Unbeatens | The Boneyard

Unbeatens

trisailin

trisailin

Joined
Dec 27, 2011
Messages
600
Reaction Score
2,531
Through games 12/1. (number of wins and alphabetical)
School​
Wins​
Binghamton​
8​
Louisville​
8​
NC State​
8​
Stanford​
8​
Arizona​
7​
Florica St​
7​
Kentucky​
7​
Oregon St​
7​
Tennessee​
7​
UCLA​
7​
Colorado​
6​
Connecticut​
6​
Kansas​
6​
Loyola-Chicago​
6​
North Carolina​
6​
TCU​
6​
Providence​
5​
Texas Tech​
5​
 
Last edited:
Not that I’m ambitious enough to do this, but it would be interesting to see each of the unbeaten teams ranked by SOS.
 
oldude said:
Not that I’m ambitious enough to do this, but it would be interesting to see each of the unbeaten teams ranked by SOS.
Expand ...

Not until UConn's SOS improves please.
 
.-.
trisailin said:
Through games 12/1. (number of wins and alphabetical)
School​
Wins​
SOS​
Binghamton​
8​
341​
Louisville​
8​
6​
NC State​
8​
124​
Stanford​
8​
9​
Arizona​
7​
199​
Florida St​
7​
108​
Kentucky​
7​
134​
Oregon St​
7​
33​
Tennessee​
7​
198​
UCLA​
7​
106​
Colorado​
6​
191​
Connecticut​
6​
5​
Kansas​
6​
300​
Loyola-Chicago​
6​
308​
North Carolina​
6​
213​
TCU​
6​
164​
Providence​
5​
237​
Texas Tech​
5​
332​
Expand ...


Apparently I had the time this morning. These SOS based on Massey Ratings.
 
CBus13 said:
Apparently I had the time this morning. These SOS based on Massey Ratings.
Expand ...
I Think Massey is probably more accurate, particularly with regard to OR St’s SOS. I pulled my SOS numbers off of some sight named RealTimeRPI.
 
oldude said:
I Think Massey is probably more accurate, particularly with regard to OR St’s SOS. I pulled my SOS numbers off of some sight named RealTimeRPI.
Expand ...
This is the one I usually use.
 
.-.
RealTimeRPI just got me in trouble on this site. They didn't have the Stanford-MSU game in their archive - probably because it was a tournament game and wasn't "pre-programmed" into their database.
 
Just a comparison with this SOS between WarrenNolan (left) and Massey, top 20 only for both. If anyone can decipher why there are big differences between both of them, please do not hesitate to reply.

EDIT: just realized Oregon St is not even shown in Massey's top 20 in SOS.

1575307284480.png
 
mbr33ct said:
Just a comparison with this SOS between WarrenNolan (left) and Massey, top 20 only for both. If anyone can decipher why there are big differences between both of them, please do not hesitate to reply.

EDIT: just realized Oregon St is not even shown in Massey's top 20 in SOS.

View attachment 48750
Expand ...
Massey counts last ten wins, so that includes the end of last season, but I don't think that explains much of it. I think the main thing is that Massey weighs good opponents better, because Massey takes margin of victory into account, so when Baylor beats cupcakes by 70 points, that looks different than when Missouri State beats 2-4 Boise State by 3. Which then means that when Washington State loses to Baylor, that SOS will look much stronger than when Oregon State beats Missouri State, even though Baylor and Missouri State have identical 7-1 records overall. You'll note that the RPI SOS indicator on the left just totals up all opponent wins and losses, with no consideration of who they beat (which RPI weighs, but not RPI SOS), and by how much (which RPI ignores entirely).

Among other laughable things is that RPI produces the result that Oregon State has a stronger SOS than Washington State, which has already played Baylor, Indiana, and South Carolina. OSU's best opponents are DePaul, Miami, and Missouri State, but sure, whatever.
 
Last edited:
TheFarmFan said:
Among other laughable things is that RPI produces the result that Oregon State has a stronger SOS than Washington State, which has already played Baylor, Indiana, and South Carolina. OSU's best opponents are DePaul, Miami, and Missouri State, but sure, whatever.
Expand ...
Yet one more indication that RPI is completely nonsense. Yet, supposedly bright people of America's institutions of higher learning continue to use it.
The ONLY reason they use it is for politics.
 
RPI is based only on opponent records and the opponents'-opponent records. It doesn't go any deeper, and it (per NCAA dictum) absolutely does not consider margin-of-victory.

Most "name" formulas, like Massey and KenPom, either directly or indirectly dig deeper than just the two levels that RPI goes. Typical is the Elo formula (Sagarin), which gives every team a rating and with each game the winner "takes" rating points from the loser, in proportion to the difference in rating (much like a wager). For example, an Elo formula might set it up so that if two evenly-matched teams play, the winner takes 16 points from the loser. But in a UConn-Gallaudet type mismatch, UConn might take 1 point from Galluadet if they win, but lose 32 points to them if they lose. Over the course of the season, teams that play strong opponents will be able to gain more points than those who play weak opponents. You have to beat Gallaudet 16 times to get the same effect as playing Baylor. (The numbers I used here are typical, but not precise.)

But here's the upshot. Whether it's Elo, ISR, or any other computer rating system, it takes on the order of two dozen games for the signal to overcome the noise. No computer rating means much of anything before February, and they only really have validity in March. Massey tries to get around this by adding some of last year's games to the mix - if they didn't, their ratings would look as stupid as the current RPI lists.

The one area where we really legitimately look to the computers for support for our arguments TODAY is in the conference-vs-conference comparisons, as we have far more games going into the calculation. The conference ratings would become legitimate roughly 6 games into the season if every team played a "decent" schedule. As things stand, with all the patsies being scheduled in November, it's probably closer to 10-12 games in before the conference ratings are legit.
 
.-.
TheFarmFan said:
They have Stanford and MSU's other wins in the tournament, so I don't think that's it. Whatever the deal is, if they don't fix it, it's going to throw things off all season. Here's the warrennolan RPI, which does include all games. Incidentally, they also calculate ELO rankings, which are also pretty interesting, and the closest statistical ranking that mirrors my own subjective ranking of the teams as of today.
Expand ...
By the way, Elo is not an acronym. The Elo rating system gets its name from its inventor, Arpad Elo. So it shouldn't really be written ELO.
 
I know that. I was calculating ELO ratings for the Northwest Chess Association in the 1970's on a hand calculator. But even way back then, it was almost always written ELO. I suppose because they were usually calculated on a computer and computer terms were almost always inappropriately capitalized.
 
It's fun to see the teams that some of these undefeated teams have beaten.

Texas Tech (4-0) has beaten:
  • Sam Houston State (1-3)
  • Florida A&M (0-4)
  • Northwestern State (0-3)
  • Purdue Fort Wayne (1-5)
If Texas Tech can beat Ole Miss in the Big 12/SEC Challenge tomorrow, they'll likely go unbeaten until Big 12 play, as their remaining OOC schedule is:
  • Houston Baptist (1-4)
  • Prairie View A&M (0-5)
  • Arkansas Pine Bluff (0-5)
  • Louisiana Monroe (1-5)
  • Texas San Antonio (1-4)
These 9 OOC opponents have a combined record vs. D-I teams of 5-38. Yikes.
 
Last edited:
.-.
These rankings only prove that people will use the ones that fit their own agenda's. Everyone that understands rankings will generally give more credibility to Massey than they do to RPI. But they also understand that Massy is highly flawed early in the season because it uses the last ten games played. Well no teams have played ten games yet, so games from last season are factored into the score. For example that means that the last four games Uconn played last season are factored into their Massey score. They were all NCAA tournament teams, the last two being final eight and final four. The same holds true for other teams depending on how many games they have played this season.

Could anyone actually believe that UConn has played the fifth toughest schedule in the country this season? The farther a team went in the NCAA's, multiplied by the number of games used from last season would inflate their strength of schedule for this season. A team has to have played at least ten games for the Massey numbers to have any validity.

Based on this Cbus13, why would you use Massey as the standard at this time for a SOS metric?
 
@willtalk I had no real critical thinking when it came between choosing Massey over RPI other than I'm used to looking at Massey. In my ignorance, I didn't realize it used the last 10 games. So that was my error.


I've updated the table below to have the current Massey SOS and the RPI SoS (from RealTimeRPI) for a comparison. Also, removed Providence and did alphabetical order...Not sure if it's an order I like yet.

School
Wins​
Massey (Previous)​
RPI​
Arizona
8​
208 (199)​
278​
Binghamton
8​
341(342)​
342​
Colorado
7​
200 (191)​
186​
Connecticut
6​
4 (5)​
2​
Florida State
7​
106 (108)​
147​
Kansas
7​
298 (300)​
215​
Kentucky
8​
127 (134)​
146​
Louisville
8​
7 (6)​
13*​
Loyola-Chicago
7​
300 (308)​
247​
North Carolina
6​
199 (213)​
283​
NC State
8​
125 (124)​
235*​
Oregon State
7​
32 (33)​
3​
Stanford
8​
8 (9)​
57​
Tennessee
7​
195 (198)​
345​
TCU
6​
160 (164)​
136​
Texas Tech
6​
321 (332)​
321​
UCLA
7​
108 (106)​
164​


*RPI had these teams with one loss. Not sure why. Also, listed Mississippi State and Indiana as undefeated.
 
Could somebody point me to where it definitively says Massey uses the last 10 games? We know they start the season "priming the pump" with some data from last year - but I'm not aware of anything that specifically says "last 10 games".
 
trisailin said:
Could somebody point me to where it definitively says Massey uses the last 10 games? We know they start the season "priming the pump" with some data from last year - but I'm not aware of anything that specifically says "last 10 games".
Expand ...
I'm curious about this as well. The only thing official I've read leaves it relatively vague:

Preseason ratings are typically derived as a weighted average of previous years' final ratings. As the current season progresses, their effect gets damped out completely. The only purpose preseason ratings serve is to provide a reasonable starting point for the computer. Mathematically, they guarantee a unique solution to the equations early in the season when not enough data is available yet.

 
.-.
You must log in or register to reply here.

Forum statistics

Threads
168,512
Messages
4,579,712
Members
10,488
Latest member
Azerion


Top Bottom