)returned
Last year (25-8) NET 121, Ken Pom 118

3-1 vs Army(8-6), NET 96, KP 185 (66%, 52%)
Last year, 15-15, NET 265, KP 264
4-0 vs Holy Cross(3-9), NET 247, KP 333 (43%, 39%)
Last year, 3-29, NET 343, KP 344
2-0 vs Boston U(4-6), NET 236, KP 279 (86%, 78%)
Last year, 21-13, NET 160, KP 159

Other Patriot league teams
Lehigh 11-21 NET 288, KP 304
2020-21, (3-7) 282, 324
Loyola(Md) 14-17 NET 248, KP 268
2020-21, (2-7) 203, 234
Navy 13-16 NET 243, KP 255
2020-21, (11-2) 90, 167
Bucknell 14-20 NET 239, KP 242
2020-21, (4-4) 193, 246
Lafayette 18-12 NET 223, KP 228
2020-21, (7-5) 205, 245
American 16-14 NET 197, KP 203
2020-21 (2-2) 147, 247

The higher NET rankings are the result of lack of OOC and less metrics than any team or conference being better. They don't deserve an at-large.
 
Last edited:
P

PSAL_Hoops

Heisman
Feb 18, 2008
13,350
12,657
78
It’s not scoring margin or even adjusted efficiency that’s causing it. Their RPI is 13...

It has to be the part of the opponents and opponents’ opponents records in the formula that removes the losses from counting in the record of someone that you beat. Think about it - Holy Cross is 3-9 but in that part of the calculation they are 3-5 since the Colgate games don’t count. Then for Holy Cross’ opponents, they all also played Colgate and Colgate’s other opponents so I don’t know how that ends up working in the math but that’s probably where the formula is getting wacky.
 
S

Scarlet Shack

Heisman
Feb 3, 2004
26,283
15,978
73
3.4.6.7,,19.27.30,36,39.50.53,

big ten is doing great in the net rankings...and quite comparable to kenpom and Bart

Not worried a single outlier or two
 
R

RUChoppin

Heisman
Dec 1, 2006
19,270
13,695
0
Scarlet Shack said:
3.4.6.7,,19.27.30,36,39.50.53,

big ten is doing great in the net rankings...and quite comparable to kenpom and Bart

Not worried a single outlier or two
Click to expand...

Yes, Big Ten is doing well in a broken system... but the system is still broken.

At this point, I don't know how the NCAA can give these rankings to the committee in good conscience.
 
G

Greene Rice FIG

Heisman
Dec 30, 2005
40,437
23,613
0
I’d like to think the NCAA is smart enough to understand why Colgate’s numbers are out of whack. It shouldnt throw out the system because one team has only played 3 opponents.

it isn’t designed to work with a schedule like Colgate. Most other schools have semi normal schedules so the NET definitely has value.
 
G

Greene Rice FIG

Heisman
Dec 30, 2005
40,437
23,613
0
PSAL_Hoops said:
It’s not scoring margin or even adjusted efficiency that’s causing it. Their RPI is 13...

It has to be the part of the opponents and opponents’ opponents records in the formula that removes the losses from counting in the record of someone that you beat. Think about it - Holy Cross is 3-9 but in that part of the calculation they are 3-5 since the Colgate games don’t count. Then for Holy Cross’ opponents, they all also played Colgate and Colgate’s other opponents so I don’t know how that ends up working in the math but that’s probably where the formula is getting wacky.
Click to expand...
Combo of both
 
R

RUChoppin

Heisman
Dec 1, 2006
19,270
13,695
0
Here are the biggest differences between kenpom and NET.... not saying kenpom is necessarily a gold standard, but these are some pretty big variances between two mathematical models...

NET is Worse:
87 places.... NM State, 153 kenpom, 240 NET
81 places.... Iowa State, 148 kenpom, 229 NET
67 places.... E Michigan, 256 kenpom, 323 NET
63 places.... San Diego, 192 kenpom, 255 NET
55 places.... St. Josephs, 216 kenpom, 271 NET
53 places.... Depaul, 130 kenpom, 183 NET

NET is Better:
98 places.... American, 247 kenpom, 149 NET
89 places.... Army, 185 kenpom, 96 NET
86 places.... Holy Cross, 333 kenpom, 247 NET
85 places.... Colgate, 96 kenpom, 11 NET
82 places.... Merrimack, 264 kenpom, 182 NET
76 places.... Navy, 167 kenpom, 91 NET
.... and an additional 12 teams where NET has them 50-70 points higher than kenpom

Looking only at the biggest gaps (20 places or more) for teams ranked in the NET Top 75 (Quad 1 away, Quad 2 home)...
85... Colgate... 11 NET, 96 kenpom
-43... Duke.... 66 NET, 23 kenpom
34... Winthrop... 69 NET, 103 kenpom
32... UC Santa B... 48 NET, 80 kenpom
25... Boise St... 38 NET, 63 kenpom
24... Colorado St... 44 NET, 68 kenpom
23.... Drake... 35 NET, 58 kenpom
-21... Indiana... 50 NET, 29 kenpom
-21... UConn... 59 NET, 38 kenpom
 
P

PSAL_Hoops

Heisman
Feb 18, 2008
13,350
12,657
78
RUChoppin said:
Yes, Big Ten is doing well in a broken system... but the system is still broken.

At this point, I don't know how the NCAA can give these rankings to the committee in good conscience.
Click to expand...

Unfortunately, no non-biased quant system can really fix the problem. The issue as someone pointed out is that only Army and Navy from the Patriot League played OOC games.

Army went 3-1 in D1 games including neutral wins over Buffalo and Central CT, home vs La Salle, and losing a close neutral game vs. Florida. Navy notched road wins at Georgetown and at Mount St Mary’s, beat GW and lost at Maryland. That’s the entire non-conference representation for the Patriot League.

So in summary, in non-conference the Patriot league is being carried by the 6-2. mark that Army and Navy put together with a total of only 2 games being played on their home floors and the two losses being a true road game to a bubble team and a close neutral loss to a tourney team. Now we all know that if the Holy Crosses of the conference played non-conference schedules they would’ve racked up terrible losses. However, that didn’t happen and unfortunately since there’s always a winner and a loser in conference matches it’s non-conference play that drives the overall output of the computer numbers. There’s no other non-biased way for the math to shake out (in conference play the aggregate record is always equal number of wins and losses). Like it or not, most major conferences didn’t produce a better blended paper OOC resume than Army and Navy.
 
  • Like
Reactions: Doctor Worm and Scarlet Blind_rivals
S

Scarlet Blind_rivals

All-Conference
Aug 5, 2001
4,621
4,681
62
They "fixed" a system that wasn't broken. I thought the NET rankings was the best the NCAA and selection committee created that couldn't be utilized in 2019-20. This year, the NET rankings resembles too much like the old RPI rankings where Rutgers doesn't miss MM, once under Bannon and once under Waters, if the 2019-20 NET system was in place then.

Vermont(3-3) winning 2 @ Binghamton(1-8), Jan 9-10
Vermont(5-3), Binghamton(1-10), 76-60, 84-44
VT 225 to 173 to 140, today 104
Bing 291 to 307 to 324, today 324
Vermont (9-3,9-3), Binghamton (2-13,2-10), L Marist(9-7,7-7) twice and...

Youngstown St beat Binghamton, only D1 OOC game, and 4 Non D1 wins
Youngstown St(3-4) won @ Wright St(7-2), 74-72 and lost 55-93, Jan 9-10
YSU(4-5), Wright St(8-3)
YSU 248 to 211 to 232, now 257
WSU 79 to 102 to 86, now 60
YSU(13-10,8-10), Wright St(17-4,15-3)

I know efficiency, SOS, etc. plays a part in the jumps, but they need more metrics not less to get a better system in place that Colgate should not be ranked where they are even if they finish the season 13-1(3-1 vs Army), #96 Ken Pom Q1 0-0 Q2 2-0, unlike Loyola Chi(16-4 D1) has earned their #10, #9 Ken Pom, Q1 1-2 Q2 4-2.
 
Last edited:
P

PSAL_Hoops

Heisman
Feb 18, 2008
13,350
12,657
78
Using Kempom as the system for selection wouldn’t exactly be fair either. Then you’d be awarding potential rather than accomplishment in many cases. The thing is - if Colgate doesn’t win the autobid they will take another bad loss and likely tank into the 30s with a 14-2 record. That’s not nearly as obscene as where they are now. We’ve seen SMU type teams not get in with those type numbers before in the old RPI system. If colgate wins out, they’ll have the autobid anyway so it won’t really matter. They don’t usually use NET for seeding.
 
  • Like
Reactions: RUChoppin
G

Greene Rice FIG

Heisman
Dec 30, 2005
40,437
23,613
0
PSAL_Hoops said:
The point is the RPI is a pure win system that disregards efficiency and point margin and Colgate still ranks very high.
Click to expand...
That stuff probably drives it much more than scoring margin.

How do they beat ARMY 101-57 and then come back the next day and lose 73-75?
 
G

Greene Rice FIG

Heisman
Dec 30, 2005
40,437
23,613
0
Do we know if kenpom and bart are at the point of the season where there still isn't some predicted info in the system?
 
You must log in or register to reply here.
Top Bottom