Given the weight still given to RPI by the Committee and the amount of dollars flowing in and out of power conference programs based on tourney appearances, why don't universities employ statisticians (or do they) to--and I'm spit-balling here--run projections for all DI teams for the next year, pooling them into likely RPI spots (i.e. 1-50, 50-100, etc.) and then share that data with the program to schedule accordingly. It seems like it's just a numbers game and you keep on the schools from certain pools until your non-conference schedule is RPI proof.
When I run through some of these teams in the top 25-30 in RPI right now while we sit outside the top 100, it's appalling. We should never find ourselves in this position provided the resources bestowed upon our program (crappy RPI based in large part on scheduling not on-court play)
Curious to hear thoughts. Is this topic played out? Is MU considering this? Have they tried something similar?
Search around th board for this. Other people have done it much better justice than I will.
As I recall, it comes down to two big things:
1. EVERYONE wants to schedule teams with high RPIs a low chances of winning. That makes them expensive (either as a buy game or wanting a return visit) and not readily available.
2. Renting out a shared space like the BC means the dates available to you are limited.
I'm sure there's more, but it's certainly not that they don't know who to schedule.
I thought this thread was going to be about Schneider from One Day at a Time. Rest in peace.....or RPI.
Quote from: Windyplayer on January 07, 2016, 09:57:56 PM
Given the weight still given to RPI by the Committee and the amount of dollars flowing in and out of power conference programs based on tourney appearances, why don't universities employ statisticians (or do they) to--and I'm spit-balling here--run projections for all DI teams for the next year, pooling them into likely RPI spots (i.e. 1-50, 50-100, etc.) and then share that data with the program to schedule accordingly. It seems like it's just a numbers game and you keep on the schools from certain pools until your non-conference schedule is RPI proof.
When I run through some of these teams in the top 25-30 in RPI right now while we sit outside the top 100, it's appalling. We should never find ourselves in this position provided the resources bestowed upon our program (crappy RPI based in large part on scheduling not on-court play)
Curious to hear thoughts. Is this topic played out? Is MU considering this? Have they tried something similar?
They do.
Any idea of the extent of their diligence regarding schedule? How many employed? Methodology? How does MU stack up against blue bloods in this department? Any innovative methods out there?
With the money doled out for advanced statistics and analysis in baseball and most other professional sports, I'm curious to know how far universities are behind.
They do all of this and more. We just knew we had a young team abd scheduled accordingly. That abd some of our tough games like Belmont, Wisconsin, and LSU ended up not being as good as expected. Next year the schedule will be tougher.
The reason people like Brew (and me) were so upset with the schedule was that Broeker and co. are VERY good at creating schedules. They don't just pick teams out of a hat. I went into a bit more detail in this post http://painttouches.com/2015/07/17/historically-how-weak-is-marquettes-schedule/ (http://painttouches.com/2015/07/17/historically-how-weak-is-marquettes-schedule/)
Here is the relevant graph
(https://painttouches.files.wordpress.com/2015/07/non-con-average.jpg?w=630)
"Since 2011, Marquette's average non-con opponent ranking going into the season has been almost identical to the average final ranking of those opponents. What that means is the staff has a very good feel for which teams will be good and which teams won't, and schedule accordingly.
The predict column is the average of the ranking of all the Non-Con opponents from the year before (so the numbers we are currently discussing for the 2016 season). The actual is where that average ended up. Although we see fluctuations every year, on the whole, they don't overestimate or underestimate by much, if at all."
Having seen the level of play though, I admit, and have admitted, that I was wrong. Though it may lead to diabetes, this team needed the cupcakes it could get.
Quote from: pux90mex on January 08, 2016, 08:54:32 AM
The reason people like Brew (and me) were so upset with the schedule was that Broeker and co. are VERY good at creating schedules. They don't just pick teams out of a hat. I went into a bit more detail in this post http://painttouches.com/2015/07/17/historically-how-weak-is-marquettes-schedule/ (http://painttouches.com/2015/07/17/historically-how-weak-is-marquettes-schedule/)
Here is the relevant graph
(https://painttouches.files.wordpress.com/2015/07/non-con-average.jpg?w=630)
"Since 2011, Marquette's average non-con opponent ranking going into the season has been almost identical to the average final ranking of those opponents. What that means is the staff has a very good feel for which teams will be good and which teams won't, and schedule accordingly.
The predict column is the average of the ranking of all the Non-Con opponents from the year before (so the numbers we are currently discussing for the 2016 season). The actual is where that average ended up. Although we see fluctuations every year, on the whole, they don't overestimate or underestimate by much, if at all."
Having seen the level of play though, I admit, and have admitted, that I was wrong. Though it may lead to diabetes, this team needed the cupcakes it could get.
Thank you.
I do wonder how much of the soft schedule can be attributed to Broeker and co. recognizing a youthful team and to that group just not getting what they wanted.
I get that the team is young, but arguably two of their best players already had 1+ years of experience under their belts (Fisher and D. Wilson--both also had ample practice time against DI competition during their redshirt years). Further, the freshmen class included what many think to be an NBA ready stretch 4 so the handicap certainly wasn't for him. Lastly,Wojo already had a year under his belt at the helm, too.
Broeker seems to know how important scheduling is so I have to think he just ran into some bad luck and poor timing.
It sucks that our RPI is suffering but it does seem like we needed time to beat up on some teams, at least we look to be improving. We could have 6+ losses if we scheduled even some marginally better teams because of our inconsistency.
Quote from: LAMUfan on January 08, 2016, 09:11:40 AM
It sucks that our RPI is suffering but it does seem like we needed time to beat up on some teams, at least we look to be improving. We could have 6+ losses if we scheduled even some marginally better teams because of our inconsistency.
No doubt, it's just a matter of striking that balance. We appear to be off the mark this year. That's not to say we won't dance, but we made it more difficult to gain entry.
At one point, I was in the "we need cupcakes cuz we're young" camp. But I've changed my mind.
In general, we have played to the level of our competition. When playing decent teams, only against Iowa and Seton Hall did we look bad (and against SH, it really was only a bad 15 minutes or so). Otherwise we looked bad in beating some crap teams because we played down to their level and developed some bad habits.
Aside from Iowa and Hall, the other decent to good to very good teams we've played -- Providence, Wisconsin, Arizona State, LSU, even Georgetown and Belmont -- it's not as if we were cowed by them.
In retrospect (and yes, it's always easy to do these things in retrospect), I wish we would have played several more teams at least the caliber of Wis, ASU, LSU and Iowa.
Either the lads would have risen to the occasion, as they did in several of those, or we would have lost. In the case of the latter, we wouldn't have made the tournament, but it appears we won't now anyway having feasted on cupcakes. Had the former occurred -- had we risen to the occasion against most of our better opponents -- we would be in the tourney mix.
I'm confident that in future years we will play more representative schedules, even if it means giving up a home date or two.
Quote from: MU82 on January 08, 2016, 09:28:24 AM
In the case of the latter, we wouldn't have made the tournament, but it appears we won't now anyway having feasted on cupcakes.
We went 10-2 in non-conference beating Wiscy, LSU, and ASU and we may very well be doomed because of that. That's terrible. The RPI is a joke.
I am not sure how much it will change. However, I am assuming our final RPI is going to be effected by how our bunnies do in their own conference. In that case we do not yet know how weak the schedule is. I would think there is a difference between a bunny going 2-16 in conference vs. going 10-8. Am I right in thinking that if a majority of our bunnies finish with a winning conference record our RPI would improve? On the flip side, if they are finishing last in their conference our RPI is going to get worse.
Quote from: bilsu on January 08, 2016, 09:38:21 AM
I am not sure how much it will change. However, I am assuming our final RPI is going to be effected by how our bunnies do in their own conference. In that case we do not yet know how weak the schedule is. I would think there is a difference between a bunny going 2-16 in conference vs. going 10-8. Am I right in thinking that if a majority of our bunnies finish with a winning conference record our RPI would improve? On the flip side, if they are finishing last in their conference our RPI is going to get worse.
Take a look at that Paint Touches post above though. The variance between predicted and actual non-conference RPI in past years is minimal.
through 1/7 the projected RPI (http://www.rpiforecast.com/teams/Marquette.html) of the non-conf opponents is at 205.6.
Assuming the 218.8 mentioned previously used the same metric, that's a 6% improvement in a little over a week.
The RPI of your opponents isn't what matters. Cripes.
MU could have done a better job of playing crappy teams that won't have as horrendous of records as some of their opponents this year will have. The 'why's' I have heard don't make sense. I think they just weren't able to schedule some teams they should have for any number of reasons, or didn't do the work needed to avoid the super-craptastic teams.
I think if we get to 10 wins in conf, we'll be fine. Our RPI will only affect seeding and personally, all I care about in 2016 is making it.
9-9 is where it could come into play but I'm not sure .500 teams totally deserve the Dance.
In short, I believe our horrific RPI will hurt our seeding more than anything.
Quote from: Jay Bee on January 08, 2016, 12:10:43 PM
The RPI of your opponents isn't what matters. Cripes.
MU could have done a better job of playing crappy teams that won't have as horrendous of records as some of their opponents this year will have. The 'why's' I have heard don't make sense. I think they just weren't able to schedule some teams they should have for any number of reasons, or didn't do the work needed to avoid the super-craptastic teams.
As a refresher, per Wiki...
"In its current formulation, the index comprises a team's winning percentage (25%), its opponents' winning percentage (50%), and the winning percentage of those opponents' opponents (25%). The opponents' winning percentage and the winning percentage of those opponents' opponents both comprise the strength of schedule (SOS). Thus, the SOS accounts for 75% of the RPI calculation and is 2/3 its opponents' winning percentage and 1/3 its opponents' opponents' winning percentage."
The RPI is a decent indicator, but what actually matters is the W/L records of your opponents and your opponents' opponents. That's why I hate seeing SWAC teams on the schedule, because most end the season with losing records no matter their conference record and most of their opponents also have losing records. On the contrary, a team like San Jose State is a decent buy because while their record will likely suck, their MWC opponents will have better records.
Grambling and SJSU may finish with identical 2-28 records, but SJSU will have a more positive impact on our RPI. Similarly, you could play a team that has a tough schedule that leaves them with a 13-17 record but a 120 RPI or a team with a weak schedule that wind and has a 17-13 record but a 250 RPI. In that case, playing the 250 opponent would be far more beneficial.
Quote from: The Lens on January 08, 2016, 12:33:50 PM
I think if we get to 10 wins in conf, we'll be fine. Our RPI will only affect seeding and personally, all I care about in 2016 is making it.
9-9 is where it could come into play but I'm not sure .500 teams totally deserve the Dance.
In short, I believe our horrific RPI will hurt our seeding more than anything.
On my mobile so I can't check, but 10 conference wins would leave us around 100 in the RPI. Between conference and the BET, here's the scenario...
13+ wins: Virtual lock for NCAA Tournament
12 wins: Bubble
11 wins: Bubble, likely out
10- wins: Virtually no chance, need to win conference tourney
10-8 in conf could put MU with a projected ending rpi of 70. Probably a first round bye in the BET and then a matchup with the 4-5 conf seed.
conf 8-10 projects out to 100, 9-9 around 85.
Quote from: brewcity77 on January 08, 2016, 01:11:08 PM
On my mobile so I can't check, but 10 conference wins would leave us around 100 in the RPI. Between conference and the BET, here's the scenario...
13+ wins: Virtual lock for NCAA Tournament
12 wins: Bubble
11 wins: Bubble, likely out
10- wins: Virtually no chance, need to win conference tourney
No way in hell are we "likely out" with 11 wins in conference.
Quote from: Windyplayer on January 08, 2016, 01:47:58 PM
No way in hell are we "likely out" with 11 wins in conference.
Lose our last non-conference game and first game of Big East tournament 10-3, 11-7 and 0-1 puts us at 21-11 and a bid would be very questionable.
Win our last non-conference game and first game of Big East tournament 11-2,11-7 and 1-1 puts us at 23-10 and I like our chances.
11-2, 11-7 and 0-1 puts us at 22-10 and we are either last four in or first four out.
Quote from: bilsu on January 08, 2016, 02:29:38 PM
Lose our last non-conference game and first game of Big East tournament 10-3, 11-7 and 0-1 puts us at 21-11 and a bid would be very questionable.
Win our last non-conference game and first game of Big East tournament 11-2,11-7 and 1-1 puts us at 23-10 and I like our chances.
11-2, 11-7 and 0-1 puts us at 22-10 and we are either last four in or first four out.
This is a far cry from "likely out."
Quote from: brewcity77 on January 08, 2016, 01:11:08 PM
On my mobile so I can't check, but 10 conference wins would leave us around 100 in the RPI. Between conference and the BET, here's the scenario...
13+ wins: Virtual lock for NCAA Tournament
12 wins: Bubble
11 wins: Bubble, likely out
10- wins: Virtually no chance, need to win conference tourney
12 wins, even if 6 of them come from Creighton, St Johns and DePaul would include 5 more combined against Georgetown, Providence (already have 1), Nova, Xavier, Butler and Seton Hall. I know nothing about RPI (other than it's worthless as a true measure) but I can't imagine that scenario not being a lock. I think 11 gets us in. Sadly, I don't think we'll get 11 - or even 10.
Quote from: brewcity77 on January 08, 2016, 01:11:08 PM
On my mobile so I can't check, but 10 conference wins would leave us around 100 in the RPI. Between conference and the BET, here's the scenario...
13+ wins: Virtual lock for NCAA Tournament
12 wins: Bubble
11 wins: Bubble, likely out
10- wins: Virtually no chance, need to win conference tourney
11 BE wins and a first round BET loss would put us at 22-10 with an RPI of about 57. I'd be very surprised if we were left out under that scenario.
Ran the RPI Wizard. 11-7 with a first round BET loss to Xavier, Providence, Villanova, or Butler would give us this. We would be in.
Marquette
W-L RPI* SOS
22-10 56 77
That wiki crap re: RPI is nonsense; important to remember the first component is based on ADJUSTED win-loss, not "a team's winning percentage."
Quote from: brewcity77 on January 08, 2016, 01:11:08 PM
On my mobile so I can't check, but 10 conference wins would leave us around 100 in the RPI. Between conference and the BET, here's the scenario...
13+ wins: Virtual lock for NCAA Tournament
12 wins: Bubble
11 wins: Bubble, likely out
10- wins: Virtually no chance, need to win conference tourney
10 conference wins in the regular season and we're likely sitting in the 60-70 range. Depending on who we play, one BET win may be enough to put us in...
Right or wrong, I think the conference will help us and I think the wins against "known names" in the nonconference will help us...
Need to avoid bad losses... St. John's, DePaul.. these become big.
Quote from: Jay Bee on January 08, 2016, 02:43:13 PM
That wiki crap re: RPI is nonsense; important to remember the first component is based on ADJUSTED win-loss, not "a team's winning percentage."
So per KenPom, RPI is calculated as such...
Part I (25% of the formula): Team winning percentage. [Beginning in the 2004-2005 season], the NCAA added a bonus/penalty system, where each home win or road loss get multiplied by 0.6 in the winning percentage calculation. A home loss or road win is multiplied by 1.4. Neutral games count as 1.0. More on the effect of these changes can be found here.
Part II (50%): Average opponents' winning percentage. To calculate this, you must calculate each opponent's winning percentage individually and average those figures. This is NOT calculated from the opponents' combined record. Games involving the team for whom we are calculating the RPI are ignored.
Part III (25%): Average opponents' opponents' winning percentage: Basically taking all of the opponents' Part II values and averaging them.So really Wiki was just missing the adjustment made in Part I.
Quote from: brewcity77 on January 08, 2016, 01:11:08 PM
On my mobile so I can't check, but 10 conference wins would leave us around 100 in the RPI. Between conference and the BET, here's the scenario...
13+ wins: Virtual lock for NCAA Tournament
12 wins: Bubble
11 wins: Bubble, likely out
10- wins: Virtually no chance, need to win conference tourney
Completely disagree.
13+ wins: Stone cold lock.
12 wins: In comfortably.
11 wins: In.
10- wins: Bubble, perhaps one of last 4 in.
9 wins: Bubble, likely out, without a run to the BE champ, which could put MU back in last 4 in category.
Quote from: brewcity77 on January 08, 2016, 01:11:08 PM
On my mobile so I can't check, but 10 conference wins would leave us around 100 in the RPI. Between conference and the BET, here's the scenario...
13+ wins: Virtual lock for NCAA Tournament
12 wins: Bubble
11 wins: Bubble, likely out
10- wins: Virtually no chance, need to win conference tourney
If you want to get quoted, say something wrong ;)
top 5 BE are in (maybe 6 depending on circumstances) ... MU in if 10-8 & 1-1 BE. Not sure how you arrive at RPI projections...if we're 10-8 we'll be under 70 even with the playdough non-conf schedule.
quoted from an RPI article...I doubt its perfectly accurate but gives the general idea.
•The lowest rated teams to get at-large bids thru 2012: #67 USC, #64 Marquette (2011), #63 NC State (2005), #63 Stanford (2007).
Quote from: Windyplayer on January 08, 2016, 02:51:48 PM
So per KenPom, RPI is calculated as such...
Part I (25% of the formula): Team winning percentage. [Beginning in the 2004-2005 season], the NCAA added a bonus/penalty system, where each home win or road loss get multiplied by 0.6 in the winning percentage calculation. A home loss or road win is multiplied by 1.4. Neutral games count as 1.0. More on the effect of these changes can be found here.
Part II (50%): Average opponents' winning percentage. To calculate this, you must calculate each opponent's winning percentage individually and average those figures. This is NOT calculated from the opponents' combined record. Games involving the team for whom we are calculating the RPI are ignored.
Part III (25%): Average opponents' opponents' winning percentage: Basically taking all of the opponents' Part II values and averaging them.
So really Wiki was just missing the adjustment made in Part I.
Well, no. KenPom isn't an appropriate source either and has Part II underdescribed... the calc does not use the opponent's winning percentage individually.. it uses the opponent's winning percentage
excluding game(s) against you. Part II on the other hand does not make this adjustment.
Quote from: Jay Bee on January 08, 2016, 07:21:41 PM
Well, no. KenPom isn't an appropriate source either and has Part II underdescribed... the calc does not use the opponent's winning percentage individually.. it uses the opponent's winning percentage excluding game(s) against you. Part II on the other hand does not make this adjustment.
Well, it does. See the last sentence in Part II.
Quote from: jsheim on January 08, 2016, 04:22:51 PM
If you want to get quoted, say something wrong ;)
top 5 BE are in (maybe 6 depending on circumstances) ... MU in if 10-8 & 1-1 BE. Not sure how you arrive at RPI projections...if we're 10-8 we'll be under 70 even with the playdough non-conf schedule.
quoted from an RPI article...I doubt its perfectly accurate but gives the general idea.
•The lowest rated teams to get at-large bids thru 2012: #67 USC, #64 Marquette (2011), #63 NC State (2005), #63 Stanford (2007).
Was on my phone and typing a bit quick...what I should have been posting was TOTAL wins, not league wins. 23, 22, 21, 20 were the numbers I should have used. My mistake. Anyway...
Per RPI Wizard...
20-11 (9-9) gets us a 78 RPI
21-10 (10-8) gets us a 65 RPI
22-10 (11-7) gets us a 56 RPI
23-9 (12-6) gets us a 42 RPI
So that means 9-9 and we're out. Highly unlikely we get in with a 78 RPI. 10-8, we're probably out. As you note, teams have got in with lower RPIs, but it's not common. I'd expect us to be out with 21 total wins. Not impossible, but unlikely we'd be in. 22 wins give us a chance. We'd be squarely on the bubble. 50/50 chance we get in at 11-7. 23 or more wins, we're a lock.
Either way, if you want to get quoted, say something wrong. You think 10-8 gets us in comfortably, you're out of your mind. 10-8 likely leaves us out. Assuming your own numbers are correct, only one team with an RPI that low has made it in since 2007, which means there is probably a 95% chance we are out if we go 10-8 in league.
Quote from: Windyplayer on January 08, 2016, 07:44:19 PM
Well, it does. See the last sentence in Part II.
I agree. My mistake.
Quote from: Jay Bee on January 08, 2016, 10:04:05 PM
I agree. My mistake.
I appreciate the initial correction. I had no idea.
Quote from: brewcity77 on January 08, 2016, 09:41:23 PM
Was on my phone and typing a bit quick...what I should have been posting was TOTAL wins, not league wins. 23, 22, 21, 20 were the numbers I should have used. My mistake. Anyway...
Per RPI Wizard...
20-11 (9-9) gets us a 78 RPI
21-10 (10-8) gets us a 65 RPI
22-10 (11-7) gets us a 56 RPI
23-9 (12-6) gets us a 42 RPI
So that means 9-9 and we're out. Highly unlikely we get in with a 78 RPI. 10-8, we're probably out. As you note, teams have got in with lower RPIs, but it's not common. I'd expect us to be out with 21 total wins. Not impossible, but unlikely we'd be in. 22 wins give us a chance. We'd be squarely on the bubble. 50/50 chance we get in at 11-7. 23 or more wins, we're a lock.
Either way, if you want to get quoted, say something wrong. You think 10-8 gets us in comfortably, you're out of your mind. 10-8 likely leaves us out. Assuming your own numbers are correct, only one team with an RPI that low has made it in since 2007, which means there is probably a 95% chance we are out if we go 10-8 in league.
Remember, too, it still depends on who those 10 wins are against.
If we go 10-8 with the 10 wins against (remaining home games and beat SJU and Depaul on road):
We have RPI of 64 (didn't include BET)
If we go 10-8, but with the 10 wins against different teams road \ away
We have RPI of 68 or a few other rankings depending on the combination you wish to choose.
Quote from: ChicosBailBonds on January 08, 2016, 11:02:49 PM
Remember, too, it still depends on who those 10 wins are against.
If we go 10-8 with the 10 wins against (remaining home games and beat SJU and Depaul on road):
We have RPI of 64 (didn't include BET)
If we go 10-8, but with the 10 wins against different teams road \ away
We have RPI of 68 or a few other rankings depending on the combination you wish to choose.
Sort of, but not really. The difference of who they are against is incredibly marginal. The number of wins is what matters. You could say "but if we beat Villanova and Xavier" and that'd be great, but you'd have to offset that with two other losses, say to Creighton and DePaul, and suddenly it's not so great.
23 wins is the number for a lock, no matter who they are against, 22 is squarely on the bubble, 21 or less and we're probably waiting for a call from the NIT.
Put me in the 11 win and we are in camp, anything else we are out.
Also the committee seems to put very little value in winning or losing an early round conference tournament game. If you make a deep run (finals) it can help, but the bulk of conference tournament games are meaningless.
Quote from: Marquette_g on January 09, 2016, 07:17:54 AM
Put me in the 11 win and we are in camp, anything else we are out.
Also the committee seems to put very little value in winning or losing an early round conference tournament game. If you make a deep run (finals) it can help, but the bulk of conference tournament games are meaningless.
I'm not sure about that, actually seems like the deeper you go, the less the games mean, at least if you're on the bubble. It's easier for them to factor in the games that happen on Wednesday and Thursday than it is weekend games because by then their bracket is starting to take shape.
I can't see any way 23 doesn't make us a lock, and wouldn't hate our chances with 22 wins. But if we come in to the tournament at 20-11 (9-9) winning two may not be enough. It would help the RPI and win count, but by that point, they may already have us pegged as a sub-75 RPI team and not worth looking at if we don't win the conference tourney and force their hand.
Quote from: brewcity77 on January 09, 2016, 07:29:19 AM
I'm not sure about that, actually seems like the deeper you go, the less the games mean, at least if you're on the bubble. It's easier for them to factor in the games that happen on Wednesday and Thursday than it is weekend games because by then their bracket is starting to take shape.
What they could do and what they actually do are often not aligned. The Lunardi's of the world make very few changes based on winning a first round conference tournament game.
I contend that the conference tournament win is the most over-valued component of a season by fans. Sure winning the whole thing is nice and gives bragging rights (plus the sure bid) but unless that is the only way into the dance, it seems not to have much impact, yet often gets mentioned in threads like this.
They are seeding events (+/- 1 slot) by and large.
Really hope I am wrong but I just don't see how this team gets in 10 wins in conference. If you look at the remaining schedule it is really going to be tough. I have kind of resigned myself to a NIT this year. Again, hope I am way off the mark.
Brewcity, sorry if i sounded snarky with the something wrong comment...i was amazed at all the quotes you got...hit a nerve i think...so good post.
But 10 wins and 95%...no way...rpi will sort itself out...committee will be thinking strong conference, winning record, quality wins, home and away.
The tough part of the arguement is they have to finish 10-8 to prove anything....
Fischer made a hook....got to go!
I will say i assume a win in BET and providence doesnt tank
Quote from: jsheim on January 09, 2016, 01:03:22 PM
Brewcity, sorry if i sounded snarky with the something wrong comment...i was amazed at all the quotes you got...hit a nerve i think...so good post.
But 10 wins and 95%...no way...rpi will sort itself out...committee will be thinking strong conference, winning record, quality wins, home and away.
The tough part of the arguement is they have to finish 10-8 to prove anything....
Fischer made a hook....got to go!
I will say i assume a win in BET and providence doesnt tank
RPI won't "work itself out." I'm not pulling numbers out of the air. 10-8 will give us a sub-60 RPI, which most of the time means waiting for a call from the NIT. We go 10-8 in league and don't win a game at MSG and we'll be virtually guaranteed of watching the tournament on TV.
10-8 means needing 2 at MSG to be a lock. Any less than 10-8 and the only way we get in is the automatic bid. 23 total wins is the magic number.
Quote from: brewcity77 on January 09, 2016, 03:18:39 PM
RPI won't "work itself out." I'm not pulling numbers out of the air. 10-8 will give us a sub-60 RPI, which most of the time means waiting for a call from the NIT. We go 10-8 in league and don't win a game at MSG and we'll be virtually guaranteed of watching the tournament on TV.
10-8 means needing 2 at MSG to be a lock. Any less than 10-8 and the only way we get in is the automatic bid. 23 total wins is the magic number.
I agree. 10-8 and only 1 at MSG and we are going to be praying for a birth.
RPI just fell about 7 spots with the St John's win.
Also, worth remembering that our RPI in New York won't go up significantly unless we make the final. Go 0-1 and it drops, go 1-1 and it stays about the same. Strictly from an RPI perspective, you need to win more than you lose to see any marked increase.
Quote from: LAZER on January 09, 2016, 03:23:43 PM
RPI just fell about 7 spots with the St John's win.
(http://img.memecdn.com/we-got-this_o_364845.jpg)
Quote from: brewcity77 on January 09, 2016, 03:27:15 PM
Also, worth remembering that our RPI in New York won't go up significantly unless we make the final. Go 0-1 and it drops, go 1-1 and it stays about the same. Strictly from an RPI perspective, you need to win more than you lose to see any marked increase.
wouldn't most bubble teams rpi percentage decrease since most wins will be by teams in the tourney already so even our rpi actual number goes down our ranking could go up?
Quote from: martyconlonontherun on January 09, 2016, 03:44:15 PM
wouldn't most bubble teams rpi percentage decrease since most wins will be by teams in the tourney already so even our rpi actual number goes down our ranking could go up?
That's why I feel the first two days of conference tournaments are the ones that matter. The bubble teams that at least win 1-2 games. In the past 2 years, 75% of the first four (bubble teams that got in) won at least one in their tourney.
Quote from: brewcity77 on January 09, 2016, 03:27:15 PM
Also, worth remembering that our RPI in New York won't go up significantly unless we make the final. Go 0-1 and it drops, go 1-1 and it stays about the same. Strictly from an RPI perspective, you need to win more than you lose to see any marked increase.
Careful... must be delicate with words here.
If we go 10-8 in the BEast and go 1-1 in the BET, our RPI will drop. However, our RPI ranking may improve because so many other teams are losing in the BET and other tourneys. Depending on where our opp's w-l sits heading in, and how we do in the BET, things can improve quite a bit (RPI ranking) even if our RPI goes down slightly (as calculated).
All gets down to the details.
Let's win the whole thing
(http://spinoff.comicbookresources.com/wp-content/uploads/2014/10/youguysstink.jpg)
(http://i.crackedcdn.com/phpimages/article/2/5/0/340250_v1.jpg)
The way I look at is is the Big East is as good or better than it was last year and we got six bids.
2014-15 Standings prior to NCAA tournament
Villanova 16-2 32-2
Butler 12-6 22-10
Georgetown 12-6 21-10
Providence 11-7 22-11
St John's 10-8 21-11
Xavier 9-9 21-13
Xavier won two BET tournament games. So they were 19-12 pre tournament.
I think this years conference dispersion is going to be much tighter. I think 10-8 and 1 BET tournament win gets us to 22-11 which should be good enough.
Quote from: Jay Bee on January 09, 2016, 07:36:08 PM
Careful... must be delicate with words here.
If we go 10-8 in the BEast and go 1-1 in the BET, our RPI will drop. However, our RPI ranking may improve because so many other teams are losing in the BET and other tourneys. Depending on where our opp's w-l sits heading in, and how we do in the BET, things can improve quite a bit (RPI ranking) even if our RPI goes down slightly (as calculated).
All gets down to the details.
Yeah, that's why my pure guess would be that 1-1 ends up about a wash. In terms of the RPI ranking, would probably change no more than 2-3 spots when all's said and done.
Most likely, wherever we stand before the BET wouldn't change if we go 1-1, more likely just solidify that we are whatever they thought.
Quote from: Marquette Fan In NY on January 09, 2016, 10:23:11 PM
The way I look at is is the Big East is as good or better than it was last year and we got six bids.
2014-15 Standings prior to NCAA tournament
Villanova 16-2 32-2
Butler 12-6 22-10
Georgetown 12-6 21-10
Providence 11-7 22-11
St John's 10-8 21-11
Xavier 9-9 21-13
Xavier won two BET tournament games. So they were 19-12 pre tournament.
I think this years conference dispersion is going to be much tighter. I think 10-8 and 1 BET tournament win gets us to 22-11 which should be good enough.
Xavier finished with an RPI of 28 last year, which helped them get in despite being 6th in the conference. St. John's was 52 at 5th place. We're going to struggle to get into the 60's with 10 wins, which might not be enough.
Quote from: brewcity77 on January 09, 2016, 12:41:34 AM
Sort of, but not really. The difference of who they are against is incredibly marginal. The number of wins is what matters. You could say "but if we beat Villanova and Xavier" and that'd be great, but you'd have to offset that with two other losses, say to Creighton and DePaul, and suddenly it's not so great.
23 wins is the number for a lock, no matter who they are against, 22 is squarely on the bubble, 21 or less and we're probably waiting for a call from the NIT.
Marginal or not, if you go through all the scenarios, it can mean a 10-8 record each time, but the difference between a 59 and a 71. I think that's a pretty big spread, despite the records being the "same".
You guys realize that all of these RPI projections and "wizards" emanate from an extremely flawed model?
The whole thing is based on simulations run thousands of times for 350 teams who each have roughly 16-18 games left to play. That's close to 3,000 games in D-I left to play. And those results are based on KenPom's or Sagarin's models, which aren't perfect themselves.
Someone do the math on how many possible outcomes there are for those 3,000 games and tell me if running 10,000 simulations seems like it would be a fine predictor at this point in a season where all of the pundits are talking about parity.
IOW... Don't be surprised if MU closes the regular season at 10-8 with an RPI in the high 40s/low 50s. It's not about "simulating" the remainder of MU's schedule... you have to simulate everyone else's schedule too.
Quote from: Benny B on January 10, 2016, 02:10:48 AM
You guys realize that all of these RPI projections and "wizards" emanate from an extremely flawed model?
The whole thing is based on simulations run thousands of times for 350 teams who each have roughly 16-18 games left to play. That's close to 3,000 games in D-I left to play. And those results are based on KenPom's or Sagarin's models, which aren't perfect themselves.
Someone do the math on how many possible outcomes there are for those 3,000 games and tell me if running 10,000 simulations seems like it would be a fine predictor at this point in a season where all of the pundits are talking about parity.
IOW... Don't be surprised if MU closes the regular season at 10-8 with an RPI in the high 40s/low 50s. It's not about "simulating" the remainder of MU's schedule... you have to simulate everyone else's schedule too.
Don't see a high-40's / low-50's being likely under a 10-8 scenario. I think low to mid 60s at 10-8 is a reasonable expectation.
The key in that thinking (and the models) is that our opp's W-L calc is likely to be dragged down by a few very awful %'s... Chicago State looks like a 3-win team.. San Jose State.. maybe 4...
Now, Presbyterian, Grambling St, Maine... they may be able to pick things up a bit from where projected...but they're still going to be bad..
I think our best chance at securing an NCAA tournament bid would be by winning a lot more games.
Thank you. (Drops mic.)
Our RPI forecast right now is 120. We have a lot of work to do.
Quote from: MU82 on January 10, 2016, 10:39:14 AM
I think our best chance at securing an NCAA tournament bid would be by winning a lot more games.
Thank you. (Drops mic.)
We could lose all the rest of our games, but sweep the BET and we're in. It's a crap shoot.
The Big East could get 7 bids and MU could be one of the the three that does not get a bid.
Quote from: ChicosBailBonds on January 10, 2016, 01:27:24 AM
Marginal or not, if you go through all the scenarios, it can mean a 10-8 record each time, but the difference between a 59 and a 71. I think that's a pretty big spread, despite the records being the "same".
I guess I don't follow the math on this.
RPI only takes into account W/L and location when calculating the 25% record portion of the formula. If MU wins at DePaul and Creighton and loses at Villanova and Xavier, the RPI will be identical when compared to winning at Villanova and Xavier and losing at DePaul and Creighton.
I thought a team's actual RPI wasn't what the committee looked at the most. I thought they have been moving towards using it more as a way to quantify quality wins. I think our record against the Top 50 and Top 100 will be more important than whatever our RPI ranking ends up being.
We already are in a good place with the Providence win as far as resume building wins. The bubble isn't great and there were a lot of bad losses early on. Plus, while our non-conference was weak, we did come out with 2 neutral site wins and 1 road win. That's something that will help when comparing resumes.
Quote from: Jay Bee on January 10, 2016, 10:00:52 AM
Don't see a high-40's / low-50's being likely under a 10-8 scenario. I think low to mid 60s at 10-8 is a reasonable expectation.
The key in that thinking (and the models) is that our opp's W-L calc is likely to be dragged down by a few very awful %'s... Chicago State looks like a 3-win team.. San Jose State.. maybe 4...
Now, Presbyterian, Grambling St, Maine... they may be able to pick things up a bit from where projected...but they're still going to be bad..
But that's just MU's RPI... What matters is the RPI
Rank, and that's dependent on 349 other teams to a) determine the 349 other teams' RPI and b) where those 349 teams rank against MU.
Bottom line, too much volatility to make any sort of prediction of the remaining 3,000 games... And that makes RPI (and rank) volatile by extension.
Quote from: oldwarrior81 on January 10, 2016, 12:55:33 PM
I guess I don't follow the math on this.
RPI only takes into account W/L and location when calculating the 25% record portion of the formula. If MU wins at DePaul and Creighton and loses at Villanova and Xavier, the RPI will be identical when compared to winning at Villanova and Xavier and losing at DePaul and Creighton.
http://www.rpiforecast.com/wizard/Marquette.html
What I was saying is what makes up the 10-8 matters to some degree. Home, away, etc. As an example, go to the link I have and select MU to win their next 8 conference games and then lose the last 6. That would make us 10-8 with a RPI of 71. Now, do one where we lose most of the home games coming up, means a 76 RPI. Then there are scenarios that bring up around 60.
Quote from: oldwarrior81 on January 10, 2016, 12:55:33 PM
I guess I don't follow the math on this.
RPI only takes into account W/L and location when calculating the 25% record portion of the formula. If MU wins at DePaul and Creighton and loses at Villanova and Xavier, the RPI will be identical when compared to winning at Villanova and Xavier and losing at DePaul and Creighton.
Not exactly...because of the opp opp w-l component. (by "not exactly" I mean "no, that's incorrect.")
Quote from: Benny B on January 10, 2016, 03:28:39 PM
But that's just MU's RPI... What matters is the RPI Rank, and that's dependent on 349 other teams to a) determine the 349 other teams' RPI and b) where those 349 teams rank against MU.
Bottom line, too much volatility to make any sort of prediction of the remaining 3,000 games... And that makes RPI (and rank) volatile by extension.
Completely disagree (and not just on the faulty 349 figure). The idea that I'm unable to reasonable predict the records of our non-conf opponents and/or any teams in college bball is outlandish. I can, with great confidence, and would put $ on it.
If we are on the bubble and do not get in, will it be becasue of our RPI or because we lost to Belmont? It would seem to me it would be the loss to Belmont.
Quote from: bilsu on January 10, 2016, 07:26:09 PM
If we are on the bubble and do not get in, will it be becasue of our RPI or because we lost to Belmont? It would seem to me it would be the loss to Belmont.
The two are not mutually exclusive
Quote from: bilsu on January 10, 2016, 07:26:09 PM
If we are on the bubble and do not get in, will it be becasue of our RPI or because we lost to Belmont? It would seem to me it would be the loss to Belmont.
Under the scenario above , I think it would be lack of signature wins. If we can beat some more of the top teams in the conference we can be in the discussion.
Quote from: Marquette Fan In NY on January 10, 2016, 09:12:42 PM
Under the scenario above , I think it would be lack of signature wins. If we can beat some more of the top teams in the conference we can be in the discussion.
I said"If we are on the bubble" which I am assuming we have at least a 10-8 record in the Big East. Basically, if we are 9-9 or worse we will not be in the bubble consideration. Basically all 10-8 is is that you won one more game on the road than you lost at home. Right now we are even. I road win and ine home loss.
This league is a bitch, and I'm not sure this team is good enough to get to double digit wins in it. They don't value the ball and don't finish possessions very well on defense.
It wouldn't surprise me if they won against Nova or Xavier, and lost to DePaul or St. John's. Just the way it goes with a really young team. Trying to project out RPI rankings at this point is premature.
Quote from: MUfan12 on January 11, 2016, 08:38:02 AM
This league is a bitch, and I'm not sure this team is good enough to get to double digit wins in it. They don't value the ball and don't finish possessions very well on defense.
It wouldn't surprise me if they won against Nova or Xavier, and lost to DePaul or St. John's. Just the way it goes with a really young team. Trying to project out RPI rankings at this point is premature.
Unfortunately I agree. I think we probably go 8-10 or 9-9. Hope the warriors prove me wrong.
Pomeroy has 3 for sure wins left: DePaul, @St. John's, Georgetown
Probable Losses (7): @Nova, Xavier, @ Seton Hall, @ Xavier, @ Creighton, Nova, @ Butler
That's 9 losses right there.
Toss-Ups, 40-50% chance of winning (4): Butler, Providence, Creighton, @DePaul
So, MU wins all of the first group and all of the last group, that's 9 wins. And you have to win one of the probable losses to get to 10. And all of that is without losing one of the for sure wins or toss-ups. Don't see it happening. Also, we are currently ranked 29th in "luck" by Pomeroy. We will probably lose a close game or two in there as well.
Quote from: MUDPT on January 11, 2016, 08:59:53 AM
Pomeroy has 3 for sure wins left: DePaul, @St. John's, Georgetown
Probable Losses (7): @Nova, Xavier, @ Seton Hall, @ Xavier, @ Creighton, Nova, @ Butler
That's 9 losses right there.
Toss-Ups, 40-50% chance of winning (4): Butler, Providence, Creighton, @DePaul
So, MU wins all of the first group and all of the last group, that's 9 wins. And you have to win one of the probable losses to get to 10. And all of that is without losing one of the for sure wins or toss-ups. Don't see it happening. Also, we are currently ranked 29th in "luck" by Pomeroy. We will probably lose a close game or two in there as well.
Nothing for sure in college hoops. At SJU will be really tough. DePaul seems to get us every year or at least play us down to the wire, maybe it was a Buzz thing.
I truly believe that because of our youth we will improve more than other teams in conference do as the season goes on. Enough to make the tournament? Not sure
Quote from: MUDPT on January 11, 2016, 08:59:53 AM
Pomeroy has 3 for sure wins left: DePaul, @St. John's, Georgetown
Probable Losses (7): @Nova, Xavier, @ Seton Hall, @ Xavier, @ Creighton, Nova, @ Butler
That's 9 losses right there.
Toss-Ups, 40-50% chance of winning (4): Butler, Providence, Creighton, @DePaul
So, MU wins all of the first group and all of the last group, that's 9 wins. And you have to win one of the probable losses to get to 10. And all of that is without losing one of the for sure wins or toss-ups. Don't see it happening. Also, we are currently ranked 29th in "luck" by Pomeroy. We will probably lose a close game or two in there as well.
That's a good rundown..thanks.
I was thinking along the same lines...to go 10-8 we have to win a bunch of close games, not lose any "freshmen" games...and somehow overachieve against a superior opponent, perhaps on their court.
But some things to remember:
* upside of our freshmen higher than upside of typical juniors/seniors.
* Henry has not had a great game yet that we know he's capable of (see Junior Worlds game he took over). A couple of those perhaps?
Lets win one of the next two...those are probable listed as losses....lets get ahead of the game!!
A few things...first, I started a thread to test how well RPI Wizard works. I will be comparing the RPI ratings on March 6 with the projected RPI ratings as of today, January 11, for 25 teams in the NCAA from 11 different leagues.
Next...CBB pointed out that with a 10-8 record, our RPI could swing from 59-71 (I believe, without looking back). That really isn't that much. I stated earlier the expected RPI would be 65 with a 10-8 league record. So depending on how you move the victories, there's a +/-6 margin of error. The reason I started the other thread was to see how accurate they are. If most of their projections come in 6 off, I'd say that's not so good. If they are generally within 3-4 of projected, that's a valuable tool in my opinion. My guess is it'll be closer to 3 than 6.
Current RPI has Villanova #1 and Xavier #2, so we should get a nice boost this week win or lose.
Quote from: MUDPT on January 11, 2016, 08:59:53 AM
Pomeroy has 3 for sure wins left: DePaul, @St. John's, Georgetown
Probable Losses (7): @Nova, Xavier, @ Seton Hall, @ Xavier, @ Creighton, Nova, @ Butler
That's 9 losses right there.
Toss-Ups, 40-50% chance of winning (4): Butler, Providence, Creighton, @DePaul
So, MU wins all of the first group and all of the last group, that's 9 wins. And you have to win one of the probable losses to get to 10. And all of that is without losing one of the for sure wins or toss-ups. Don't see it happening. Also, we are currently ranked 29th in "luck" by Pomeroy. We will probably lose a close game or two in there as well.
Damn. I guess our lads might as well not even bother showing up to play.
Quote from: Litehouse on January 12, 2016, 08:39:07 AM
Current RPI has Villanova #1 and Xavier #2, so we should get a nice boost this week win or lose.
Unfortunately that's not how it works. Your opponents' RPI doesn't directly impact yours.
Think about the two components that can swing significantly...
1) MU's adj W-L
Currently, MU's record is 9.6-4.8, good for .6667...
IF MU loses both games this week, their record falls to 9.6-6.8 or .5854...
That's a drop of .0813... x 25% = .0203
2) Opp's W-L (ex games vs. MU)
Currently, MU's opp's W-L = .4865
If they play Nova and X, it improves to .5329
That's an improvement of .0464... x 50% = .0232
Component 3 is more difficult to move meaningfully...
The net of 1) and 2) is a positive .0029.
MU's currently at .5405... add .0029, we're at .5434
As of today, that would mean MU improves from #115 in the RPI to a three-way tie for #111.
Again, the DETAILS of the RPI calc are so very important. Blanket statements are difficult to make.. usually when speaking on the RPI, "it just depends..."
Wisconsin lost again, to Northwestern this time. Bryant McIntosh had his way and then some with Koenig. After Trimble, he may be the best PG in the B1G.