Post by SRT4driver on Dec 2, 2022 16:12:08 GMT -6
As Cactus Panther and possibly others have mentioned, most college basketball computer rating systems include a preseason weighting for all teams, based on several variables, such as results from past seasons (many use a team's final rating/strength for the past 5 years, or weight a team based on conference strength in previous seasons). These biases last for an arbitrary length of the season, depending on the system. This arbitrarily boosts some teams, mainly "Power" (Money) conference teams, and suppresses teams that weren't predicted to do well based on past results and/or conference affiliation.
In recent years in these systems that use conference historical strength for purposes of teams' pre-season weightings, this has become much more noticeable for HL teams such as Wright State and NKU, with them becoming more and more undervalued despite their results.
There are however several systems that base their ratings/rankings purely on merit, i.e. only this season's results and margins of victory, and subsequent SOS based on that. Right now those few systems are easy to spot, as MKE and other teams with good to great records (and only losses to good to great teams), are all much higher than those with arbitrary, often incorrect pre-season weights affecting them.
For anyone interested, here's my long-time favorite of those "unbiased" websites, and their explanation of how their rating are determined. Ironically, this system currently gives MKE their highest ranking of any system (102 of 363 teams, up from 193 last week) from the 35 most utilized CBB rating systems around.
www.allmysportsteamssuck.com/computer-ranking-methodology/
Here's a matrix of all 35 ranking/rating systems out there:
masseyratings.com/cb/compare.htm
In recent years in these systems that use conference historical strength for purposes of teams' pre-season weightings, this has become much more noticeable for HL teams such as Wright State and NKU, with them becoming more and more undervalued despite their results.
There are however several systems that base their ratings/rankings purely on merit, i.e. only this season's results and margins of victory, and subsequent SOS based on that. Right now those few systems are easy to spot, as MKE and other teams with good to great records (and only losses to good to great teams), are all much higher than those with arbitrary, often incorrect pre-season weights affecting them.
For anyone interested, here's my long-time favorite of those "unbiased" websites, and their explanation of how their rating are determined. Ironically, this system currently gives MKE their highest ranking of any system (102 of 363 teams, up from 193 last week) from the 35 most utilized CBB rating systems around.
www.allmysportsteamssuck.com/computer-ranking-methodology/
Here's a matrix of all 35 ranking/rating systems out there:
masseyratings.com/cb/compare.htm
