Well at least can we get under 100?
Quote from: gepsguys on January 05, 2016, 08:47:22 PM
Well at least can we get under 100?
Close, but probably not.
Quote from: gepsguys on January 05, 2016, 08:47:22 PM
Well at least can we get under 100?
edited: to reflect 140, instead of 40 that was posted
edited again: Back to 40 in case that was you guess. I don't really know what you were saying. But our current RPI is 140 per real time rpi
What?
Quote from: gepsguys on January 05, 2016, 08:47:22 PM
Well at least can we get under 100?
edited: to reflect 140, instead of 40 that was posted
edited again: Back to 40 in case that was you guess. I don't really know what you were saying. But our current RPI is 140 per real time rpi
Calculus makes more sense at first glance then you.
Quote from: wadesworld on January 05, 2016, 09:12:43 PM
What?
To clarify, the edits were mine. I thought he was posting our current, which is 140 (and accidentally posted 40), but then I re-though and maybe is guessing at 40. I just don't know. I'll remove my comments from his post to clarify.
Quote from: theburreffect2 on January 05, 2016, 09:15:08 PM
Calculus makes more sense at first glance then you.
Haha.
Quote from: rocky_warrior on January 05, 2016, 09:16:13 PM
To clarify, the edits were mine. I thought he was posting our current, which is 140 (and accidentally posted 40), but then I re-though and maybe is guessing at 40. I just don't know. I'll remove my comments from his post.
Gotcha. Thanks for the clarification.
To clarify, was fantasizing about a rise to 40. But could be pushing the 90s?
Quote from: gepsguys on January 05, 2016, 09:26:15 PM
To clarify, was fantasizing about a rise to 40. But could be pushing the 90s?
LiveRPI has us at 101 currently.
http://espn.go.com/mens-college-basketball/rpi/_/page/1/sort/RPI
92
Quote from: RKMU123 on January 06, 2016, 09:19:32 AM
http://espn.go.com/mens-college-basketball/rpi/_/page/1/sort/RPI
92
Never underestimate the effect that the combination of a) a road win and b) conference play has on your RPI in January.
Quote from: Benny B on January 06, 2016, 11:46:45 AM
Never underestimate the effect that the combination of a) a road win and b) conference play has on your RPI in January.
Yup. A lot of those teams right above us are going to fall really far, regardless of the outcome of their upcoming games.
Quote from: wadesworld on January 06, 2016, 11:54:34 AM
Yup. A lot of those teams right above us are going to fall really far, regardless of the outcome of their upcoming games.
Exactly. This is why RPI sucks at this point in the season... just above MU in the rankings are the likes of LBSU (at 4-9), Evansville (with a SOS nearly double MU's), and TAMU-CC (with an adjusted scoring margin of -6.0).
Here's a nice little tidbit...
MU's non-conference RPI is 30. For all the bitching and whining we've endured over the past few weeks about playing too many cupcakes, losing to Belmont/Iowa, ASU/LSU/Wis not being good wins, etc.... all I can say is
#30.
Quote from: Benny B on January 06, 2016, 12:01:50 PM
Exactly. This is why RPI sucks at this point in the season... just above MU in the rankings are the likes of LBSU (at 4-9), Evansville (with a SOS nearly double MU's), and TAMU-CC (with an adjusted scoring margin of -6.0).
Here's a nice little tidbit... MU's non-conference RPI is 30. For all the bitching and whining we've endured over the past few weeks about playing too many cupcakes, losing to Belmont/Iowa, ASU/LSU/Wis not being good wins, etc.... all I can say is #30.
Wait, what? How is MU's non-con RPI 30, but overall RPI in the 90-100 range after adding games again Seton Hall, Georgetown and Providence in conference? Please explain.
To clarify, not saying I don't think RPI is most irrelevant at this stage of the season, but still curious how that math works.
Quote from: JamilJaeJamailJrJuan on January 06, 2016, 12:58:33 PM
Wait, what? How is MU's non-con RPI 30, but overall RPI in the 90-100 range after adding games again Seton Hall, Georgetown and Providence in conference? Please explain.
To clarify, not saying I don't think RPI is most irrelevant at this stage of the season, but still curious how that math works.
Not sure about all the numbers, but if true it might be because we were 10-2 nonconference, and are currently 1-2 in conference. SOS means a lot, but you still have to win. And our nonconference SOS might be improving with LSU's win over UK last night and Iowa's recent wins over MSU and Purdue.
Quote from: JamilJaeJamailJrJuan on January 06, 2016, 12:58:33 PM
Wait, what? How is MU's non-con RPI 30, but overall RPI in the 90-100 range after adding games again Seton Hall, Georgetown and Providence in conference? Please explain.
To clarify, not saying I don't think RPI is most irrelevant at this stage of the season, but still curious how that math works.
Perhaps JB can jump in here, but non-con RPI excludes conference teams from an individual team's RPI calculation... so SHU, GU and PC would obviously be ignored in calculating MU's NCRPI, but when you calculate your opponent's win percentage (and opponent's opponent's) that make up the other 75% of the calculation, you're also excluding wins/losses your opponents and opponent's opponents have against Big East teams from the calculation. So in MU's case for example, when calculating Becky's win percentage, you exclude their loss to Georgetown; likewise, when calculating Purdue's win percentage (who played IUPUI earlier in the year), you exclude their loss to Butler.
As you can see, the stat is heavily weighted by how your own conference has performed in non-conference play, i.e. the more wins your conference has, the more losses that are being excluded from your team's RPI calculation. In that regard, it won't necessarily define, but it does give an idea of what a team's RPI "potential" is.... not accurately, but a heck of a lot better than anything you could get from RPIforecast.com or RPIwizard.com.
Quote from: GooooMarquette on January 06, 2016, 01:47:53 PM
Not sure about all the numbers, but if true it might be because we were 10-2 nonconference, and are currently 1-2 in conference. SOS means a lot, but you still have to win. And our nonconference SOS might be improving with LSU's win over UK last night and Iowa's recent wins over MSU and Purdue.
Technically, the outcome of the LSU/UK game has no effect on LSU's strength of schedule... LSU's SOS improved when they put five players on the court for tipoff last night.
Quote from: Benny B on January 06, 2016, 02:01:24 PM
Technically, the outcome of the LSU/UK game has no effect on LSU's strength of schedule... LSU's SOS improved when they put five players on the court for tipoff last night.
I meant that it helped
MU's SOS when LSU beat UK.
Quote from: Benny B on January 06, 2016, 12:01:50 PM
Exactly. This is why RPI sucks at this point in the season... just above MU in the rankings are the likes of LBSU (at 4-9), Evansville (with a SOS nearly double MU's), and TAMU-CC (with an adjusted scoring margin of -6.0).
Here's a nice little tidbit... MU's non-conference RPI is 30. For all the bitching and whining we've endured over the past few weeks about playing too many cupcakes, losing to Belmont/Iowa, ASU/LSU/Wis not being good wins, etc.... all I can say is #30.
Source?
RPIForecast.com (http://www.rpiforecast.com/teams/Marquette.html) has Marquette's non-conference RPI as 143. Usually that number will be pretty static from this point, unless multiple teams we already played end up doing much better (or worse) than expected.
Quote from: brewcity77 on January 06, 2016, 02:07:14 PM
Source?
Some little Mickey Mouse sports website called ESPN.com.
RPI Forecast (et al) are going to have much lower numbers because - presumably - their RPI calculations simply exclude conference results from the team's W/L percentage, not the OWP or OOWP like ESPN does.
BTW - You're looking at
expected NCRPI, not
actual NCRPI.
That aside, I've picked up on a few discrepancies in RPIForecast's/RPIWizard's calculations which suggests they're using obsolete (or bad) data and/or are butchering their algorithms.
Quote from: Benny B on January 06, 2016, 02:25:11 PM
Some little Mickey Mouse sports website called ESPN.com.
Did it have pictures of Ben Simmons plastered all over it?
Quote from: MomofMUltiples on January 06, 2016, 02:32:48 PM
Did it have pictures of Ben Simmons plastered all over it?
Just a couple dozen. Not thousands.
Quote from: Benny B on January 06, 2016, 02:25:11 PM
Some little Mickey Mouse sports website called ESPN.com.
RPI Forecast (et al) are going to have much lower numbers because - presumably - their RPI calculations simply exclude conference results from the team's W/L percentage, not the OWP or OOWP like ESPN does.
BTW - You're looking at expected NCRPI, not actual NCRPI.
That aside, I've picked up on a few discrepancies in RPIForecast's/RPIWizard's calculations which suggests they're using obsolete (or bad) data and/or are butchering their algorithms.
Generally, I am not a fan of using ESPN for any rankings because they always create their own crap that no one else gives a hoot about. Part of their marketing model, I guess.
And yes, I'd always rather look at expected non-con RPI because it's far more meaningful to consider the full season schedule. A snapshot looking at now is useless because teams like Grambling, Jackson State, and other cupcakes will traditionally have a much tougher SOS that would artificially boost their non-con RPI early in the season (as they presumably play a ton of teams with winning records, whereas come season's end, most of their games will be against teams with losing records).
Yes, I realize that means taking predictions into account, but usually those predictions balance out pretty well. Sure, if Maine, San Jose State, Jackson State, Chicago State, IUPUI, Belmont, and Presbyterian all win their conferences it could change radically, but I'm going to go out on a limb and say that things will balance out, the way they tend to do every single year.
EDIT: Also, Warren Nolan (http://warrennolan.com/basketball/2016/nitty-live) has our non-con RPI at 129. Sorry, but no way I buy us at #30 no matter who's selling. You don't get an RPI of 30 with the schedule we played.
Quote from: brewcity77 on January 06, 2016, 02:46:57 PM
Generally, I am not a fan of using ESPN for any rankings because they always create their own crap that no one else gives a hoot about. Part of their marketing model, I guess.
And yes, I'd always rather look at expected non-con RPI because it's far more meaningful to consider the full season schedule. A snapshot looking at now is useless because teams like Grambling, Jackson State, and other cupcakes will traditionally have a much tougher SOS that would artificially boost their non-con RPI early in the season (as they presumably play a ton of teams with winning records, whereas come season's end, most of their games will be against teams with losing records).
Yes, I realize that means taking predictions into account, but usually those predictions balance out pretty well. Sure, if Maine, San Jose State, Jackson State, Chicago State, IUPUI, Belmont, and Presbyterian all win their conferences it could change radically, but I'm going to go out on a limb and say that things will balance out, the way they tend to do every single year.
EDIT: Also, Warren Nolan (http://warrennolan.com/basketball/2016/nitty-live) has our non-con RPI at 129. Sorry, but no way I buy us at #30 no matter who's selling. You don't get an RPI of 30 with the schedule we played.
I highly doubt that the selection committee calculates non-conference RPI simply be removing conference teams from 1/4 of the RPI equation.
Quote from: Benny B on January 06, 2016, 02:55:28 PM
I highly doubt that the selection committee calculates non-conference RPI simply be removing conference teams from 1/4 of the RPI equation.
I don't think I asserted that remotely... ?-(
Quote from: Benny B on January 06, 2016, 02:55:28 PM
I highly doubt that the selection committee calculates non-conference RPI simply be removing conference teams from 1/4 of the RPI equation.
But they might use the NCAA's calculations, which doesn't show a Non-con, but does only have us at 140...
http://www.ncaa.com/rankings/basketball-men/d1/ncaa-mens-basketball-rpi
Curious how our non-con RPI is 30 despite going 10-2 against the #287 SOS while SMU played the #91 SOS and is undefeated yet only manages a non-con RPI of 93.
I haven't sat down and crunched numbers, but something tells me ESPN is pulling something out of their ass (again). Guess this is the new TQBR or whatever other arbitrary rating they created to feel important about themselves.
Quote from: brewcity77 on January 06, 2016, 03:24:10 PM
I haven't sat down and crunched numbers, but something tells me ESPN is pulling something out of their ass (again). Guess this is the new TQBR or whatever other arbitrary rating they created to feel important about themselves.
Agreed, and unfortunately I think it's misleading to fans, because I don't think it corresponds at all to the "official" RPI that gets distributed to the committee...which currently shows Marquette as 140 overall (vs 92 ESPN).
Quote from: brewcity77 on January 06, 2016, 02:59:18 PM
I don't think I asserted that remotely... ?-(
No, I was the one who asserted that. RPIForecast says on it's explanations page "The OOC RPI is
just the RPI applied to non-conference opponents" (emphasis mine).
In other words, consider a scenario where Kansas, Kentucky, North Carolina, Maryland, Arizona and UCONN breeze through their NC schedules which consisted of mostly cupcake wins mixed in with a bunch of losses to Big East teams... yet when conference play starts, all go undefeated, win their respective conference titles and are in the top 15 RPI.
Now assume MU goes 10-2 in OOC against a schedule that includes all six of these teams... considering a) the losses by these six against other BE teams hurt MU's NCRPI - despite NCRPI being a
non-conference RPI and b) the success of these six during the conference season has no effect on MU's NCRPI, you can see the flaw in how RPIForecast is calculating NCRPI, no matter if it's a projected or actual rank you're looking at.
In other words, if you play a bunch of eventual 1-4 seeds in November and December yet they played crappy OOC schedules themselves (or have a bunch of losses to your conference mates), of course you're going to have a shiiiity NCRPI.
I agree that RPI calculations seem to vary amongst the various providers, but all seem to be consistent at the end of the season. But I'm pretty certain that the committee isn't seeing the NCRPI figures calculated the same way that all the other third parties are.