NU Prof's Computer Rankings for 1/11/10

[NUProf]

Okay a day early, but if you want some evidence for your poll speculation thread, or if there are poll voters out there who need some advice- here they are: the definitive authoritative D3 Computer rankings .

As always - if you see an error in a team's record, let me know (remember a tie is reported here as half a win and half a loss)

Some interesting moves this week - 5 count 'em, 5 NESCACs ranked 7-8-9-10-11 and Midd doing what they often do - rising after a questionable start - how many times over the years have we counted them out after the first couple of weeks only to have them revived.

And to the poster in the poll speculation thread asking about Fredonia - I've got them at 15.

As to Adrian - they are in free fall in my rankings.

                G        W	L	Score	LW ScoreLW Rank
1	OSW	17	16	1	709	696	1
2	NRW	13	11.5	1.5	681	668	3
3	ELM	14	10.5	3.5	677	662	4
4	PLT	14	10.5	3.5	665	661	5
5	STS	14	11	3	663	661	5
6	SNB	15	12.5	2.5	661	675	2
7	MID	10	7.5	2.5	622	612	11
8	AMH	11	8.5	2.5	620	606	12
9	WIL	9	7.5	1.5	620	624	9
10	CLB	11	8	3	618	583	18
11	TUF	10	6.5	3.5	613	600	13
12	HML	13	9.5	3.5	612	642	7
13	GAC	14	10.5	3.5	602	595	14
14	BOW	10	7	3	596	567	21
15	FRE	14	11	3	595	566	22
16	SAL	11	6.5	4.5	594	633	8
17	WEN	12	10.5	1.5	587	575	19
18	TRN	11	6.5	4.5	586	616	10
19	WST	16	7.5	8.5	582	590	17
20	POT	15	8	7	580	569	20
21	WSR	15	9	6	576	561	25
22	CUR	12	7.5	4.5	574	593	16
23	WRF	15	9.5	5.5	567	552	28
24	HAM	11	5.5	5.5	565	594	15
25	HOB	15	8	7	563	525	36
26	UMB	11	6.5	4.5	560	554	27
27	FIT	11	7.5	3.5	559	560	26
28	MAN	16	12	4	557	534	34
29	AUG	13	9.5	3.5	556	525	36
30	NEU	15	9	6	552	538	33
31	BAB	12	6.5	5.5	551	563	23
32	WEC	15	6.5	8.5	549	548	30
33	WSP	17	8.5	8.5	546	546	31
34	WFS	12	7	5	535	562	24
35	ADR	13	10	3	535	544	32
36	CST	12	5	7	533	477	50
37	UTC	14	6.5	7.5	529	521	38
38	NEC	12	6.5	5.5	528	517	39
39	STT	14	6.5	7.5	528	534	34
40	MOR	13	4	9	522	517	39
41	CTC	12	3	9	518	475	52
42	STO	14	7	7	518	551	29
43	MAR	12	6	6	515	501	45
44	SOM	12	5.5	6.5	512	504	44
45	STA	10	6	4	508	465	55
46	SKD	12	5	7	502	467	54
47	STJ	15	6	9	501	488	49
48	BRC	13	6	7	494	501	45
49	GEN	15	7	8	493	501	45
50	BET	15	6	9	493	510	41
51	PLY	12	6	6	490	506	42
52	UMD	11	5.5	5.5	484	500	48
53	BEC	14	6.5	7.5	483	476	51
54	BFS	15	5.5	9.5	467	452	59
55	WES	10	4	6	463	505	43
56	NIC	12.5	6.5	6	461	472	53
57	CRT	14	4	10	459	461	56
58	SUF	12	5	7	456	455	58
59	LKF	12	4	8	453	434	64
60	STN	14	7.5	6.5	447	448	60
61	LAW	10	5	5	443	457	57
62	WOR	10	3	7	436	446	61
63	AMP	12	5	7	432	443	62
64	J&W	12	4	8	428	413	66
65	CNC	15	4	11	427	443	62
66	MSE	13	6.5	6.5	424	411	67
67	STM	11	2.5	8.5	422	402	69
68	FPU	14.5	4	10.5	407	428	65
69	UNE	11	1	10	386	404	68
70	STR	13	1.5	11.5	376	400	70
71	FRA	11	1.5	9.5	360	374	71
72	SNH	13	3	10	360	364	72
73	NLD	12	2	10	358	360	73
74	WNE	11	1.5	9.5	325	328	74
75	SLV	12	1	11	311	311	75
76	FNL	13	3.5	9.5	303	286	77
77	LVC	14	0	14	302	303	76
78	CNW	13	0.5	12.5	291	263	78

 

[norm1909]

Re: NU Prof's Computer Rankings for 1/11/10

Eastern BIAS (Built In Adjustment System)

VECTOR NRW <= 3.
- DO IF NRW < XXX.
- ADJUST NRW > XXX.
- ELSE.
- GOWITHIT.
- END IF.
EXECUTE.

 

[jazman]

Re: NU Prof's Computer Rankings for 1/11/10

You must place a great emphasis on the opponet. How does a team who wins twice over the weekend drop from 2 to 6 without a loss.

 

[CARDS_rule_the_Burgh]

Re: NU Prof's Computer Rankings for 1/11/10

You must place a great emphasis on the opponet. How does a team who wins twice over the weekend drop from 2 to 6 without a loss.

Because he factors in OWP oF Opponents Defeated. Norbert beat 3-10-2 Concordia MN, who has the fourth worst record in the west, and 5-8-2 St. John's, who is only slightly better than that, but is doing miserably in-conference, which does significant damage to their overall record.

Compare that to the defeated opponents above them: Oswego defeated 10-3-0 Adrian, Norwich defeated 9-2-3 Plattsburgh, Plattsburgh and Elmira tied, and St. Scholastica beat 8-2-3 Hamline and 6-9-0 Bethel.

SNC did good in winning, but this ranking isn't going to reward them for beating teams they SHOULD beat.

EDIT: Part of this could be that these opponents hurt SNC in both DOWP and OWP, both of which are variables in NU Prof's equation. Moral of the story: don't play weak opponents like this.

 

[jazman]

Re: NU Prof's Computer Rankings for 1/11/10

Because he factors in OWP oF Opponents Defeated. Norbert beat 3-10-2 Concordia MN, who has the fourth worst record in the west, and 5-8-2 St. John's, who is only slightly better than that, but is doing miserably in-conference, which does significant damage to their overall record.

Compare that to the defeated opponents above them: Oswego defeated 10-3-0 Adrian, Norwich defeated 9-2-3 Plattsburgh, Plattsburgh and Elmira tied, and St. Scholastica beat 8-2-3 Hamline and 6-9-0 Bethel.

SNC did good in winning, but this ranking isn't going to reward them for beating teams they SHOULD beat.

EDIT: Part of this could be that these opponents hurt SNC in both DOWP and OWP, both of which are variables in NU Prof's equation. Moral of the story: don't play weak opponents like this.

So the flaw would be that if you spend your time beating eachother up in a conference of really good teams that it looks like your opponets are not any good when in fact they a very good.

 

[mazile]

Re: NU Prof's Computer Rankings for 1/11/10

So the flaw would be that if you spend your time beating eachother up in a conference of really good teams that it looks like your opponets are not any good when in fact they a very good.

In reference to the NCHA, after St. Scholastica, who are these really good teams you speak of? I think that it's perhaps the other way round, an average bunch of teams play a weak out-of-conference schedule to build their resume, then play an in-conference schedule against mediocre opponents of similar ability.

 

[NUProf]

Re: NU Prof's Computer Rankings for 1/11/10

Cards described the situation very well, as well as I can describe it. The other thing that my model includes that tries to avoid rewarding mediority is that it goes deeper into the schedule - A minor factor is an inclusion of oooWP - Teams that play tough teams that play tough teams get rewarded. Can it be gamed? of course, but I doubt that anybody builds their schedule with my metrics in mind - I know I'm good, but not that important . You have to beat the teams on your schedule, but you you get a bonus when you beat teams that do well. The other thing about SOS metrics in your league - they all tend toward 50% and the real thing that is measured is how well your league cohort does in its out of conference play.

My method is somewhat Krach-like. Krach really goes deeper into the schedule than I do, but again has a component that basically asks "how good are the teams you beat?"

I knew somebody would wonder how SNC fell, but the reality is that although they had a good week, relatively speaking other teams (like Norwich) had better weeks by beating higher ranked teams.

 

[jazman]

Re: NU Prof's Computer Rankings for 1/11/10

In reference to the NCHA, after St. Scholastica, who are these really good teams you speak of? I think that it's perhaps the other way round, an average bunch of teams play a weak out-of-conference schedule to build their resume, then play an in-conference schedule against mediocre opponents of similar ability.

A conference were 6 of the 7 teams get votes in the rankings can't be all bad.

 

[17'sfanNU]

Re: NU Prof's Computer Rankings for 1/11/10

Good job, NU Prof, you got the first two right, Oswego 1, Norwich 2, St. Norbert 3 and Plattsburgh 4

 

[NUProf]

Re: NU Prof's Computer Rankings for 1/11/10

Good job, NU Prof, you got the first two right, Oswego 1, Norwich 2, St. Norbert 3 and Plattsburgh 4

Actually, my rankings agreed with the identity of top 6 teams, but the order was scrambled a bit with 3-6. From there on down, it became a little different