One way to quantify dominance

[Suebird10fan]

I discovered the other day, prompted by another thread, that Massey has historical WCBB ratings going back to 1998. I actually like Massey's ratings (the Power rating since that one takes into account scores, not just wins and losses) a bit better than Sagarin. Sagarin's predictive rating penalizes teams like UConn that win by huge margins. It probably works out ok in MCBB where you don't see the kind of gulf between the top team and the field that you sometimes see with the UConn women.

Anyway ... here's what I did. I looked at the Power Ratings and took the average of #2 through #6. I compared that to the rating for the top team. No magic in picking 2 through 5. But I wanted more than one team so that if #2 was unusually good it wouldn't distort the results too much. I might have done 2 through 10--just didn't want that much data entry!

This is by no means a definitive basis for comparing teams across the years but still pretty interesting. (Translation for all you DT fans: I am NOT using this to argue that the 2014 team was better than the 2002 team!!!)

Here are the results where the gap was greater than 10 points. (Note: the scale of the Power Rating is set so that the difference between two teams approximates what the point spread should be between the teams).

UConn 14 22.2
UConn 10 20.9
UConn 02 19.7
UConn 09 19.4
UConn 15 19.3
UConn 00 16.9
Baylor 12 12.8
Tenn 98 12.2
UConn 13 12.1

The one that took me by surprise was the 2000 team. I remember them being really, really good (as Geno would say) but didn't expect to see the result above. That said, they only lost one game and that one by one point. (At home vs Tennessee, 71-72.)

And of course I loved the fact that the top six teams are all Huskies!

 

[Sonny44]

I discovered the other day, prompted by another thread, that Massey has historical WCBB ratings going back to 1998. I actually like Massey's ratings (the Power rating since that one takes into account scores, not just wins and losses) a bit better than Sagarin. Sagarin's predictive rating penalizes teams like UConn that win by huge margins. It probably works out ok in MCBB where you don't see the kind of gulf between the top team and the field that you sometimes see with the UConn women.

Anyway ... here's what I did. I looked at the Power Ratings and took the average of #2 through #6. I compared that to the rating for the top team. No magic in picking 2 through 5. But I wanted more than one team so that if #2 was unusually good it wouldn't distort the results too much. I might have done 2 through 10--just didn't want that much data entry!

This is by no means a definitive basis for comparing teams across the years but still pretty interesting. (Translation for all you DT fans: I am NOT using this to argue that the 2014 team was better than the 2002 team!!!)

Here are the results where the gap was greater than 10 points. (Note: the scale of the Power Rating is set so that the difference between two teams approximates what the point spread should be between the teams).

UConn 14 22.2
UConn 10 20.9
UConn 02 19.7
UConn 09 19.4
UConn 15 19.3
UConn 00 16.9
Baylor 12 12.8
Tenn 98 12.2
UConn 13 12.1

The one that took me by surprise was the 2000 team. I remember them being really, really good (as Geno would say) but didn't expect to see the result above. That said, they only lost one game and that one by one point. (At home vs Tennessee, 71-72.)

And of course I loved the fact that the top six teams are all Huskies!

Something's wrong; we have only 7 of the top 9. Keep trying.

 

[UcMiami]

Fun - but it does not work as a comparison between years in terms of one team being 'better' as it is only comparative to the other teams that year. And because you are averaging the 2-6 teams it is also not distinguishing where the 'gap' falls in the top six - some years there are two or three comparable teams at the top and then a large gap, other years there is a single team at the top and perhaps no additional gap in teams 2-5. As an example:

Last year was a truly amazing year - Uconn PR = 86.25 #2ND PR = 70.87 a gap between them of 15.38 (good enough for #7 on your list of gaps averaged for positions #2-#6) - and the #6 team Baylor was 64.18 (a gap between #2 and #6 of only 6.69.

In 2014 - Uconn #1 at 90.54, ND#2 at 78.27 and then 3-6 at 67.04-63.97

In 2013 - Uconn #1 at 83.15, Baylor and ND at 78.82 and 76.03 and then 4-6 at 68.68-64.61

In 2012 - Baylor at #1 with 81.90 ND, Uconn, and Stanford from 74.35 to 71.39, and then 5-6 at 63.33 and 62.96.

Four very different competitive environments (The tightest bunching being the year Baylor was favored and lost before the FF.)

I am too lazy to do it, but it would be interesting to see the average power rating for the top teams over the Massey period - some teams obviously would have serious drop out years, but looking at Uconn, Stanford, TN, ND, Baylor, Duke, UNC, MD, OK for example.

 

[Suebird10fan]

I might take you up on that one UcMiami ...

 

[Kaizen]

How about the team in 1996? I guess Massey may not have enough data. Thanks for the post.

 

[DobbsRover2]

Fun - but it does not work as a comparison between years in terms of one team being 'better' as it is only comparative to the other teams that year. And because you are averaging the 2-6 teams it is also not distinguishing where the 'gap' falls in the top six - some years there are two or three comparable teams at the top and then a large gap, other years there is a single team at the top and perhaps no additional gap in teams 2-5. As an example:

Last year was a truly amazing year - Uconn PR = 86.25 #2ND PR = 70.87 a gap between them of 15.38 (good enough for #7 on your list of gaps averaged for positions #2-#6) - and the #6 team Baylor was 64.18 (a gap between #2 and #6 of only 6.69.

In 2014 - Uconn #1 at 90.54, ND#2 at 78.27 and then 3-6 at 67.04-63.97

In 2013 - Uconn #1 at 83.15, Baylor and ND at 78.82 and 76.03 and then 4-6 at 68.68-64.61

In 2012 - Baylor at #1 with 81.90 ND, Uconn, and Stanford from 74.35 to 71.39, and then 5-6 at 63.33 and 62.96.

Four very different competitive environments (The tightest bunching being the year Baylor was favored and lost before the FF.)

I am too lazy to do it, but it would be interesting to see the average power rating for the top teams over the Massey period - some teams obviously would have serious drop out years, but looking at Uconn, Stanford, TN, ND, Baylor, Duke, UNC, MD, OK for example.

Yeah, it's like the question of who has the toughest bracket every year. A #1 team may have a #2 seed that is recognized by far the toughest on the second seedings row but have overall a weak top 4. So if the #2 team gets bounced in the 2nd round, that NCAA slate that was making the fans of the #1 team apoplectic may turn into a bit of a cream puff. But I can well believe that the 2014 team might end up on top in this analysis scheme even though they faced an undefeated team in the NC, because some years are deeper than others at the top, and 2014 just wasn't.

 

[Waquoit]

This is nothing but Fun With Numbers.

 

[UcMiami]

Waquoit - but it is summer and what other fun is there?!

 

[DaddyChoc]

Renee, Tina and Maya was my favorite team followed by Bria & Dolson (sr. year)... 2002 is in the Top 5 but I liked 2004 better

 

[Suebird10fan]

I am too lazy to do it, but it would be interesting to see the average power rating for the top teams over the Massey period - some teams obviously would have serious drop out years, but looking at Uconn, Stanford, TN, ND, Baylor, Duke, UNC, MD, OK for example.

Ok, average Massey ratings for the entire period covered:

UConn 65.1
Tenn 56.4
Duke 52.2
Stfd 52.2
ND 51.1
Baylor 48.5

I thought Maryland would be next but they came in at 42.7. UNC or OK might be slightly ahead of that.

 

[UcMiami]

SueBird - excellent work! And a true sign of dominance that 8.7 gap back to TN. It would I am sure expand to 10 if you dropped out 1998 and 1999 and started with the 2000 season.
My money would be OKC coming in very close to Baylor - they have surprisingly maintained pretty high rankings in the face of serious injuries and very limited recruiting success. OSU might surprise as well and UNC probably as you note.

 

[Suebird10fan]

I don't have exact #s but some quick research indicates that after Top 5 above Baylor and UNC are extremely close. Then there is a gap and the next three, all extremely close, are Purdue, GA and LSU. Less sure after that but it looks like another gap followed by OK, OSU, Md, Lville, Rutg, Md bunched up.

 

[UcMiami]

And thanks again!
Had forgotten about LSUs strength in the middle of the period with their string of FF appearances - they have also had some pretty poor years, but the bonus points that distance teams from the pack at the top, and especially NC caliber teams are significant. The gap between teams becomes almost negligible the further down you go.

For example in 2015 :
#1 team: 86.26
#4: Team 67.51 - gap =18.75
#10 Team: 58.12 - gap = 9.39
#20 Team: 55.61 - gap = 2.51
#40 Team: 50.25 - gap = 5.36
#60 Team: 46.81 - gap = 3.44

Or 2013:
#1: 83.15
#4: 68.68 - gap = 14.47
#10: 61:02 - gap = 7.66
#20: 56.25 - gap = 4.77
#40: 50.44 - gap = 5.81
#60: 47.37 - gap = 3.07

For 2005 (An aberrational year with a 'surprise' champion)
#1: 68.45 (Stanford not the champion Baylor)
#4: 65.42 - gap = 3.03
#10: 60.08 - gap = 5.34
#20: 52.64 - gap = 7.44
#40: 48:40 - gap = 4.24
#60: 44.01 - gap = 4.39