NCAA RPI

With UConn, G-Town and Marquette nipping at their heels and all still alive in the BE tournament MD may need to at least make it to the ACC final though the BE tournament is going to give the bigger bump with the BE Final Four representing current RPI slots of 1,5,6,7.

If the Hoyas take the BE it is possible for MD to win the ACC and still slip to #5.

 

UNC gets a bump up from 12 to 8. If they have a decent showing at the ACC tourney, the Heels should get a top 8 seed, despite a not-so-stellar OOC schedule

 

well the SOS is # 7 according to one source

(https://docs.google.com/spreadsheet...&single=true&gid=12&range=A1:G206&output=html)

and includes a road tie with #4 MD
wins over
#15 VA on the road
#17 Washington (on the road, avenging the earlier home loss)
#18 Tulsa
#49 Cal Poly
#65 Stanford

it all adds up..... those are NCAA RPI "rankings" by the way

 

If you look at their schedule, their only losses are to New Mexico (12) and Washington (17) and they came East and beat Virginia (15) and tied Maryland (4). They avenged the loss to Washington (again, 17) with a road win over the Huskies and have a win over Tulsa (18) and a win and a tie with Cal (38).

That's 4-2-2 against teams in the Top 38 and 3-2-1 against teams in the Top 18, with some of those results coming on the road. They also have wins over Cal Poly (49), twice over Stanford (65), San Diego State (89) and twice over Oregon State (93), Loyola Marymount (120) and UC Irvine (137) and a tie with UC Santa Barbara (72).

EDIT: My phone rang mid-post and Well Duh had posted his answer by the time I hit send.

 

Assuming the seeds are directly in line with the RPI at least. Still would be advisable to at least make the final to be safe.

 

What stats are you talking about ?

 

ok ...

UCLA #4

a road tie with #4 MD
wins over
#15 VA on the road
#17 Washington (on the road, avenging the earlier home loss)
#18 Tulsa
#49 Cal Poly
#65 Stanford

UNC # 8

home tie with #16 WF
Wins
#15 VA
#30 BC
#42 Duke
#45 WV
#47 Campbell

they look different to me....

 

as in most any sport..... it is about W's ...unless we have to talk about the BCS process

 

Akron beat WMU 1-0 on a goal in the 85th minute. Does that look like an impressive win?

What if I told you they outshot WMU 23-2 and 9-0 on goal. That WMU had zero shots in the second half. That possession was (generously) 85-15 in favor of Akron. The win looks a bit more impressive right?

Don't judge a game by the scoreline.


PS you should know all this since you were at the damn game

 

Right idea but you have it backwards. There are not enough weak teams in the west which which means that the RPI for the entire region is low. Take a look at posts 33 and 34 of RPI explained thread and you will see the stark differences between the RPI of men's soccer teams in the west and the RPI of the women's soccer teams in the west. The negative bias against the men's teams in the west is unmistakeable.

 

Niagara has an .813 winning percentage playing the 161 ranked schedule (per Massey). Neither Massey nor RPI rank them in the top 50. If they lose their conference playoff, both will drop. Unless you feel that every conference should get an at-large bid, I'm at a loss as to why you believe a team with an unremarkable record against a schedule that is among the easiest in D1 that can't even win the conference championship in one of the lowest ranked conferences should get an at-large bid. Sure they have just one loss, but they also have 4 ties against an extremely weak schedule.

 

One good reason is that they are clearly the top team in the top conference. That said, I've watched them quite a bit since they are on the Pac12 and haven't looked that impressive. Then again, the only teams that have really impressed me have been Akron and Maryland.

Conference	Tms	Record	Pct	Rank	SoS
Pac 12	6	30-14-6	0.66	1	1
Atlantic Coast	9	56-21-8	0.706	2	2
Big East	15	88-32-15	0.707	3	4
Big 10	7	42-29-10	0.58	4	3
Conference USA	9	46-26-12	0.619	5	5
Big West	7	37-28-8	0.562	6	6
Colonial	11	42-32-7	0.562	7	9
West Coast	7	19-24-6	0.449	8	7
Missouri Val	7	42-28-10	0.588	9	12
Mtn Pacific SF	9	31-32-10	0.493	10	8
Southern	8	39-36-8	0.518	11	10
Ivy League	8	34-33-7	0.507	12	11
America East	8	35-34-12	0.506	13	13
Atlantic 10	16	65-59-19	0.521	14	15
Mid-American	8	35-37-12	0.488	15	14
Patriot League	8	29-45-8	0.402	16	16
Horizon	8	27-47-5	0.373	17	17
Metro Atlantic	10	36-48-9	0.435	18	18
Big South	11	30-49-8	0.391	19	20
Atlantic Sun	9	29-51-11	0.379	20	19
Northeast	11	29-45-10	0.405	21	22
Summit Lg	8	19-58-7	0.268	22	21
Atlantic Soccer	3	13-33-1	0.287	23	23

 

When using an unbiased source like Massey, it is even more revealing:

Team	Conf	Record	MAS	RPI	Diff
New Mexico	Mtn Pacific SF	15-3	10	12	-2
Washington	Pac 12	11-4-3	14	17	-3
California	Pac 12	8-6-3	33	38	-5
Stanford	Pac 12	8-8-1	36	65	-29
Cal Poly SLO	Big West	10-6-1	40	49	-9
UC Santa Barbara	Big West	10-6-3	41	72	-31

 

Nonsense

 

What makes Massey "unbiased"?

 

Any drunken fool from the flat earth society can scream nonsense to someone trying to help him understand why the earth is actually round. Since it is mathematically provable that the RPI calculation will under rank members of a smaller, isolated population whose values are skewed higher than the general population, I’m not expecting you to offer a rational argument effectively that refutes the negative bias towards the western teams that I have demonstrated. But I'd like to see your attempt. On the hand, I marvel at your dogged determination to hold onto your belief about accuracy of RPI even when that belief is detrimental to the interests of your own child’s team.

 

You bring up a valid point. Since absolute strength of each team is unknowable, it is impossible to ascertain with absolute certainty whether the skew in rankings is due to weakness of RPI or those of Massey. That said, it doesn’t take very long to show with a high degree of certainty the problem is with the RPI.

First it can easily be proven with a deck of cards that the crude strength of schedule component of the RPI calculation can easily lead to erroneous results such as a ten being ranked higher than a Queen. In contrast, with Massey each match connects two teams (or cards) and as more games are played each team quickly gets connected with every other team such that a Queen would always be ranked higher than a Jack.

Second, the significant ranking skew between the RPI and Massey rankings only occurs with the men’s teams in the west but not the women’s teams. The difference between the men’s and women’s side is easily explained by the crude nature of the RPI calculations. On the other hand, there is really no good explanation of why the Massey calculations for only men’s soccer teams in the west would be so uniquely positively skewed.

Here is a link to the comparison of the women: https://www.bigsoccer.com/community/threads/rpi-explained.1976256/page-2#post-26733628 (I just did the top 50 because I'm lazy and can't get the data automatically but you should still get the point)

Here is the comparison for every men's team in the west. All but 5 are negative to very negative. Since you are so certain of the validity of the RPI, why don't you try to explain the difference between the results for the men and women.

Team	Conf	Record	PCT	Rank	RPI
UCLA	Pac 12	12-2-3	0.794	4	3	1
Washington	Pac 12	11-4-3	0.694	14	17	-3
UC Davis	Big West	9-5-4	0.611	31	40	-9
CS Northridge	Big West	13-6	0.684	32	39	-7
California	Pac 12	8-6-3	0.559	33	38	-5
Stanford	Pac 12	8-8-1	0.5	36	65	-29
Cal Poly SLO	Big West	10-6-1	0.618	40	49	-9
UC Santa Barbara	Big West	10-6-3	0.605	41	72	-31
Oregon St	Pac 12	6-8-3	0.441	65	93	-28
San Diego St	Pac 12	7-8-3	0.472	72	89	-17
UC Riverside	Big West	10-8	0.556	83	102	-19
CS Fullerton	Big West	8-11-1	0.425	92	87	5
UC Irvine	Big West	5-14-1	0.275	123	137	-14
New Mexico	Mtn Pacific SF	15-3	0.833	10	12	-2
St Mary's CA	West Coast	9-3-3	0.7	34	55	-21
CS Bakersfield	Mtn Pacific SF	8-4-5	0.618	43	57	-14
Denver	Mtn Pacific SF	10-5-3	0.639	47	51	-4
Santa Clara	West Coast	9-6	0.6	53	52	1
San Diego	West Coast	10-7	0.588	66	91	-25
CS Sacramento	Mtn Pacific SF	8-8-2	0.5	75	73	2
San Francisco	West Coast	5-6-3	0.464	79	99	-20
San Jose St	Mtn Pacific SF	8-8-1	0.5	86	106	-20
Air Force	Mtn Pacific SF	8-8-3	0.5	94	112	-18
Portland	West Coast	7-9-1	0.441	96	100	-4
Loyola Marymount	West Coast	5-10-3	0.361	107	120	-13
UNLV	Mtn Pacific SF	6-12-1	0.342	128	136	-8
Gonzaga	West Coast	4-13	0.235	175	174	1
Seattle	Mtn Pacific SF	3-13-1	0.206	179	180	-1

[/FONT]

 

The best way to have a good RPI is to beat the best team in a weak conference that will dominate their league and rack up wins. The west's problem is just that there aren't enough teams period and that the schools are unable/unwilling to spend the money to travel outside of their region or entice schools to come to them. Presumably there's enough weak teams (bottom half of division I) for west coast teams to rack up easy wins: Seattle 180, Gonzaga 174, Houston Baptist 145, UNLV 136, Loyola Marymount 120, Airforce 112, San Jose St 106, UC Riverside 102. Portland 100 and San Fransisco right inside the top half. SO the issue is not that there are no weak teams for West coast teams to play it just that there's only a handful of them period so they're all in the same conferences so they all get bound together. The problem for west coast fans is they think all of their teams are strong so don't acknowledge the weak teams. Yes there are more options for east coast teams, but a west coast team could easily make a schedule that would allow them to game the RPI. Just play teams out of conference that will rack up wins in their conference. Basically contenders from the west coast conferences should be playing each other.

The issue is west coast fans assume they should get 10 bids when really they're getting all the bids they deserve based on the system in place.

 

The RPI is a calculation based on won-lost record using a calculation whose method is a matter of record. Period.

On what basis do you derive your judgements of "under rank" and "negative bias"? Compared to what?

 

Where's Nate Silver when you REALLY need him?

 

You'd think he would have jumped in by now. Not like he's been projecting anything else recently ya know?