(11-02-2017 09:49 AM)Neelys Ghost Wrote:  That's an awful lost of numbers... I ain't that smart.. But the eye test is the best metric I know... The eyes don't lie...

Never been to a magic show or left a bar at closing time with a 10?

(11-02-2017 09:30 AM)gsloth Wrote:  Taking the SRS ratings from sports-reference.com, I went back to 1965 to see what it thought some of the best and worst teams were, against expected competition in that year. The full list of SRS values is at the end. Definition of SRS (simple rating system):

Quote:a rating that takes into account average point differential and strength of schedule. The rating is denominated in points above/below average, where zero is average, although it should be noted that margin of victory has been assigned a lower bound of 7 points and an upper bound of 24 points. Non-major opponents are included as a single team in the ratings.

source: https://www.sports-reference.com/cfb/abo...y.html#srs

Obviously, this still doesn't provide a way to compare quality of what was considered average in different years. Sagarin doesn't either, though.

Code:

Lowest             Highest
1983    -17.21     1972    8.38
2017    -16.35     1970    5.82
2009    -16.33     1980    5.53
2005    -15.03     1973    4.93
2007    -14.59     2008    3.92
2016    -13.67     1996    3.88
1982    -13.59     1966    3.85
1988    -11.93     1967    1.09
1987    -11.41     1998    0.53
2015    -11.36     1990    0.50

By this metric, 1983 barely squeezes out the "win" as worst team. And considering margin of loss is capped at -7 (see this link for a fuller discussion), this is truly bad play, for that as well as several recent Bailiff iterations. (Berndt doesn't escape, either.)

As for tops, well, it's definitely hit or miss. Rice has never consistently built a winner. Those early 70s teams might be the most consistently successful overall, with Hatfield a close second (but he had some bad years, too). The only other years with a positive SRS are 2013 and 1971 (each barely).

I agree that the eyeball test of, say, the 2016 team and the 1988 team (both of which I've seen) says that even this metric doesn't do the difference justice. The 1988 team would be better, even with its offensive struggles.

Code:

1965    -3.53    1976     -6.11    1987    -11.41    1998      0.53    2009    -16.33
1966     3.85    1977     -8.64    1988    -11.93    1999     -3.59    2010    -11.08
1967     1.09    1978     -1.07    1989     -4.93    2000     -7.49    2011     -8.15
1968    -4.49    1979     -4.58    1990      0.50    2001     -4.35    2012     -3.52
1969    -1.65    1980      5.53    1991     -5.62    2002    -10.21    2013      0.14
1970     5.82    1981     -3.18    1992     -1.33    2003     -6.96    2014     -0.28
1971     0.05    1982    -13.59    1993     -2.81    2004     -9.31    2015    -11.36
1972     8.38    1983    -17.21    1994     -2.48    2005    -15.03    2016    -13.67
1973     4.93    1984    -11.06    1995     -8.04    2006     -3.89    2017    -16.35
1974    -0.18    1985     -6.59    1996      3.88    2007    -14.59        
1975    -0.41    1986    -10.24    1997     -1.02    2008      3.92

Fascinating.

I think what is most telling from this is how smoke-and-mirrors 2013 was. The conference championship and all that, but this isn't even a top 10 Rice team

I was off doing the tour of fall foliage in southeastern New Mexico and west Texas, and I didn't have time to make a post after the Louisiana Tech game. The news wasn't good in any case. The late-game 42-28 loss to Tech yielded a Massey performance rating of 29.11, while the turnover-plagued 52-21 loss to UAB produced a 12.03, the second-lowest game of the season. Here's how the season has gone so far:

vs.Stanford: 7.61
@Texas El-Paso: 42.04
@Houston: 25.13
Florida International: 34.62
@Pittsburgh: 30.24
Army: 13.06
@Texas-San Antonio: 32.24
Louisiana Tech: 29.11
@Alabama-Birmingham: 12.03

Here are the numbers for all Conference USA members:

Team, Median, Mean, Standard Deviation:
1. Florida Atlantic 60.31 58.32 14.47
2. Marshall 49.40 53.32 13.52
3. Southern Mississippi 46.22 43.33 11.47
4. North Texas 44.81 43.64 12.44
5. Louisiana Tech 43.34 43.62 9.37
6. Texas-San Antonio 41.85 44.77 8.15
7. Alabama-Birmingham 41.65 39.86 14.28
8. Western Kentucky 41.29 39.80 9.41
9. Middle Tennessee 41.03 42.53 16.20
10. Old Dominion 36.61 33.93 11.18
11. Charlotte 34.48 28.62 13.18
12. Florida International 32.34 38.69 15.28
13. Rice 29.11 25.12 11.69
14. Texas-El Paso 18.80 22.10 11.76

FAU beat Marshall 30-25 to stay on top of the East with a 5-0 record, the only team left unbeaten in C-USA play. The Burrowing Owls have to play at Louisiana Tech before their final showdown for the East title against neighbor FIU. In the west, North Texas has all but wrapped up the West crown with only the two worst teams in the division, UTEP and Rice, left on the C-USA schedule. The Mean Green have managed to tightrope their way through C-USA play, going 5-1 despite being outscored by nine points in those six games.

This week is Homecoming for Rice, although it's not likely to end on a happy note. The opponent is Southern Miss, who beat the Owls 65-10 the last time it was the Homecoming opponent two years ago. The numbers don't suggest a different outcome of the game this year. The Golden Eagles have shown themselves to be vulnerable this year, registering sub-30 games at home against UNT and UAB, but overall their numbers are higher than Rice's. A U-test shows a significant difference at the 5% level, so it looks like the Owls' chance for another win this season will have to wait for next week.

(10-28-2017 02:03 AM)WRCisforgotten79 Wrote:  With C-USA teams, Massey is an outlier when compared with other computer ratings, such as mine. You'd be better served by using Massey's composite rankings.

Making the calculations that I do requires a system which can produce an expected margin of victory in a matchup between two teams. The Massey composite is a simple ordinal ranking incapable of doing this.

Rice's 43-34 loss to Southern Miss produced a Massey performance rating of 30.40. While the offense had its highest output of the season, the defense failed to make even one stop of the Golden Eagle offense (the only drive in the game USM didn't score on was its final one, when it ran out the clock). Here's how the season has gone so far:

vs.Stanford: 9.40
@Texas El-Paso: 40.58
@Houston: 25.45
Florida International: 33.84
@Pittsburgh: 29.12
Army: 14.61
@Texas-San Antonio: 30.80
Louisiana Tech: 28.50
@Alabama-Birmingham: 12.97
Southern Mississippi: 30.40

Here are the numbers for all Conference USA members:

Team, Median, Mean, Standard Deviation:
1. Florida Atlantic 62.89 60.83 14.23
2. Marshall 50.78 53.47 13.33
3. North Texas 46.31 45.72 12.00
4. Southern Mississippi 45.16 42.92 10.71
5. Middle Tennessee 43.61 43.33 14.76
6. Alabama-Birmingham 43.35 40.64 14.07
7. Louisiana Tech 43.07 43.04 10.05
8. Western Kentucky 40.59 40.39 9.27
9. Texas-San Antonio 40.58 43.83 8.38
10. Old Dominion 37.20 36.46 11.26
11. Florida International 32.14 38.69 14.62
12. Charlotte 32.06 28.92 12.75
13. Rice 28.81 25.77 10.03
14. Texas-El Paso 18.84 21.38 12.05

The division leaders Florida Atlantic and North Texas both won their games handily this past week. The two teams have three C-USA games left, all against the bottom four teams in the rankings, and appear to be on course for a rematch of their regular-season matchup in the C-USA championship game.

This week, Rice travels to Old Dominion. Unlike the past few games, the numbers say that the Owls have a reasonable chance of winning this one. The Monarchs have won their last two games by a combined 13 points, but before that they had lost six in a row, four of them by 30 points or more. A U-test on the data shows that, while ODU is favored, the difference is not significant at the 5% level. So, unlike the last few games, it appears that Rice actually has a chance in this one.

I didn't have time to post during the week of Thanksgiving, so I'll cover the last two games now. Rice's last-minute 24-21 loss to Old Dominion produced a Massey performance rating of 33.96, which is actually the second-highest of the season, while the 30-16 loss produced a more typical 24.21 rating. Here's how the season has gone so far:

vs.Stanford: 8.28
@Texas El-Paso: 37.42
@Houston: 23.15
Florida International: 30.25
@Pittsburgh: 29.88
Army: 10.88
@Texas-San Antonio: 28.47
Louisiana Tech: 26.28
@Alabama-Birmingham: 10.38
Southern Mississippi: 28.73
@Old Dominion: 33.96
North Texas: 24.21

Here are the numbers for all Conference USA members:

Team, Median, Mean, Standard Deviation:
1. Florida Atlantic 60.63 57.12 13.69
2. Marshall 45.68 47.51 12.72
3. North Texas 44.90 43.19 10.56
4. Southern Mississippi 43.50 42.83 13.11
5. Louisiana Tech 41.04 40.83 9.52
6. Middle Tennessee 40.10 40.88 15.09
7. Alabama-Birmingham 38.66 36.85 14.04
8. Texas-San Antonio 37.42 40.74 8.40
9. Western Kentucky 36.18 35.06 9.51
10. Old Dominion 29.47 31.04 12.77
11. Charlotte 28.29 23.34 14.11
12. Florida International 27.63 35.95 15.17
13. Rice 27.37 24.32 9.56
14. Texas-El Paso 18.66 18.81 10.68

The biggest takeaway from these nearly-final season numbers was that C-USA was a big cesspool of mediocrity. Aside from FAU, there were no really good teams in the conference, so they kept beating up on each other (and all beating up on Charlotte, Rice, and UTEP), and that's why the conference has no less than ten bowl-eligible teams. Looking at the numbers, FIU might be the worst team to go bowling in NCAA history and will probably be destroyed by any opponent; the rest of the teams have chances if they play lower Group of 5 teams.

I'll do a final season wrapup after the conference championship and bowl games have been played.

Rice closes out 2017 as #2 from the bottom in Massey and also #2 on ESPN's Bottom 10. Wonder if this is the worst performance by someone with a decade tenure .. (Mike Price might be in the running too)

http://www.espn.com/college-football/sto...fifth-spot

At least the future is finally looking up.