Friday, December 29, 2017
28 DEC Bracket
The #1 seeds have stayed pretty consistent. I've seen a lot of love for Arizona State as #1. The numbers just aren't there, and the bottom-heavy PAC-12 is going to take the shine of their NC wins. If they win the conference convincingly, they may get there, but right now, I see them jostling the 2-4 lines. Purdue is likely a little over-seeded. Their conference slate starts light and sees MSU only once. I feel like a three team shuffle between ASU, Texas A&M, and Purdue would make me feel better.
I don't like 3 Big 12 Teams on the 3 line, and expect Kansas to move up as well, but they need a run to find themselves ahead of the Sun Devils.
Notre Dame has cratered their way into the toughest road possible. I still think that they pull things together for a top 3 conference finish, as the schedule is favorable the next few weeks, and for the most part, you can chalk those two bad in-state losses to poor luck come March, but in December they look brutal.
Auburn and Southern Cal have been pulled out of the dumpster and dusted off, much like Arizona and Miami last round. The schools have really pushed their recruiters and assistants out on a limb, and the toothless NCAA only bullies the little guys. It's been quiet enough on that front this month to think that the feds will let the basketball component plead out and testify against the deep pockets of Adidas. The middle men like Dawkins will get hung out to dry for facilitating it. With the amount of Louisville turnover/turmoil, I still cannot imagine that they don't have a "come to God" moment and sit this one out. Who knows what other players they are going to find dirt on...
Treading on Thin Ice: Notre Dame, Virginia Tech, Baylor, Maryland, Minnesota. These teams can come off a big week and look like a top 5 seed, or crap the bed to the bottom third of the conference and stumble out quickly. Minnesota and Notre Dame are the most perplexing. The other three are good, but not great teams.
Last 5 In: BYU, Houston, UCLA, Butler, St. John's. The Johnnies are likely behind the Friars now, as they were dismantled at home after I pulled the numbers. UCLA dug themselves a nice hole and have gone Bane to get going. It will be interesting to see how high they can climb, as their real conference games don't start until February, and they close with 3 on the road, @Utah, Colorado, and USC. Could murder their season. Butler has their Super Bowl Saturday hosting Villanova. They are legit, but 4 bottom 100 wins drags on their overall numbers. They AAC is deep enough to get Houston in this year, after they sat on the bottom of the bubble looking up last season. BYU is the flavor of the week, though they need to drop the Gaels this weekend to hang around the bracket much longer.
First 5 Out: Syracuse, Missouri, Oregon, Providence, Alabama. Sorry, I hate to deep digging up this horse to beat it, but Boeheim needs to learn that he cannot keep trying to force this hand. Outside of a well-fought win here in DC against pretty suspect Georgetown team, they took on a couple neutral site games, but other than that, they paraded a bunch of 100+ ranked schools through the Carrier Dome. It's not the worst, as 5 of those teams are likely in the Field, but the loss to the Bonnies has them out now. They need a real road skin too. Virginia, Duke, Florida State, or Miami... fuck, I'd take Louisville as well. They won't and we will go for round 3 with the committee on their sham profile. Providence has been discussed, but they need more wins. Alabama is fun to watch, but they chase the pace... will have impressive wins sandwiched around WTF losses. Missouri was a lock 3 weeks ago, but their December was pretty bad, losing to a struggling Illinois and escaping SFA. Considering this had been one of the worst major conference programs in the country the past two years, being is actually quite an achievement. Oregon will have a lot to prove in conference, but their road win at Fresno shouldn't be ignored, even with their struggles in Portland.
Next 5 Out: Ohio State, Iowa State, Boise State, Temple, Boston College. The Buckeyes are in better shape than their other Big Ten fringe players (Northwestern, Penn State). They are already 2-0 in conference and their losses are solid losses. Their lack of road games is a little perplexing and it bumps them a notch. Iowa State is going to have to fight it in a stacked Big 12, but with quality wins abound, the only way to not be in the mix is to crash out with 3 or 4 wins. Boise is much better than this, but they are a little low on the KenPom and Sagarin numbers, which means they really have to push the envelope on Nevada. Temple had the best computer numbers, but those have slipped. There 3 100+ losses cancels out their shocking win(s) in Charleston. BC is hanging that Duke win out their and will ride it as long as possible, but .500 in the ACC is what they need to make up for weak NCSOS and results.
In the mix: Northwestern, Penn State, Marquette, Kansas State, Georgia, Utah, New Mexico State, Northern Iowa, UNLV, San Diego State
Tuesday, December 12, 2017
12 DEC Bracket Analysis
To qualify the changes, the "pre-season" bracket, though published around the first of December, was primarily informed by my preseason picks with little statistical influence, and maybe one of two anecdotal results. I also leaned heavily toward schools in the FBI's crosshairs taking the burden of proof off the NCAA and taking the postseason penalties as soon as possible. Right now, I really only see Louisville (as a culmination of events) and Auburn (warming up the bus to back over basketball to spare football) as being "locks" to remove themselves from participation.
Everything has been reprogrammed to extract the required data regularly. It's just a matter of following a pattern to consistently manage the data to build a product. The metrics I am rating on are same as last year. First, a "Basic" measure that indicates as to whether or not they play good basketball (do they score more per possession/do they allow fewer points per possession). Next, there is a "Strength" measure that digests RPI and schedule to verify the results. Finally, there is a "Trend" measure that addresses the when and where of results. Typically, the analysis is done independent of record until that matters (closer to March). Basically, a 5-4 team now (like Florida, URI) doesn't just topple out of the field, nor does an undefeated team (I hate using the Georgetown example because it is extreme, so Syracuse) with good metrics immediately drop other proven teams. A four metric blends the three to give a broader picture.
Logic is also applied, such as MTSU has a blistering RPI, but the C-USA doldrums will dampen their seed. Schools outside the Big 6 will need sterling resumes. Come March, since the NCAA shares their RPI breakdown, that will be used as a tie breaker, because the exercise is to project the field, not argue who deserves it more.
One of the criteria that DOES NOT help analyze and project the field is looking at Top 25, 50, 100 wins and losses. First, rankings change hourly, and your snapshot is as good as the last result. A top 25 team can topple out of the Top 50 with a sprained ankle, and then that win is worthless. The Big East was a mess of who beat who and when last year. The Pac12 and Big12 had teams jockeying the 50 line every night, with some teams losing as many as 6 "Rated" games a night based on results, only to have a few pop up the next night when more results come in.
The other things that do need to be considered: health, ratio of games played away, non-conference schedule balance, and avoiding unexplainable losses.
Here is a snapshot of the tool:
Numbers-minded folks probably recognize some of the value sources, but it is really how you interpret the committee's mindset AT THAT TIME as to how you balance the formulas.
That said, the bracket:
After Duke LOLed in Chestnut Hill (don't be fooled there), it created firm separation between Villanova, MSU, and the other 349 teams. The curve smooths after that, with Duke and Wichita State riding shotgun for the other #1 seeds. I was pretty relaxed with the regions and had no problem dropping UVA out west. Despite Nova's roll, I still like Xavier to pace them in the conference, and a 2 seed fits their numbers. UNC and WVU continue their silent assaults. A&M, Kansas, and Gonzaga are next up. The biggest surprises falling into the chaos? Kentucky is struggling a little, Notre Dame crapped the bed, Cincinnati should be much closer to the Shockers, and Florida has entered the abyss. Minnesota was hot, then suddenly not. The Gaels are on life support now... getting swept by the Zags again with boot them. The Pac12 is a dumpster fire. Surprises emerging? Tennessee, Temple, Clemson, OHIO STATE (maybe it was all Thad's back), and of course Arizona State.
Breakdown by conference:
A10: 2
AAC: 5 (yes, 5)
ACC: 8
Big Ten: 5 (Wisc/NW implosions have left their mark)
Big 12: 7
Big East: 6
MVC: 2
MWC: 1 (but oh so close Boise)
Pac12: 2 (I see 4 here again, but what a miserable non-conf run)
SEC: 7 (yeah, I had to wtf that one as well)
WCC: 2
22 one bid leagues
Teams that are closest to the bubble:
Virginia Tech, Ohio State, Temple, Minnesota, Saint Mary's
The major conference teams just need to defend their home court and not lose to the dregs on the road and their numbers will remain steady. The Gaels and Owls will likely be on a weekly SOS rollercoaster as to whether or not they are safe.
Last 5 in:
Alabama, Houston, Providence, Saint John's, Northern Iowa
The Panthers need help from a middling MVC. Bradley has rebounded to give the league a sixth decent team, but Valpo, the Redbirds, Bears, and Braves are all going to be eyed as NIT at best, even with a strong conference run. Loyola and Northern Iowa need to stay on top of that middle class and hope it keeps their numbers strong. These aren't Creighton and WSU from the old Valley though. The Big East has 3 team log jam, and that could break at any point. WSU's move to the AAC should buoy Temple and Houston's numbers for the rest of the season; double digit conference wins will be huge. The SEC is so strong that Alabama went from potential Top 3 to "just hope to stay in the top half."
First 5 out:
Syracuse, Butler, Maryland Boise State, UCLA
UCLA is a solid cut below the other 4. Depending how I prioritize the numbers, Cuse, Butler, MD, and Boise replace any of the 5 teams ahead of them (and Minnesota). UCLA is really just there, not bad, but in need of big wins in the state of Arizona to really have a shot. Boise needs good seasons out of Fresno, SDST, UNLV, and New Mexico to boost the conference numbers.
Next 5 out:
Penn State, Iowa State, Marquette, Utah, Oregon
If you were looking for the Pac12, you have found the mortuary. Marquette can jump into the BE mix again if they get hot. Penn State is exactly what you expect from the middle of the pack in the Big 10. Decent numbers, a few notable win, and no idea whether or not they are any good. When I was originally programming the spreadsheet, the first run pumped out 9 teams (OKST was DOA). Kansas lost and KSU lost and OKST lost. Iowa State has a ways to go, but the ground work is there. Better shape than UCLA long term; no bad losses in the Big12.
In the Hunt:
BYU, KSU, Miss St, Georgia, Fresno, SDST, NCST, Towson!, UNLV, Northwestern, UCF, ODU, Wisconsin, Wake Forest, and I swear-to-freaking-god Boston College.
Anomoly:
I have Hawaii as my Big West pick... but UC Davis looking really good.
Everything has been reprogrammed to extract the required data regularly. It's just a matter of following a pattern to consistently manage the data to build a product. The metrics I am rating on are same as last year. First, a "Basic" measure that indicates as to whether or not they play good basketball (do they score more per possession/do they allow fewer points per possession). Next, there is a "Strength" measure that digests RPI and schedule to verify the results. Finally, there is a "Trend" measure that addresses the when and where of results. Typically, the analysis is done independent of record until that matters (closer to March). Basically, a 5-4 team now (like Florida, URI) doesn't just topple out of the field, nor does an undefeated team (I hate using the Georgetown example because it is extreme, so Syracuse) with good metrics immediately drop other proven teams. A four metric blends the three to give a broader picture.
Logic is also applied, such as MTSU has a blistering RPI, but the C-USA doldrums will dampen their seed. Schools outside the Big 6 will need sterling resumes. Come March, since the NCAA shares their RPI breakdown, that will be used as a tie breaker, because the exercise is to project the field, not argue who deserves it more.
One of the criteria that DOES NOT help analyze and project the field is looking at Top 25, 50, 100 wins and losses. First, rankings change hourly, and your snapshot is as good as the last result. A top 25 team can topple out of the Top 50 with a sprained ankle, and then that win is worthless. The Big East was a mess of who beat who and when last year. The Pac12 and Big12 had teams jockeying the 50 line every night, with some teams losing as many as 6 "Rated" games a night based on results, only to have a few pop up the next night when more results come in.
The other things that do need to be considered: health, ratio of games played away, non-conference schedule balance, and avoiding unexplainable losses.
Here is a snapshot of the tool:
Numbers-minded folks probably recognize some of the value sources, but it is really how you interpret the committee's mindset AT THAT TIME as to how you balance the formulas.
That said, the bracket:
After Duke LOLed in Chestnut Hill (don't be fooled there), it created firm separation between Villanova, MSU, and the other 349 teams. The curve smooths after that, with Duke and Wichita State riding shotgun for the other #1 seeds. I was pretty relaxed with the regions and had no problem dropping UVA out west. Despite Nova's roll, I still like Xavier to pace them in the conference, and a 2 seed fits their numbers. UNC and WVU continue their silent assaults. A&M, Kansas, and Gonzaga are next up. The biggest surprises falling into the chaos? Kentucky is struggling a little, Notre Dame crapped the bed, Cincinnati should be much closer to the Shockers, and Florida has entered the abyss. Minnesota was hot, then suddenly not. The Gaels are on life support now... getting swept by the Zags again with boot them. The Pac12 is a dumpster fire. Surprises emerging? Tennessee, Temple, Clemson, OHIO STATE (maybe it was all Thad's back), and of course Arizona State.
Breakdown by conference:
A10: 2
AAC: 5 (yes, 5)
ACC: 8
Big Ten: 5 (Wisc/NW implosions have left their mark)
Big 12: 7
Big East: 6
MVC: 2
MWC: 1 (but oh so close Boise)
Pac12: 2 (I see 4 here again, but what a miserable non-conf run)
SEC: 7 (yeah, I had to wtf that one as well)
WCC: 2
22 one bid leagues
Teams that are closest to the bubble:
Virginia Tech, Ohio State, Temple, Minnesota, Saint Mary's
The major conference teams just need to defend their home court and not lose to the dregs on the road and their numbers will remain steady. The Gaels and Owls will likely be on a weekly SOS rollercoaster as to whether or not they are safe.
Last 5 in:
Alabama, Houston, Providence, Saint John's, Northern Iowa
The Panthers need help from a middling MVC. Bradley has rebounded to give the league a sixth decent team, but Valpo, the Redbirds, Bears, and Braves are all going to be eyed as NIT at best, even with a strong conference run. Loyola and Northern Iowa need to stay on top of that middle class and hope it keeps their numbers strong. These aren't Creighton and WSU from the old Valley though. The Big East has 3 team log jam, and that could break at any point. WSU's move to the AAC should buoy Temple and Houston's numbers for the rest of the season; double digit conference wins will be huge. The SEC is so strong that Alabama went from potential Top 3 to "just hope to stay in the top half."
First 5 out:
Syracuse, Butler, Maryland Boise State, UCLA
UCLA is a solid cut below the other 4. Depending how I prioritize the numbers, Cuse, Butler, MD, and Boise replace any of the 5 teams ahead of them (and Minnesota). UCLA is really just there, not bad, but in need of big wins in the state of Arizona to really have a shot. Boise needs good seasons out of Fresno, SDST, UNLV, and New Mexico to boost the conference numbers.
Next 5 out:
Penn State, Iowa State, Marquette, Utah, Oregon
If you were looking for the Pac12, you have found the mortuary. Marquette can jump into the BE mix again if they get hot. Penn State is exactly what you expect from the middle of the pack in the Big 10. Decent numbers, a few notable win, and no idea whether or not they are any good. When I was originally programming the spreadsheet, the first run pumped out 9 teams (OKST was DOA). Kansas lost and KSU lost and OKST lost. Iowa State has a ways to go, but the ground work is there. Better shape than UCLA long term; no bad losses in the Big12.
In the Hunt:
BYU, KSU, Miss St, Georgia, Fresno, SDST, NCST, Towson!, UNLV, Northwestern, UCF, ODU, Wisconsin, Wake Forest, and I swear-to-freaking-god Boston College.
Anomoly:
I have Hawaii as my Big West pick... but UC Davis looking really good.
12 DEC Bracket
Work is interfering with posting the analysis, but should have that up before I get out of here for hockey. Villanova is the number 1 overall, Houston and Northern Iowa are last teams in, Maryland and Butler first teams out.
Thursday, December 7, 2017
Confirmed... and some things I did not see coming
A new segment that will likely exist primarily on Twitter. Basically a quick take of a couple of the results of the night that help explain some of the thought behind selections and seedings for this year's bracket.
6 DEC
Confirmed suspicions... that they are going to make a run in The Valley
Loyola 65, Florida 59: Loyola is a very slow, frustrating team to play. Not necessarily like Virginia, who grinds the game to a halt, but a team that doesn't give away a lot on the offensive end as far as poor shot selection. 3 point shots are taken in rhythm to maximize conversion rate and prevent breaks the other way. Extra passes in and out of the paint make 8-footers 6-footers, and 6-footers vulnerable for an easy backdoor cut. I was not expecting that Florida would enter the SEC slate undefeated, but a couple losses to Duke, FSU, even Cinci or Clemson next week would risk their 1 line status... Loyola though? Maybe Loyola is an at large caliber team. Which is good because Valpo is a wildcard right now, hot, but untested.
Blindsided... Just because Lorenzo Romar couldn't have been THAT bad
Washington 74, Kansas 65: KU had beaten UK, blown out tourney front runners S Dakota St, Oakland, and Texas Southern, and humbled Syracuse. I get the game was in KC as opposed to Allen, but the Huskies ran with this one late. Typically, the better team gets behind early and just can't get enough rhythm or opportunity to comeback. But not this time. UW won every stage of the game and Kansas, playing more like a feisty underdog, kept up early, but just couldn't make the plays a top team would make to close out inferior competition. And this is not indicative of the Huskies level of play. They have struggled to beat their "peers" in the mid-level rankings (100-200) and were quite convincingly dispatched at MSG last month by Top 50 squads.
6 DEC
Confirmed suspicions... that they are going to make a run in The Valley
Loyola 65, Florida 59: Loyola is a very slow, frustrating team to play. Not necessarily like Virginia, who grinds the game to a halt, but a team that doesn't give away a lot on the offensive end as far as poor shot selection. 3 point shots are taken in rhythm to maximize conversion rate and prevent breaks the other way. Extra passes in and out of the paint make 8-footers 6-footers, and 6-footers vulnerable for an easy backdoor cut. I was not expecting that Florida would enter the SEC slate undefeated, but a couple losses to Duke, FSU, even Cinci or Clemson next week would risk their 1 line status... Loyola though? Maybe Loyola is an at large caliber team. Which is good because Valpo is a wildcard right now, hot, but untested.
Blindsided... Just because Lorenzo Romar couldn't have been THAT bad
Washington 74, Kansas 65: KU had beaten UK, blown out tourney front runners S Dakota St, Oakland, and Texas Southern, and humbled Syracuse. I get the game was in KC as opposed to Allen, but the Huskies ran with this one late. Typically, the better team gets behind early and just can't get enough rhythm or opportunity to comeback. But not this time. UW won every stage of the game and Kansas, playing more like a feisty underdog, kept up early, but just couldn't make the plays a top team would make to close out inferior competition. And this is not indicative of the Huskies level of play. They have struggled to beat their "peers" in the mid-level rankings (100-200) and were quite convincingly dispatched at MSG last month by Top 50 squads.
Labels:
blindsided,
confirmed,
Florida,
Kansas,
Loyola,
Washington
Monday, December 4, 2017
Preseason Bracket
So the preseason bracket is a month late. Most of the conference picks were made between the first of November and Thanksgiving, but the bracket layout and seeding has been delayed for a few reasons.
- My computer, iPad, thumb drives, etc, were all stolen in San Francisco this summer. Lesson: In SF, if it isn’t tattooed or pierced to your body, it will likely be stolen. Be warned. Because it was a new folder, I had not set the permissions to back-up any of the data to the Cloud or my Airport. Well, that’s on me.
- It’s OK, because I was constantly playing catchup, everything was hand-jammed to spreadsheets. This year, I have tried to make my data acquisition a little more efficient, which has required some up front coding…
- I suck at coding.
- I do most of this during work down time (last year I was on the final extension of a contract, which equates to blowing off real work) which there has been little of…
- Too many kids.
- This whole NCAA-FBI thing. This type of selection bias goes well beyond KenPom, RPI, endowment dollars, etc. Clearly there is so much wrong going on in college basketball right now to circumvent the rules, how do you punish those who have been caught in this one scheme, versus what is known to be occurring at numerous other programs? Especially when the NCAA administrative body has proven that they have know idea what a student athlete is.
- I am lazy.
I’m not here to hold ethical or moral judgment. It amazes me Pitino got pinched before Calipari. It bothers me a coaching lifer like Larranaga is thrown into the industrial wasteland that has become NCAA recruiting.
In other words, I moved several schools into a category reserved for schools at risk, whether it be administrative turmoil, APR, or investigations. Of these schools, only two are in my field: UCLA and Miami. The damage done at USC, OK State, Louisville, Arizona, and Auburn will likely result in internal punishments to cover this season. In the case of Louisville, the NCAA may even rule on that sooner than later due to the amount of specifics already out in the public. Alabama is on the cusp of flushing their best squad in years down the drain, so here’s hoping they come clean in the next few weeks. They are good enough to be playing the second weekend
This has moved my PAC12 outbid from Arizona to Oregon. Also, at print time, I have had enough of Harvard and will roll with Steve Donahue’s dark horse Quaker Oats team.
The bracket:
Duke has had enough near deaths to make me believe that ACC regular season will humble them. They have done enough to hold their 1 seed, but they may not even win the regular season title (hint- they don’t have to). Same with Michigan State. The Big12 is loaded and TCU gets the first #2 seed.
I really like Villanova, but I am going to stick with my gut that a senior-laden Seton Hall and frustrated by circumstance Xavier team will test them off the top 2 lines. Wichita State is out making me look like a buffoon right now, but I still like Cincinnati to beat them to the top AAC spot. PAC12 is really gutted without Arizona and USC. Oregon and UCLA just aren’t mature enough to do much better than the 5 line.
Mid Major alerts:
I like Nevada. They have improved a lot the last few months, and the MWC is receiving nicely, thanks to hot starts from San Diego St., Boise St., and UNLV. I have Wyoming holding a play-in spot, but that could very likely be one of those aforementioned schools. Gonzaga reloaded quickly, but the WCC only goes as far as the Gaels can take them, and they are not strong enough to get into the preferred pod position. The Bonnies took a week 1 hit to the resume, but still will be a factor Selection Sunday.
The One Bid leagues will likely be buried, much like UT Arlington and Middle Tennessee were last season. However, the numbers continue to insist that these are fringe Top 25 schools, not Top 100 fodder. The CAA was in this position last season, but without UNCW driving the train, CoC gets screwed. Vermont, Florida Gulf Coast, Bucknell, and South Dakota State are the schools nobody wants to draw. It will be very difficult to keep the winners of the HBCs out of the 16 seed play-in games, as only Texas Southern is remotely to playing at D-1 standards right now.
1 SEEDS
Kansas, Duke, Michigan State, Florida
2 SEEDS
TCU, Notre Dame, Texas A&M, Xavier
WILD CARDS (Teams that can blow this up)
Villanova, Arizona (Not in Field), Wichita State/Cincinnati, Gonzaga
LAST FIVE IN
Oklahoma, Northwestern, VCU, Central Florida, Wyoming
ENJOYING A SOFT BUBBLE
Maryland, SMU, Mississippi St, Texas, St. John’s, St. Bonaventure
TEMPORARILY NOT CONSIDERED (that matter)
Arizona, USC, Louisville
FIRST FIVE OUT
Alabama, KSU, Temple, San Diego St., Utah
NEXT FIVE OUT
Clemson, PSU, Washington St., UConn, NC State
Subscribe to:
Posts (Atom)