Tuesday, February 20, 2018

20 FEB Bracket Update

Writing proposals today, no analysis:

Monday, February 12, 2018

12 FEB Bracket Update

So there is nothing more pointless than the Bracket Preview.  I don't know, maybe it is extremely exciting if a) you do not follow much college basketball, b) you enjoy pointless arguments, c) you suffer from premature ejac... instant gratification.  Sure there were hardly any surprises (outside of them realizing only 15 teams had separated themselves from the pack, so they picked the 16th randomly) but it did give us at home the sense that there is a method behind the madness, and that analytical metrics will matter, as opposed to S-curving the RPI again.

I'm not factoring in the Q1/Q2/Q3/Q4 Wins and losses until March, since those rankings and values are subject to change greatly over the next few weeks, and there is no source to quickly nab a spreadsheet with the most up to date data.

Not many changes from the last couple brackets up top. UVA remains #1 overall, and Purdue hangs on to their seed.  Despite the perception, Duke is still much closer to the top line than most are willing to give them credit for... it's mostly just failing to live up to the expectations after acquiring Bagley.  Spreading the 4 ACC teams across the 4 regions works pretty well, though UNC has currently lost their cushy Charlotte gig.  The committee will likely ship UVA elsewhere to sell the tickets, but they belong in ACC country.

I have Gonzaga and URI claiming 4 seeds for now, but I believe once I adjust the formula to conform to the NCAA model, their power numbers will slip to the 5/6 range.  I swapped MSU and Auburn for this bracket, as the Spartans numbers jumped Sunday due to those B10 games playing to form.

At the bottom of the bracket, there has been some seismic activity.  Since most of these teams are either a) Not great, b) reliant upon home wins, or c) stuck in a mid-major without many opportunities for quality opponents, they bounce around each other a lot.  Here is the raw list:

The colors are fun, right? First, blue teams are currently under NCAA investigation (or ruled ineligible)... just a reminder and a possible discriminator for the last few teams.  For instance, USC was the 68th team last year... if they fall that close to the band saw again, they are getting cut.  Right now, I have them 67th.

Green are auto-bids, and it does show how close Middle Tennessee and New Mexico State are from stealing a bid if they lose.

Orange are the teams I feel are safe barring a bad loss or tailspin.  These are all somewhat flawed teams. Louisville is streaky. Texas is anything but streaky. Syracuse can't win on the road (unless they catch a streaky Louisville). Virginia Tech lacks non-conference superiority over other NCAA-caliber teams. Michigan is the safest of the bunch, but really ought not to finish behind Nebraska, Penn State, and another B10 Team.

Brown are the last teams in.  Baylor has great efficiency numbers, but has a terrible RPI and was buried at the bottom of the Big12 till blowing out Kansas.  They still need to do work, though.  Providence just lost to DePaul, which had them out, but the lack of teams stepping up against decent opponents allowed them to sneak back in, but they are very close to being irrelevant.  UCLA got the big wins they needed this week, though that schedule looks to be the ultimate booby trap.  Should be easy home games against UO/OSU, but two games at altitude and closing out with a rivalry game at USC could right back off the bubble.  USC themselves are in the same predicament. The Bonnies are in virtue of their head-to-head non-home wins against Syracuse, Vermont, Buffalo, and Maryland, which shows they can compete and beat bubble teams.  They are riding high, but need to beat URI and/or Davidson down the stretch and not choke away the easy A-10 games.

Red Teams are First 5 out. Why 5? Because I'm not one of Lunardi's sheep.  Maryland was done in by injuries and continues to fade.  They will not have a chance to play their way in until the B10 tourney, and semi-finals minimum at that point.  Notre Dame could be a factor if they get healthy, and have a huge chance tonight at UNC.  Boise State and Marquette both just lost "cannot lose" games and need quality wins... Boise gets Nevada Wednesday, but Marquette has gotta be thinking win-out at this point.  Kansas State just has crappy numbers right now, but will get their shot.  Funny that 3 of last year's play-in teams (USC, Providence, Kansas State) find themselves in the same spot this year.

Purple teams are the next 5 out and need a bolt of lightning to have a chance.  Temple is a bizarre case of how much weight will be put on November wins.  They are a "strength of schedule" darling, but really need the sweep of Wichita State and another big win to get in the bubble discussion.  Nebraska has the best shot of doing this because they are likely going to win out and finish 3rd or 4th in the Big 10; in doing so, they will eliminate Maryland and Penn State from the bubble, and despite crappy numbers and no big wins, they are not Illinois State.  Penn State has some great opportunities, getting slumping Purdue and a possible OSU sweep to their resume, but they have to be hoping Rider wins out and scratches off that "bad loss". Ben Howland has MSU playing grating, frustrating basketball that has resulted in some nice wins, but they have ONE road win, and likely are not adding to that.  Utah also sucks on the road, but if they sweep the northwest road trip, they finish with home games against UCLA, USC, and Colorado... a possible 7-game win streak.  6-1 (gotta beat Washington, though) gets them to the bubble.

And the white teams?  These guys just don't have the profile right now.  NC State was way too conservative out-of-conference, and mixed in a couple good wins (Arizona, PSU) with puzzling losses (NIA, UNCG).  Western Kentucky just ran out of good juju.  The UTSA loss clearly puts them on a lower rung than MTSU, then the other losses (Ohio, Belmont, Missouri St) scream NIT, which sucks, because this team beat Purdue and SMU at their peaks and nearly dropped NOVA as well.  SMU is decimated, St. John's is 2-11 in the Big East, but scores 132 on the total efficiency model. 133-134 is the low end of the bubble, which means that the Johnnies are better than their record and could make some noise.  LSU swept Arkansas and Texas A&M (while A&M was short-handed, but won't matter, especially since they are short-handed again) and finishes with a manageable schedule.  Oregon lost to UConn... this year, which is pretty weak.  They finish with opportunities, but their non-conference jewel is Fresno State, which is just no.

Washington doesn't even make the cut right now.


Friday, February 9, 2018

9 FEB Bracket

UVA will be the #1 overall until they really fuck it up (like lose to Pitt or something... btw, take the under in that one). Reracked the regions, with Xavier stealing the Midwest, shuttling Purdue to LA.  Helps align the 2 seeds better, and Duke is falling to the 3 line.

Texas A&M gets the big boost, and seeing they are back to full strength, I have no problem holding at the 5 line.  Ohio State feels underseeded at 4.

Does anyone get the feeling a C-USA/PAC-12 challenge would be close? 

Louisville and Syracuse are the last teams that are "safe" today, while Providence, Arkansas, Texas, and UCLA (yes, those Bruins) are you happy bubble teams.  Boise State is the last bye.  Alabama and USC have the best credentials of the first four, while Marquette and Virginia Tech have the best upside.

Saint Bonaventure and Kansas State are the two closest to cracking the field, while Baylor, Maryland, and Notre Dame have work to do. I feel that it is comical to be considering Temple as a viable at-large contender, but their numbers are so much better than Washington and NC State, so you cannot ignore that November happened.  My guess is that they finish 4th in the AAC and the committee takes a long look before swallowing the bad losses into the NIT. Western Kentucky needs a run or an autobid, as they are closer to ODU that MTSU now. Mississippi State, Penn State, and Nebraska round out my 10 out. Sorry St. John's, it's not enough to get out of the hole.

I feel good about this bracket, but then Saturday will happen and blow it up.

Sunday, February 4, 2018

4 FEB Bracket

Villanova remains the number 1 overall, holding a small margin over Virginia.  There is an ocean to Purdue, and another ocean to get down to Xavier and Duke, rounding out the top 5.  Auburn and Cinci keep winning, and Clemson maintains solid numbers.

Play-in lines will remain both on the 12 for now, but this will shift as one-bid teams slip in the rankings.

Nobody in the Big 12 has really taken charge, leaving a mess of differently-abled teams.  TCU looks done without Fisher, and Texas will go as far as bad basketball will take them.  K-State fell from lock to out, having been swept by KU and WVU, and the TCU/Baylor wins have lost their numbers luster.

The middle of the SEC is equally frustrating, as Alabama, Missouri, Arkansas, Texas A&M, and Georgia have spent the last 3 weeks bouncing between 35 and 60, and now LSU, USC and MSU want to enter the mix... hell, even Vandy is trying to play ball.

Teams hanging on for dear life: Houston, Alabama, Providence, Boise State, Arkansas.
This is actually better for the first 4 on this list, as they typically been in the last 5 in category.  Consistent winning should be rewarded.  Arkansas should right the ship with 5 winnable games before a difficult close.

Last Five In: Syracuse, UCLA, Western Kentucky, Washington, Marquette.
The numbers have Marquette out as of today, but this is their spot to hang onto as they just were put through the gauntlet, still have a road test in Newark before closing with 2 St. John's, 2 Creighton w/o Kramliej, Georgetown, and DePaul.  5-2 seems likely.  Washington has the worst numbers of anyone I have put into the bracket, but their big wins give them the nod over NC State for now.  UCLA and ASU are basically playing a bracket buster game this week.  Syracuse comes and goes, will probably be gone after Louisville nukes their offense.  WKU has the best chance of hanging out here, but the committee won't be thrilled about a multibid C-USA... they need to keep winning and remain ahead of ODU.

First Five Out: Kansas State, Maryland, Virginia Tech, Baylor, NC State.
I'm pretty sure MD is finished, but they could find another gear... and let's be honest- making the Big 10 championship game is not a stretch if they can hold on to the 4th or 5th spot. Just gotta catch Purdue napping.  KSU is a bracketology nightmare... a week ago during their run the checked in a 21.  WTF?  Nobody is confusing KSU as a top 5 seed.  This week? 69... I'm reading it more like 50ish... VA Tech needs more than a couple road wins over collapsing ND and BC.  They are closer that NC State mathematically, but I think 4 out of 5 dentists would ride the hot hand.  Baylor is more likely finishing 9th in the conference over getting a 9 seed.  Just erratic play all around.

Next Five Out: Notre Dame, Nebraska, Southern Methodist, St. Bonaventure, Penn State
Temple was out to a hot start with the RPI to back it up, then started the AAC 0-4... now they are sneaking back, but probably aren't getting a sniff at more than 11 losses. SMU and ND have been done in by injuries.  The Bonnies are dangerous and will very likely win out the regular season and find there way in, but as it stands now, they are prone to dropping a sub-100 road game and being NIT bound. Penn State and Nebraska are just yuck and nobody wants to watch them.  Give me Yante instead.

Monday, January 29, 2018

29 JAN Bracket

Analysis to follow

Friday, January 26, 2018

Friday 26 JAN Bracket

Southern Methodist... yikes!

Monday, January 22, 2018

Formula tweak

There is nothing more frustrating than letting the data tell you how to analyze it.  I feel like this is the game we play with Bracketolgy.  We are basically trying to reverse engineer something that has no plan and is not repeatable.  Basically, we are going back to Doc Brown rigging up the clock tower to send the Delorean back to the future.  One time, your timing will be right and everything will come together.  Most times, you will do everything correctly and somehow Air Force will sneak in anyways.  It's this fine line between what we believe to be correct through tireless analysis of efficiency numbers and game results to determine who is good, who beat who, where, when, and why.

We know that the committee will acknowledge advanced metrics, but everything that they do proves that they will fall back on the RPI.  Not as bad as past years, where they drop the 64 teams on the S-curve and maybe shuffle around a little so that they met their conference and venue criteria. 

I have tweaked my Total Rating formula with an RPI factor.  As a whole, nothing major changes.  I would have been more inclined to put Maryland in around the 10-11 seed with this formula and request that SMU and Houston duel to the death for the final spot.  While the Power 5 teams would be seen as benefitting most, Rhode Island get a huge jump would be hanging with Arizona and Cinci on the 4 line.  Nevada would remain 5-worthy.  St. Mary's and Gonzaga would get crushed though.

St. Mary's profile out like ISU last year, and splitting with the alpha in a mid-major is no longer a ticket.  The Nitty Gritty team sheets are barren for Big Wins, and their saving grace is that Jock Landale is a beast and makes for watchable basketball.

22 January Bracket

Man, this turned into a total goat screw.

My guess is that somebody else in the Big 10 gets hot.  Very odd year, where they have so little depth that it has propped Ohio State into this perceived juggernaut.  They are not, but nobody can stop them right now.

New Mexico State edges out Western Kentucky for the final at large bid.  I had to re arrange my play-in games to account for Buffalo taking the MAC autobid from Ball State (Buffs are right on the cusp of an at-large themselves) and ETSU snatching the SoCon bid at Furman (Bucs are also in the at large mix).  Buffalo is squarely on the 11 line, so dropped that other game back ass a 12/12.  Providence noses out Baylor.

The margin between the 1 and 2 seeds continues to grow.  There is a great deal of volatility as the major conferences' teams become less dependent on NCSOS and more dependent on "what have you done for me lately.  This may need to be tweaked, so we don't see a team like Seton Hall bouncing from 4 to 10 seed based on who the beat or lost to any given week.

I... was not aware St. Mary's RPI was so low.  I watch the games, see they are clearly a top 25 team, but holy shit the resume is the Gonzaga upset and New Mexico State.  Kansas State is also a team of circumstance right now... I had them way out just a week ago, yet 2 home wins (good wins) puts them in a safe place.

The PAC 12 has done nothing to make me believe they are getting a 4th team in, and if the FBI hammer were to fall... uh, yeah, this could be a 1 bid league (it won't).  The drop off from USC to UCLA is significant, and nobody else is even close.

1 seeds: Villanova, Purdue, Duke, Virginia
Hanging on the Bubble: Arkansas, Missouri, Alabama, SMU, Marquette
Last 5 in: Houston, Providence, Texas, Boise State, New Mexico State
First 5 out: Western Kentucky, Maryland, Baylor, Notre Dame, BYU
Some others to consider: Syracuse, Louisiana Lafayette, Virginia Tech, UCLA, Nebraska

Sunday, January 21, 2018

Saturday Notes

The AAC is that deep.  Wichita State is still an elite, top 16/25 team, as is Cinci.  Houston and and SMU should easily fit within the field.  Certainly better than 5th best teams in the Big 10 and PAC 12.

St. Bonaventure may have two of the best players in the conference, but with 4 league losses, they are outbid or bust, and I can’t see them winning 3 straight non-home games to do that. Opens a spot for… 

Based on Emmert’s toothless statement, the NCAA isn’t doing anything about the teams mixed up with Dawkins.  Which is sad, because he is one guys, and that shit is EVERYWHERE in the NCAA.  No action condones it.

… Louisville is off the bench.  As long as they don’t lose to BC tomorrow.

Injuries are the story again this week.  Creighton is devastated for a second straight season.  TCU was already struggling in the Big 12; no Fisher will hamstring them further.

Mountain West needs to be taken seriously.  Nevada is strong and well coached.  Boise should be in the field, and San Diego State and Fresno State are certainly capable.  Wyoming and UNLV are NIT bound.

St. John’s is the worst team in the Big East.  Congrats DePaul!

Seriously folks.  St. Mary’s was in the field prior to stealing the WCC lead in Spokane.  Pepperdine and Portland give the conference a false bottom, which really weighs down their numbers, but this is also a Top 25, Top 8 seed team.  Stop deluding yourself that the WCC, MWC, C-USA, and A-10 are one-bid leagues.  There is enough parity in the 10-60 team range that the only conference looking like they can get a sub-.500 league team in is the Big 12.

Villanova… wow.  They played careless and sloppy in the second half.  Turnovers, fouling jump shooters… yet they are so efficient shooting it doesn’t matter.

Arizona State has regressed.  They need to be blowing the doors off Cal, not squeaking out a single point win, even on the road.  Not sure how February will treat them, but i see them finishing much closer to the bubble than a top 4 seed.

Friday, January 19, 2018

Middle of the road

Last year, my bracket numbers were all over the place.  My formula was immature.  Metrics were poorly chosen. I had to hand-jam numbers so they were often not up to date (like, I would have never been tracking LSU).  My bracket last season typically had most of the correct at large teams,  but many were under or over seeded based on public perception.

This year, with better use of specific components of the RPI and efficiency metrics, my bracket has been more consistent week to week.  The seedings also fall in line with the rest of the matrix.  Check it out: no blue or red outside the top line (last week Texas Tech crept ahead of Duke, so I slid them into a 1 spot... oops).

Also, the formula has done a better job of balancing recent success with the entire body of work.  Ohio State may look overseeded right now based on their preseason perception, but their overall body of work is just as strong as their recent hot streak.  It does lean heavily on NCSOS, so this will damped the effect of a hot 3 game homestand against conference foes.  The RPI still gives a little too much credit to mid-pack power 5 teams that get trounced against all good teams, but steal enough home victories to remain middle of the pack.  Those miserable losses are enough to keep them relevant in the RPI, but often they are floating in line with top mid-majors by other metrics.

It will be interesting to see how this keeps up as more conference data is accumulated.