PDA

View Full Version : Rpi



Outraged
12-21-2017, 09:49 PM
Well we're second in rpi for the WCC before the loss. I think the tournament isn't a sure thing we have 4 good teams in the conference and a bunch of North Dakota's. BYU is Higher and Saint Marie's is getting their.

Edited to buy a vowel

GoZags
12-21-2017, 10:08 PM
Zags dropped from 12 to 13 in KenPom after the loss.

Bogozags
12-22-2017, 05:21 AM
Zags dropped from 12 to 13 in KenPom after the loss.

Goes to show that KenPom realizes this was a fluke loss...a game we could easily have won if we would have executed on offence...tough loss

jazzdelmar
12-22-2017, 05:36 AM
Is there a Ken Pom Eyeball Test?

509er
12-22-2017, 06:08 AM
Zags dropped from 12 to 13 in KenPom after the loss.

But RPI is 53. Why the big discrepancy? I see some teams are similar in RPI and Kenpom and some are not even close like GU and Arizona St (higher RPI). Doesn't the selection committee way RPI higher than KenPom?

cbbfanatic
12-22-2017, 08:16 AM
But RPI is 53. Why the big discrepancy? I see some teams are similar in RPI and Kenpom and some are not even close like GU and Arizona St (higher RPI). Doesn't the selection committee way RPI higher than KenPom?

rpi is straight wins/losses/sos/opp sos

kenpom factors in things like margin of victory and all that.

some level of value to both, i'd lean more towards kenpom early in the season and rpi late

Ezag
12-22-2017, 08:18 AM
The eye test says Perk needs rest from time to time and a back up not name Melson.

mgadfly
12-22-2017, 08:48 AM
Ken Pom is evaluating every possession and what we do with it. A 2 point loss on the road against a solid opponent means we did somethings right and somethings wrong (even if it was ugly).

I don't believe the committee is supposed to use RPI more than other metrics, but how it usually works is if a major conference team with political support in the committee has a good RPI, it is used to rationalize giving it an at-large bid. If the same team has a bad RPI (50+) but a good ranking in any of the other metrics used, RPI is dismissed and the other metrics are used to get that team in the field. And if none of the metrics support including the team from the big conference over a St. Mary's, Colorado State, the Bonnys, or whatever non-power conference, then they throw it all out and use the eye test.

Hopefully, GU wins the Conference tournament or has a brand name that whoever is on the committee is looking for a reason to get them in the field, not exclude them (because there are enough metrics to use to exclude almost anyone while including 17-15 70 RPI Syracuses of the basketball world).

cbbfanatic
12-22-2017, 09:06 AM
Ken Pom is evaluating every possession and what we do with it. A 2 point loss on the road against a solid opponent means we did somethings right and somethings wrong (even if it was ugly).

I don't believe the committee is supposed to use RPI more than other metrics, but how it usually works is if a major conference team with political support in the committee has a good RPI, it is used to rationalize giving it an at-large bid. If the same team has a bad RPI (50+) but a good ranking in any of the other metrics used, RPI is dismissed and the other metrics are used to get that team in the field. And if none of the metrics support including the team from the big conference over a St. Mary's, Colorado State, the Bonnys, or whatever non-power conference, then they throw it all out and use the eye test.

Hopefully, GU wins the Conference tournament or has a brand name that whoever is on the committee is looking for a reason to get them in the field, not exclude them (because there are enough metrics to use to exclude almost anyone while including 17-15 70 RPI Syracuses of the basketball world).

the complex is strong with this one

mgadfly
12-22-2017, 01:04 PM
the complex is strong with this one

If that's a conspiracy jab, I am just telling you how it has been historically. Maybe everything has been fixed and will be perfectly fair from here on out.


A few years ago SMC had a slightly better resume than the Zags going into the conference tournament final. A bunch of people looking at the metrics said either team was in regardless of win or loss. It was fiercely debated on this board about whether GU had to win. We did. SMC did not make the tournament despit 27 wins and a solid RPI. The NCAA changed the criteria for this season basically because St. Bonaventure was left out despite much better metrics than a half dozen major conference teams. Colorado State had the highest RPI without getting in ever. That was the same year that multiple 60+ RPI teams from major conferences made the field as at large teams. The number of at large births for mid-major or low-major conferences decreased almost ever year for a decade (I think from 14 to 3).

That's what happened. I don't know if there is going to be more at-large births given to teams from mid-majors, but so far having better RPIs, Ken Pom or Sagarins has guaranteed absolutely nothing.

CDC84
12-22-2017, 01:06 PM
Again, every expert will tell you that the RPI means absolutely NOTHING until one round of league play has taken place. I don't even look at it until mid-Feb. or so.

Pleasant Peninsula
12-22-2017, 01:26 PM
The number of at large births for mid-major or low-major conferences decreased almost ever year for a decade (I think from 14 to 3).



To be fair, conference realignment is probably the major factor in that, rather than some shift in the committee's preferences. Almost all the "mid-major" teams from a decade ago that were perennial tournament contenders have been swallowed up by a "major" conference, save for our Zags.

Having said that, RPI is a junk stat and there are some indications that the committee is is relying on it less and less, though they don't seem to be replacing it with KenPom though I wish they would (especially for seeding... last year's Wichita State seed a prime example of their failing to look at objective measures of team quality, such that you have two evenly matched, top 10 teams -- Kentucky and Wichita State lined up in the second round in a 2-10 seed matchup. So dumb).

Also worth noting, while St. Mary's had a much higher RPI than Gonzaga in 15-16, KenPom still liked the Gonzaga considerably more even before the tournament -- losing a bunch of close games to good teams, as Gonzaga did with virtually all of their top games that year doesn't hurt your KenPom, but crushes your RPI.

But as anecdotal evidence between the two metrics, however, I present you with the tournament that year, in which #11 seed Zags, who yes probably wouldn't have made it into the dance without the auto-bid despite their nice KP numbers, took both Seton Hall and Utah to the woodshed, but moreover, were favored by the oddsmakers to win both of those games, because the oddsmakers know which metrics to look at (KenPom) and which to ignore (RPI, polls, etc). They were also favored in their third game vs. Syracuse (first time a double digit seed has been favored in their first three games R64, R32, s16), but I don't want to think about that game right now (or I'll just be angrier at Perkins than I already am right now lol).

mgadfly
12-22-2017, 09:35 PM
To be fair, conference realignment is probably the major factor in that, rather than some shift in the committee's preferences.

I totally agree with almost everything you wrote except the above sentence.

At the beginning of the decade major conference teams made up 21.0% of the Division 1 field. Last season they made up 21.4% of the field. They expanded from 73 teams in 2010 to 75 teams in 2017. The NCAA also added three spots in the field in 2011. So while the total number of mid or low-major teams assigned at large bids continues to drop, it has nothing to do with major conference expansion as they make up virtually the same percentage of the field and more spots have been added to the tournament than teams to all the major conferences combined. They increased their proportion of the field by .4% while increasing their percentage of at-large bids from 76.4% to 86.4%.

And last year they assigned 2 at large bids to the A10 (which does very well with the committee) and 2 to the entire rest of the non-major field.

So while I think the advanced metrics certainly do a much better job at predicting and measuring teams' ability (just compare them to vegas odds who have money riding on their spreads), there has been no indication by the Committee that they are actually going to value non-major schools in the same way advanced metrics have.