The unofficial online home for fans of the Old Dominion University Monarchs. Part of Collegiate Sports Nation (CSN).

I realize that both use different metrics for ratings - however why is it that real-time ROI has us at#86 while KenPom has us at #56 ahead of WKU. What service is more acceptable and utilized ?

(02-20-2018 11:47 PM)Obiwan Wrote:  I realize that both use different metrics for ratings - however why is it that real-time ROI has us at#86 while KenPom has us at #56 ahead of WKU. What service is more acceptable and utilized ?

Kenpom to me.

Sent from my SM-N950U using CSNbbs mobile app

When comparing the two I like to look at the top 10 of each and ask, does that make sense?

RPI top 10:

Virginia, Villanova, Xavier, Duke, UNC, Kansas, Auburn, Rhode Island, Clemson, Nevada

Kenpom top 10:

Virginia, Villanova, Duke, Michigan State, Cinci, Purdue, UNC, Texas Tech, Gonzaga, Auburn.

Now let's look at the top 10 in the poll:

Virginia, Michigan State, Villanova, Xavier, Duke, Gonzaga, Texas Tech, Kansas, Purdue, UNC

I bolded the ones that don't belong in RPI/kenpom according to the poll. Kenpom has fewer, while they both share Auburn.

I realize this is a weird way to judge, but it's just an example of why I think kenpom makes more sense from a real life perspective.

There are a bunch of different metrics than can be used to compare teams. RPI and KPI are results based metrics.

The core part of RPI’s formula is ¼ team’s winning percentage, ½ opponents’ winning %, and ¼ opponents’ opponents’ winning %.

The KPI formula incorporates Opponent’s Winning Percentage, Opponent’s Strength of Schedule, the Game Result, Scoring Margin, Pace of Game, Location of Game (Home/Away/Neutral) and the Opponent’s KPI Ranking. The value for each game is divided by games played for the overall KPI Ranking.

KenPom, BPI, and Sagarin Ratings are predictive analytics.

BPI does everything that RPI does. Additionally, the BPI also incorporates (1) scoring margin, (2) diminishing returns for blowouts (a 30-point win is not necessarily twice as good as a 15-point win), (3) pace of game, (4) strength of schedule beyond opponents’ opponents’ wins and losses, and (5) de-weighting games with missing key players.

KenPom's rankings, as opposed to BPI, are purely predictive.Inputted into his formula are the adjusted offensive and defensive efficiencies of each college basketball team, which according to Pomeroy are “based purely on scoring per possession with no consideration of winning or losing”. On KenPom, the term ‘adjusted’ “refers to how a team would perform against average competition at a neutral site” with more weight on recent performances. Unlike RPI, KenPom does take into account scoring margin, pace of game, and strength of schedule beyond opponents’ opponents’ wins and losses – but it does not diminish returns for blowouts and de-weight games with missing key players like BPI does.

Sagarin's calculations are a bit of a mystery to me, but it's at least interesting enough for me to look at.

(02-21-2018 08:59 AM)ODUCoach Wrote:  There are a bunch of different metrics than can be used to compare teams. RPI and KPI are results based metrics.

The core part of RPI’s formula is ¼ team’s winning percentage, ½ opponents’ winning %, and ¼ opponents’ opponents’ winning %.

The KPI formula incorporates Opponent’s Winning Percentage, Opponent’s Strength of Schedule, the Game Result, Scoring Margin, Pace of Game, Location of Game (Home/Away/Neutral) and the Opponent’s KPI Ranking. The value for each game is divided by games played for the overall KPI Ranking.

KenPom, BPI, and Sagarin Ratings are predictive analytics.

BPI does everything that RPI does. Additionally, While the BPI also incorporates (1) scoring margin, (2) diminishing returns for blowouts (a 30-point win is not necessarily twice as good as a 15-point win), (3) pace of game, (4) strength of schedule beyond opponents’ opponents’ wins and losses, and (5) de-weighting games with missing key players.

KenPom's rankings, as opposed to BPI, are purely predictive.Inputted into his formula are the adjusted offensive and defensive efficiencies of each college basketball team, which according to Pomeroy are “based purely on scoring per possession with no consideration of winning or losing”. On KenPom, the term ‘adjusted’ “refers to how a team would perform against average competition at a neutral site” with more weight on recent performances. Unlike RPI, KenPom does take into account scoring margin, pace of game, and strength of schedule beyond opponents’ opponents’ wins and losses – but it does not diminish returns for blowouts and de-weight games with missing key players like BPI does.

Sagarin's calculations are a bit of a mystery to me, but it's at least interesting enough for me to look at.

Thanks for that explanation Coach!


Sent from my iPhone using Tapatalk

I'm not a big fan of using scoring margin as a metric, even though it helps us this year. Too often, the final score is not indicitive of how the game went. If you have a 20 point lead all game, which is never in doubt and the opponent hits a bunch of 3s in garbage time, it hurts your rating (think the first Marshall game). Conversely, if you have a 4 point game under a minute and the lead increases to double digit due to taking desperation 3s and FTs ar the other end. In this case, the winner gets credit for a blowout in a game that was in doubt with 1:00 to go.

In other words, our losses to MTSU and WKU look more decisive than our win against Marshall based on margin alone.

Here are the numbers for the Top 4 in the league:

MTSU
RPI: 23
KPI: 46
BPI: 47
Pom: 46
Sag: 54
AVG: 43.2

ODU
RPI: 85
KPI: 82
BPI: 52
Pom: 57
Sag: 66
AVG: 68.4

WKU
RPI: 58
KPI: 63
BPI: 62
Pom: 61
Sag: 82
AVG: 65.2

Marshall
RPI: 106
KPI: 119
BPI: 140
Pom: 131
Sag: 137
AVG: 126.6

I would normally say Kenpom, but I can't see any objective means of ranking ODU ahead of WKU at this point.

Maybe after this week, but not now.

(02-21-2018 09:33 AM)ODUBB35 Wrote:  I would normally say Kenpom, but I can't see any objective means of ranking ODU ahead of WKU at this point.

Maybe after this week, but not now.

Losing to UTSA for one even if it was on the road wasn't good

Sent from my SM-N950U using CSNbbs mobile app

(02-21-2018 10:06 AM)ODUMONARCHZ1 Wrote:  

(02-21-2018 09:33 AM)ODUBB35 Wrote:  I would normally say Kenpom, but I can't see any objective means of ranking ODU ahead of WKU at this point.

Maybe after this week, but not now.

Losing to UTSA for one even if it was on the road wasn't good

Sent from my SM-N950U using CSNbbs mobile app

Yeah, that has to be it.

UTSA will be scary in a few years if they can keep this team together.

(02-21-2018 09:33 AM)ODUBB35 Wrote:  I would normally say Kenpom, but I can't see any objective means of ranking ODU ahead of WKU at this point.

Maybe after this week, but not now.

Typically I would agree that head to head is a big difference maker. But to be Devil's advocate for one second. The year the Tampa Bay Bucs won the Super Bowl they lost twice in the regular season to the New Orleans Saints (Pre Brees) who finished the season below .500

(02-21-2018 10:20 AM)ODU True Blue Wrote:  

(02-21-2018 09:33 AM)ODUBB35 Wrote:  I would normally say Kenpom, but I can't see any objective means of ranking ODU ahead of WKU at this point.

Maybe after this week, but not now.

Typically I would agree that head to head is a big difference maker. But to be Devil's advocate for one second. The year the Tampa Bay Bucs won the Super Bowl they lost twice in the regular season to the New Orleans Saints (Pre Brees) who finished the season below .500

True.

But it's not just the head-to-head, but they have a much tougher schedule than we do this year, and just about an identical record. Add to that the head-to-head loss at home, they should be ranked higher.

Now if we win our next two, that may change.

(02-21-2018 10:18 AM)ODUBB35 Wrote:  UTSA will be scary in a few years if they can keep this team together.

Only if they at least pretend to play defense.

Sagarrin is crap. I have no idea how it got popular.

RPI is ok, but there are so many variables that it doesn't take into account and simply using win% is dumb.

Scoring margin should count even if there are some fluky things that can happen. Winning by 50 over UTEP is better than winning by 2; they shouldn't count the same.