Massey Ratings

[FanFromMaine]

Massey Ratings for basketball have been posted although it looks like data are still being added.

It was interesting to me to watch Massey and Net at the end of last year as UConn exploded. My recollection is that neither system rated the Huskies #1 until after the final game.

 

[WiseWillie]

Although Massey and NET can provide a useful verification and a different perspective, most basketball junkies today according to those who know these things use KenPom as their primary analytical source. Sometimes a composite of all three averaged together can produce an overall balanced metric - although there is never a system used solely or in a composite that will be infallible which in some respects creates the uncertainty and is what keeps b-ball sports fans interested and forums like the Boneyard going.

 

[EricLA]

I will say this - according to the numbers THEY USED, UConn ended up a 2 seed, but it goes to show the metrics were awful. They need to fix that IMHO. There is NO REASON UConn should have been a 2 seed.

Again, I get that we got it based on the actual metrics and data they used, so I'm not disputing that, but it was flawed and wrong. They should use last season to rework the numbers and figure out how to improve their system.

 

[azfan]

Massey pre season

Grand Canyon 87

ASU 88

NAU 112

Kitties 43 (shocking but who knows about the new roster)

And.............

The fashionista's Mustangs 105

It'll be interesting to see how the seasons unfold for these schools

 

[Bone Dog]

Am I reading this right -- ND over UCLA? That seems strange.

 

[FanFromMaine]

The Massey data seems very incomplete.

 

[FanFromMaine]

I will say this - according to the numbers THEY USED, UConn ended up a 2 seed, but it goes to show the metrics were awful. They need to fix that IMHO. There is NO REASON UConn should have been a 2 seed.

Again, I get that we got it based on the actual metrics and data they used, so I'm not disputing that, but it was flawed and wrong. They should use last season to rework the numbers and figure out how to improve their system.

Last year the existing data took a big hit after the first South Carolina loss to UConn. I’m not a statistician but an outlier of that proportion must be very hard to integrate in the last part of the season.

 

[ucmiami2]

Massey Ratings for basketball have been posted although it looks like data are still being added.

It was interesting to me to watch Massey and Net at the end of last year as UConn exploded. My recollection is that neither system rated the Huskies #1 until after the final game.

Pretty sure Massey ratings require fresh games to change, and the ratings in the first month of the season are still reflecting some of the previous year's data.

 

[Dogdude]

Meh, polls are good for giggles and kicks...I'm not sure how any analysis could have kept up with UConn's stunning improvement in the final weeks of the season and the tournaments. Based on their mid-season results it would be perfectly understandable to not buy into their new energy, and think that their inconsistency would result in a crash and burn. We the faithful saw it happening. Maybe at the end of the season I'll check out what the polls say.

 

[GoChips20]

IMO there’s some ambiguity over when the previous year’s results are no longer considered though because Iowa was a top-10 Massey offense all last year despite averaging 74.4ppg (not a standout number and almost 20 points less than Clark’s senior season). Iowa was still a good team but I didn’t understand that rating at all.

 

[FanFromMaine]

Although Massey and NET can provide a useful verification and a different perspective, most basketball junkies today according to those who know these things use KenPom as their primary analytical source. Sometimes a composite of all three averaged together can produce an overall balanced metric - although there is never a system used solely or in a composite that will be infallible which in some respects creates the uncertainty and is what keeps b-ball sports fans interested and forums like the Boneyard going.

I can’t find a KenPom for women. Does it exist?

 

[Twolf]

Rankings, seedings, and statistics are just like bikinis- they show a lot but they don’t show everything.

 

[Alydar]

The committee made it plain last season that in seeding the top 16 teams the key stat would be Tier 1 record. They use the entire season, not how a team is playing in February. As long as this is the system then UConn will never be a lock for a one seed unless they go undefeated.

And UConn fans shouldn't care. If we think the Huskies are the best then we shouldn't be concerned with seeding. Brag that we beat 3 one seeds to win the title. The men did that one year. Now the women have done it.

 

[Veteran]

The only seeding that matters is the final one the day after the championship game. The rest are basically just for entertainment value, and these days, for clicks. Yes, they're used for placement in the tournament, but no matter the seeding, somebody still has to win 6 games, and if they're good enough to win 6 games then the seeding doesn't really matter. Just ask our overall #7 seeded team if seedings mattered.

The fact that a 12th banner will soon be hanging in Gampel proves that seeding are basically just opinions. And everybody has probably heard the old saying about opinions. Of course, my comment is just an opinion, too, so the old saying also applies to me.

 

[Bone Dog]

And UConn fans shouldn't care. If we think the Huskies are the best then we shouldn't be concerned with seeding. Brag that we beat 3 one seeds to win the title. The men did that one year. Now the women have done it.

I agree. The difference between a 1 seed and a 2 seed isn't that important.

 

[Plebe]

I will say this - according to the numbers THEY USED, UConn ended up a 2 seed, but it goes to show the metrics were awful. They need to fix that IMHO. There is NO REASON UConn should have been a 2 seed.

Again, I get that we got it based on the actual metrics and data they used, so I'm not disputing that, but it was flawed and wrong. They should use last season to rework the numbers and figure out how to improve their system.

I don't agree. UConn got a 2 seed because our resume wasn't as strong as four other teams. Until mid-February, we were likely headed for a 3 seed. The win over SC helped a lot but not quite enough to put us in the top 4. For three-quarters of the regular season we just weren't particularly impressive. It wasn't flawed metrics that made us play poorly in key games until then.

Ultimately it comes down to peaking at the right time. In some seasons UConn has been an undisputed overall #1 seed but we failed to peak in the tournament. I'm thrilled that this season, after a horrible loss at Tennessee, we managed to flip the script in dramatic fashion.

 

[TheFarmFan]

I don't agree. UConn got a 2 seed because our resume wasn't as strong as four other teams. Until mid-February, we were likely headed for a 3 seed. The win over SC helped a lot but not quite enough to put us in the top 4. For three-quarters of the regular season we just weren't particularly impressive. It wasn't flawed metrics that made us play poorly in key games until then.

Ultimately it comes down to peaking at the right time. In some seasons UConn has been an undisputed overall #1 seed but we failed to peak in the tournament. I'm thrilled that this season, after a horrible loss at Tennessee, we managed to flip the script in dramatic fashion.

+1. All games count equally for your W-L and your relative rating, but only the last game of the season matters for a championship! Way to peak just when it counted - honestly, I think one of Geno's very best coaching efforts in many moons.

 

[HopJim]

I agree. The difference between a 1 seed and a 2 seed isn't that important.

Most ratings are based on the entire season, not the last 2-4 weeks. Vegas, which focuses on the last few games and other factors (injuries, coaching, etc.) had UConn favored, that's why I had a good feeling going into the NCAA's. Based on Massey's published metrics, for most of the year, they had UConn as a #2. For how UConn played for about 75% of the year, this may have been correct.

 

[Alydar]

I don't agree. UConn got a 2 seed because our resume wasn't as strong as four other teams. Until mid-February, we were likely headed for a 3 seed. The win over SC helped a lot but not quite enough to put us in the top 4. For three-quarters of the regular season we just weren't particularly impressive. It wasn't flawed metrics that made us play poorly in key games until then.

Ultimately it comes down to peaking at the right time. In some seasons UConn has been an undisputed overall #1 seed but we failed to peak in the tournament. I'm thrilled that this season, after a horrible loss at Tennessee, we managed to flip the script in dramatic fashion.

You may be right about the seed but not because of metrics. UConn was No 1 in NET rating from January on. NET measures the quality of wins AND losses and it is the NCAA's primary metric measurement. Up until last season every NET No 1 team has been a 1 seed. Last season the NCAA chose to seed the top 16 based on Tier 1 wins. I'm sure that the NCAA wasn't swayed in any way by the P-4's dominance of the list of Tier 1 teams.

 

[Plebe]

You may be right about the seed but not because of metrics. UConn was No 1 in NET rating from January on. NET measures the quality of wins AND losses and it is the NCAA's primary metric measurement. Up until last season every NET No 1 team has been a 1 seed. Last season the NCAA chose to seed the top 16 based on Tier 1 wins. I'm sure that the NCAA wasn't swayed in any way by the P-4's dominance of the list of Tier 1 teams.

The NET is just one of many listed criteria the committee considers. Also you seem to have the dates wrong. UConn wasn't number 1 in the NET at any time in the month of January.

From time immemorial the committee has always placed a major emphasis on quality wins. The NET and the RPI before it were never intended to serve as a substitute for quality wins. The previous year Oregon at #19 in the NET didn't even get an at large bid, while West Virginia at #60 did get a bid.

 

[dtbtbtb]

Any chance UConn had for a 1 seed went away with the bad loss to TN but at the same time this is the loss that galvanized the team so all good.
I do think that Geno is the best in the game in how much the team improves during the season and ready for the tournaments.

 

[Skeets]

Meh, polls are good for giggles and kicks...I'm not sure how any analysis could have kept up with UConn's stunning improvement in the final weeks of the season and the tournaments. Based on their mid-season results it would be perfectly understandable to not buy into their new energy, and think that their inconsistency would result in a crash and burn. We the faithful saw it happening. Maybe at the end of the season I'll check out what the polls say.

DD, you just exposed one of the major flaws with metric based rating systems. Such systems do not know just from analyzing the numbers that a team is preparing to go into the toilet, or ready to explode as UConn did in the latter half of the season. I think at best these ratings can give some insight into teams that you know very little about, not the teams that you are intimately familiar with. For those teams eyeballs work the best.

 

[WiseWillie]

I can’t find a KenPom for women. Does it exist?

Ken Pomeroy the developer of the KenPom website developed the site in 2002 apparently only for Div I mens basketball. At present there is no indication whether with the popularity of womens basketball KenPom will be expanded to include Div I WBB.
For another source other than KenPom, Massey and NET for WBB, HerHoops Stats uses metrics similar to KenPom.