Bracketbullets, 2018 Edition Comment Count

Brian

25730218477_96466a589f_z

[Marc-Gregor Campredon]

Eh, I'm fine with it. Not getting Detroit over an MSU team with two wins over tourney teams and two double-digit Ls to Michigan, neither at Crisler, is annoying. But other than that I'm not trading draws with them. Michigan pulled the weakest #1 and weakest #2. Gonzaga's a strong #4 and that's usually good since there's a better shot they take down the 1 before the E8. In this case the 1 is so weak that Kenpom favors Gonzaga in a hypothetical matchup. That's bad, but emphasizes how good the draw otherwise is.

And all of these hypothetical regional matchups would be happening in Los Angeles, hundreds and hundreds of miles away from anyone in the region. Flip M into MSU's slot and they're staring down Kenpom #3 Duke and #9 Kansas… in Omaha. I'll take #7 UNC and #8 Gonzaga in a building that might slant to Michigan given the cosmopolitan nature of the fanbase over a roiling pit of Jayhawk partisans. Also, Izzo is 1-11 versus Duke.

Meanwhile, the first round. Michigan got the top 14 and top 6. The first probably won't matter—M has been installed as a 12 point favorite. Montana has five games against major conference opponents this year. They beat Pitt in OT, lost by 13 to PSU, lost by 16 to Stanford, and lost by three at Washington. All of those were on the road, naturally. They're 0-1 in Kenpom "A" games—the PSU outing—and 1-6 in "B" games, which include the aforementioned losses plus Ls against nonconference mid-majors Santa Barbara and Georgia State (a 15 seed) plus a 1-2 record in road games against the three toughest Big Sky opponents.

Despite that they're #71 in Kenpom, better than the other 14s by a fair distance. This is because they've hamblasted a bunch of Big Sky teams en route to a 16-2 conference record. When the difficulty level steps up they haven't been able to hang.

Montana plays a couple of 6'8" posts with no stretch ability and relies on their point guard for 27% of their shots; Synergy has them in the 14th percentile at catch and shoot and 16th on the off the dribble jumpers that Michigan has been very good at forcing. With all due respect to the Grizzlies, this isn't the picture of a 14-3 upset, especially against Michigan.

920x920

Rob Gray and his unfortunate hairstyle

The hypothetical second round. While Houston's had an impressive season boy do they look like the team you'd pick out for Michigan to play in round two. Let us run down the ways. Houston…

  • Relies on a high-usage point guard. Rob Gray is at 29% usage; his efficiency relies on TO avoidance and getting to the line. 50% of his threes are unassisted. He's 6'1" and not super athletic. That's the profile of a heavily-relied-upon guy that X can turn into a potato.
  • Is transition-dependent. Houston was #13 nationally in transition eFG; Synergy has them in the 91st percentile. They're only slightly above average in % of transition shots, but the upper reaches there are populated largely by teams that aren't any good and are just trying to get a shot up before the defense can get set.
  • Relies on threes. Average number go up; 34th in hitting them. Michigan is top ten at preventing three launches. Two Houston players are Just Shooters and don't threaten much when you run them off the line.
  • Lets you shoot threes. They're 195th in allowing them. 3PT D is good at 43rd but how much is luck, how much is real, etc.

The one thing that stands out on Houston's resume that's bad for Michigan is that their defense is massively foul-prone. Michigan is unlikely to take as much advantage of that as your average team.

In the "remains to be seen" category: Houston pounds the boards—17th—despite not having anyone taller than 6'8" on the floor. Houston is in fact tiny. Three different 6'6"-6'7" guys get about 75% of their minutes at the 4 and 5. Michigan has done very well at keeping the opposition off the boards and probably should in a hypothetical second-round matchup, but the sheer weirdness of Houston's approach here might give them avenues that Michigan isn't used to dealing with.

The other second round. Kenpom gives San Diego State a 34% chance at the upset of Houston and the Aztecs are an entirely different challenge. They're huge (18th nationally in height), three-averse, and frequently use a 2-3 zone. They have what might be the strangest three-game stretch in the country: a win against Gonzaga bracketed by losses to Cal (at home!) and Wyoming.

So despite the 6 vs 11 thing, I think I'd rather see Houston. San Diego State is not overly dependent on their PG (21% usage), doesn't run that well, runs a long-ass zone, and has a big Duncan Robinson matchup problem in Malik Pope. Pope is a diverse and athletic 6'10" four who can face up on or post Robinson and might be able to clobber him on the boards. Duncan's come a long way but I'm not real happy when the opposition rolls in with a 4 who's their biggest-usage guy.

But what about UNC? The Tarheels did clobber Michigan early in the year, thanks in large part to a 15 minute stretch spanning halftime during which Michigan got three buckets. This team is not that team. Eli Brooks played 18 minutes; Zavier Simpson played one minute more than Ibi Watson did. Jordan Poole and Isaiah Livers both got their first real minutes against a real team—and on the road.

I would not expect Michigan to win that game, and that's fine. Michigan earned a three, got a three, and got what's actually the perfect Kenpom matchup at that point in the bracket: the #7 overall team vs the #10 overall team. But I don't think the first game is at all representative of what you should expect. UNC starts three seniors and two juniors coming off a national championship. Michigan had no idea who their point guard was or, frankly, what shape their ass was at that juncture.

UNC poses a bunch of matchup issues and Michigan will have to play their best game of the year to beat them. Even so I expect that to be tooth and nail. Should they be so fortunate to make the Elite Eight that game will feel like a breath of fresh air.

Don't even think about it. It will be UNC. Lipscomb? No. A 9-9 SEC team? No. Providence? No, even though their coach is Bunk Moreland. The Tarheels got a gift draw to the Sweet 16.

39852911454_0492101241_z

or big boy cool glasses [Campredon]

TIME TO PUT ON THE BIG BOY PANTS. It took a typical committee injustice to prevent Penn State from getting the Kenpom booby prize this year. They're the second-best team left out of the field, one slot behind St. Mary's. And they're a four-seed in the NIT. Same goes for Nebraska, which turned a (soft) 13-5 conference record into a 5 seed in that same tournament.

This is in part because both teams scheduled like garbage in the nonconference. Aside from their mandated Big Ten-ACC challenge games, this is what those two teams took on amongst power conferences and other actually good teams:

  • Penn State: Pitt, Texas A&M.
  • Nebraska: Creighton, Kansas, BC, St John's.

Nebraska deserves a little sympathy for coming up just short against those two tourney teams but the rest was dreck. This goes for the rest of the conference, too. The Big Ten's NCSOS markers per the NCAA's reckoning:

  • Ohio State: 32
  • Purdue: 71
  • Wisconsin: 92
  • Maryland: 141
  • Illinois: 169
  • Indiana: 201
  • Iowa: 203
  • MSU: 217
  • Minnesota: 244
  • Michigan: 259
  • PSU: 265
  • Nebraska: 274
    • Northwestern: 306
    • Rutger: 333

    There are about 350 D-I basketball teams. Just five Big Ten teams were in the top half, and two just barely. I know the committee head basically laughed this metric off earlier this year, since the RPI is about 75% SOS all of that crap got lumped into actual RPI numbers and diced into quadrants and what not. It got batted back and forth as the league went through its conference schedule.

    The league's scheduling has real impacts you can see when better ranking systems survey the landscape, like Seth Burns's implied pythag:

    Pythag has a lot more respect for the Big 10 than the RPI has, and it shows here with the Big 10 getting three teams on the top two seed lines. Surprisingly (to me anyway), the ACC would only get one.

    In a WAB (or Implied Pythag) world, Nebraska would be safely in. Ditto for St. Mary’s. Marquette and Middle Tennessee would be the last teams in, while Oklahoma State would be the first team out.

    Nebraska's an 8. That's what they deserved. Oh and MSU's a 1 and Michigan a 2. But because our conference has its collective head up its Izzo, none of that came to pass. Nobody even thought about Nebraska as an at large because the collected weight of RPI boat anchors moved a top 30 Kenpom team (Penn State) out of the top 75 in RPI. And moved a top 50 Kenpom team (Maryland) almost out of same.

    Hell, you don't need to even put on the Big Boy Pants. Just stop scheduling SWAC and MEAC teams, which are 8 points worse than the ASun. We'll see if next year's committee really dusts the RPI. If so, hooray. If not the league should fine any team that ends up with an NCSOS under 200.

    Comments

    Maize4Life

    March 12th, 2018 at 11:25 AM ^

    those non conference games against Alabama A & M..While Im not opposed to one after a tough stretch or maybe 2 at the most but that should be it..Schedule Quad 1 & 2 for all nonconference games with the exception of what I mentioned above..Had we done that this year we prob are a high #2 seed even close to #1...

    mfan_in_ohio

    March 12th, 2018 at 12:58 PM ^

    The team as it is now is light years ahead of where it was in December.  Remember, we barely beat UCLA (at home) and Texas and blew a big lead to OSU, and of course lost to LSU in Maui.  If we replaced the 6 worst quad 4 games with all quad 1 and 2 games in November and December, do you think we'd go better than 3-3 in those games?  I doubt it.  

    If you want to quibble with Beilein's scheduling, you can say that he could drop Alabama A&M and UC Riverside in favor of some midlevel MAC or Horizon teams, but remember that we also struggled with CMU in our second game.  Back when these games were happening I remember someone saying we should have gone out and scheduled a high RPI low-major team like Rider, but Penn State's loss to Rider might be the reason they are out of the tournament.  

    Say what you want about Beilein's schedule, it gave a team that lost three starters a chance to figure out its roles without torpedoing its chances for a high seed before January.  It might warrant tweaking for next year, but I find it difficult to look at a 3 seed and complain.

    TrueBlue2003

    March 12th, 2018 at 5:01 PM ^

    If we did, then we'd have a resume identical to UNC's (25-10 with a top 3 SoS in the country).   We'd be a 2-seed ahead of Cinci, Tennesee and MSU.  And we would have had a lot better games to watch in December.  And really, that's my biggest complaint. Difference between a 2 and a 3 isn't much (unless a 2 can get you to Detroit, sigh).  But those SWAC MEAC games are no fun for the players (unless you're Brent Hibbits and you actually get to play), terrible to watch, and they don't help you figure anything out. They're just total wastes.

    This isn't football.  You can lose some games.  The committee puts so much emphasis on SoS that playing tough games even if you lose some is good.

    funkywolve

    March 12th, 2018 at 11:40 PM ^

    it seems pretty clear the NCAA isn't going to punish you for a few extra losses if you're playing a tough non-conference schedule.  If you're just going by record, Kansas probably shouldn't be a 1 seed and UNC shouldn't be a 2 seed but those teams have SOS and NCSOS that are some of the best in the country.  

    TrueBlue2003

    March 12th, 2018 at 2:10 PM ^

    and they're a relatively good matchup for us.  Their offense is mostly terrible except for crashing the boards and I know this fanbase has PTSD about that, but Yaklich has us as one of the best DREB teams in the country and it's not a fluke.

    We'd probably put Wagner/Teske on the higher usage 6'10 guy and let Duncan check the freshman when Duncan is in the game and he'd be fine there.

    That they play a zone means they're terrible at preventing threes (274th at preventing threes!).  So yeah, if we missed a bunch of them we'd be in for a rock fight like Iowa but that's worst case scenario.

    While we match up well with Houston, at least they're actually a really good team.  If Z picks up a foul or two early on Gray, that game is in the danger zone.  They'll probably be able to switch our ball screens with some success given that their bigs are smaller and athletic. We should win it, but they're good.

    Goblueman

    March 12th, 2018 at 11:50 AM ^

    1.Mo avoiding foul trouble..2.Def Rebs.,especially vs. NC.Def Reb stats have been good this year so hopefully that continues...3.FT's ,especially in end game with lead scenarios.The 'talking head blow hards' will say 3 point shooting is key but even though it's up & down for stretches within games it usually ends up around 35% by games end.Don't be shocked if we see S.D.St. in 2nd game,they hit boards hard like NC but obviously not as much talent.

    ScruffyTheJanitor

    March 12th, 2018 at 11:54 AM ^

     

    Don't even think about it. It will be UNC. Lipscomb? No.

     

    /Waves tiny Bisons* flag.

     

    *Yes, Bisons. There was a VOTE on by the Lipscomb student body and they voted... poorly. It was before I got there, so I am not to blame.  Maybe we don't deserve to win, except UNC.

    Hail-Storm

    March 12th, 2018 at 12:00 PM ^

    His size might work really well in first few rounds as a counter to Mo.  His shot blocking on defense and ability to get offensive rebounds and play the pick and roll would be fun to see.  I love how many ways this team attacks and how they make other defenses wrong. The week and a half off is probably not as bad for this team as in other years since they can withstand a half of poor shooting by having great defense. 

    MHWolverine

    March 12th, 2018 at 12:05 PM ^

    I really hate the 2 week layoff but I'm confident once they shake off the rust and get past Montana then they will be well on they're way to a second straight Sweet 16 apperance under JB

    Go Blue!!

    charblue.

    March 12th, 2018 at 12:11 PM ^

    two rounds and then figured we get Duke or Carolina. Having watched these teams virtually all season, it makes no difference which one we face, because they both play nearly the same games relying on drive and kick and then heavy offensive rebounding. They both love to run, but neither are great shooting teams outside of certain guys.

    I watched Virginia and its swarming defense basically throttle both teams in the semis and finals. It was like Michigan playing Purdue in which Virginia maintained a constant lead that ebbed and flowed throughout with Virginia's defense the difference.

    I see Michigan as having the kind of defense that Virginia possesses while Carolina has a dominant inside presence, it doesn't do a great job of moving the ball, and typically settles for one-on-one play by Berry and Pinson, and then crashing the boards MSU style. The Spartans actuallty matchup great against the Tarheels which explains why they dispatched them earlier in the season.

    The state of North Carolina put six teams in the tournament, two more than the Big Ten, and Carolina lost to two of them Duke and NC State, going 2-1 against the Blue Devils and 1-1 against the Wolf Pack. At this point, I would rather face Carolina than Duke even though Carolina is the more veteran team.

    mGrowOld

    March 12th, 2018 at 12:16 PM ^

    Anybody and I mean ANYBODY who follows basketball analytics knows the RPI is hot garbage cause of the insane weighting it puts on your opponents, opponents schedule and yet the committee uses it like it was handed down from God.  

    What's drives me nuts is that everybody knows the committee uses it yet we make no effort to gig the system to our benefit and it's almost pathetically easy to do so via your OOC schedule.  Every other league seems to have figured out this dirty little secret (the selection committee uses the RPI almost exclusively....shhhh....dont tell anyone) but B1G continues down their merry path of scheduling misadventures which kill our collective RPIs.

    #Maizenwasright (about scheduling)

    PurpleStuff

    March 12th, 2018 at 12:54 PM ^

    MTSU (33), USC (34), St. Mary’s (40), and UCLA (36 in play-in game) all got boned. ACC got 2 teams in with 60+ RPI and ASU (66) made play-in game. Also seems like they ignored quad system in favor of weird “What was your best win (and mid-majors don’t count)?” metric.

    PurpleStuff

    March 12th, 2018 at 4:21 PM ^

    Pains me as an SC hoops season ticket holder, but the Bruins snub makes even less sense. 36 RPI, but just the kind of big scheduling/wins the committee pretended were important. Lost @Michigan (12), home to Cincinnati (6), and to Creighton (44) in Kansas City. Beat Kentucky (10) in NOLA and won in Tucson (Arizona schools didn’t have to come to LA). Committee whose spokesman admitted to not watching PAC-12 championship just mailed it in.

    mfan_in_ohio

    March 12th, 2018 at 1:15 PM ^

    Texas Tech is a good example.  Of the six metrics used by the committee, RPI had them lowest (23rd) while the others had them in the 3-4 seed range, and they ended up as a 3.  

    What I would like to see them do is, if they are going to use the quadrant system, then base the rankings not on the RPI, but rather using the average of the six metrics they used.  Basically, rank the teams 1-351 on the average of their six metric rankings. If the teams are to be judged at one level based on both result- and predictive-based metrics, then judge them the whole time on those metrics.  They already have the numbers on the sheets; it shouldn't be too hard to have the numbers next to each team played reflect the average instead of just RPI.  

    TrueBlue2003

    March 12th, 2018 at 3:18 PM ^

    issue, in general.  St. Mary's, Nebraska, and ND and Baylor and PSU all had bad NCSoS on kenpom, too. 

    The committee clearly wants to see that you've scheduled and played (and beaten) good teams.

    And I hope they keep doing that, because if that's the incentive teams like Nebraska and PSU need to schedule more difficult schedules, it's a good thing for college basketball.  If we let teams in just for pounding bad teams and inflating their efficiency/scoring margin metrics, that's not the kind of incentive structure that makes the game better.

    Even though those teams are arguably amongst the 26th-45th best teams, they aren't going to win the national championship so leaving them out is good for the game.  If it's arguable and close in that range, absolutely put in the teams that challenged themselves and made better viewing early in the season.

    The committee did the right thing with St. Mary's and PSU.

    TrueBlue2003

    March 12th, 2018 at 5:44 PM ^

    TCU twice, Texas twice, Baylor twice, Oklahoma once...and they had no bad losses.  As explained below the NCSoS is being used as a bubble tiebreaker of sorts.

    If you win enough against good teams and don't lose to bad teams like KSU, you're not going to be on the bubble and won't be subject to the NCSoS default tiebreaker.

    PurpleStuff

    March 12th, 2018 at 5:54 PM ^

    None of those teams are good. UCLA beat Kentucky on a neutral court, UA in Tucson, and a higher ranked SC team twice. They are in Dayton, despite blue chip talent and three Sweet 16 appearances last four years. If you think the teams you mentioned and Tech, FSU, Mizzou, etc. with no tourney track record are demonstrably better than UCLA you are on shrooms.

    TrueBlue2003

    March 13th, 2018 at 1:21 AM ^

    says random internet guy that is blasting the committee for "arbitrarily" deciding who is good and who isn't yet ignores the metrics that are used and goes on his own opinions.

    Then the dude starts citing historical performance, which is most certainly not part of the criteria. No wonder you took a hiatus for arguing. SMH.

    Per the metrics, those teams are good.  This is not my opinion.  All seven of those wins for KSU are considered Q1 or Q2 wins by the quadrant system except the home win over Baylor. Also included in Q2 wins is the road win at Ok St.  Kenpom likes KSU's wins even better than RPI.  They have 11 A or B tier wins by that system.

    Whether I think anyone is "better" or "worse" doesn't matter and that's the whole point.  It matters whether you're deserving and the committee is pretty consistent that you need to do two things:

    1) beat a bunch of good teams/win a lot of good games and

    2) don't lose to bad teams/don't lose easy games

    You'll notice that it doesn't matter if you lose to a lot of good teams if you satisfy the first two criteria.  That's been well-established.

    KSU satisfied both criteria.

    If you fail in both regards, you're done.  If you fail in one, you're on the bubble and NCSoS matters once you're there.  In UCLA's case, they lost to too many bad teams - CU at home and Oregon St. - but their NCSoS kept them in it.  And for the record, I think they should have been in more comfortably and I do like their team, but at least they weren't left out and they did lose some head-scratchers such that they put themselves in danger.

    In MTSU's case, they failed both 1 and 2.  And yes, that system does screw mid-majors because they don't get as many cahnces to rack up good wins, and they have more chances to pick up bad losses, but the committee is pretty consistent with that criteria.

    And that's how ASU and Syracuse got in.  Fairly high number of Q1/Q2 wins (6+ for each team) and despite tons of losses to good teams they had very few losses to bad teams (only one Q3 loss in the case of ASU, and two for Cuse and all three were barely Q3).  USC was also near there but the committee compared the resumes and dinged them for a really bad loss (home to Princeton was Q4), and only 3 Q1 wins, all three of which were barely Q1.

     

    Zenogias

    March 12th, 2018 at 12:20 PM ^

    I really do not understand why MTSU is suddenly the hill everyone is willing to die on with respect to NCAA tourney selection. I know KenPom isn't the final word on this stuff, but MTSU is the #52 team on KenPom. They are behing Maryland, USC, Baylor, Louisville, Notre Dame, Penn State, and St. Mary's. It's really fine that they aren't in the tourney.

    What major conference teams got a real gift from the committee? Again per KenPom, Providence is about the only truly eggregious selection. And if you want to complain about Syracuse, Alabama, UCLA, Oklahoma, and Arizona State, fine. But if you're really trying to be objective about this, you aren't replacing those teams with MTSU. You're replacing them with Penn State, Notre Dame, Louisville, Baylor, USC, Maryland, and St. Mary's, almost exclusively other major conference teams.

    So yes. St. Mary's got fucked. Let's go die on that hill. MTSU? No. And Nebraska? Get out. I'm never gonna feel bad if the #57 team on KenPom gets "snubbed" from the tourney. Come on.

    You want to look at things other than KenPom? Cool, that's fine. Again, I'm not saying KenPom has to be gospel. But his rankings are probably *close enough* to an objective assessment that people shouldn't be getting their dander up if a marginal team gets snubbed in favor of another marginal team.

    MaizeAndBlueWahoo

    March 12th, 2018 at 1:07 PM ^

    Agreed, MTSU is absolutely, positively not an NCAA tourney team.  They didn't do a damn thing.

    I would've liked to see St. Mary's win their tournament and push out a Syracuse or Arizona State.  But I can't really justify dying on that hill either, not after flaming out the way they did in their conference tourney.  I sympathize with the idea that people won't play them because they're dangerous, but their preseason tourney was a disaster too, and against bullshit competition.

    The real issue is that the bubble sucks this year.  When Oklahoma and Texas are comfortably in, and Syracuse and ASU are pretty much legitimately the last four in, the bubble just sucks.  I guess Penn State got hosed?  But then you're basically saying that making Ohio State your bitch - their only real accomplishment - is better than beating Kansas at Kansas (which is ASU's big accomplishment.)  Matter of preference, really.  IMO, none of the snubs really have a strong case.

    PurpleStuff

    March 12th, 2018 at 1:41 PM ^

    The thread Brian posted on Twitter was interesting because it illustrated committee moving goalposts for mid-majors. MTSU scheduled good teams and didn’t have bad losses, but no big wins against major conferences. Monmouth has schedule and big wins, but kept out for bad losses. St Mary’s gets kept out for schedule though. And we aren’t talking about unknown commodities here. MTSU won a game last two years. St Mary’s won one last year and went to Sweet 16 a few years ago. K-State didn’t schedule anybody. Lots of teams have bad losses. And a team like Virginia Tech only has a good win because they get to play Duke/UNC/UVA a bunch of times. But those teams all got in easily.

    TrueBlue2003

    March 12th, 2018 at 4:15 PM ^

    home to Belmont and Marshall are both Q3 losses.  Neutral to So. Miss is Q4.  That's three bad losses, one of which was very bad.

    They could have overcome that had they beaten just one of the tough teams on their non-conference schedule but they didn't (Murray St. doesn't really count).

    Scheduling tough games was good.  If the committee has established that's a necessary condidtion for getting in as a bubble team (which they seem to have done, and that's a good thing), it is certainly not a sufficient condition for getting in.  Gotta win enough of the games on your schedule as well.

    MTSU did not do enough winning.

    PurpleStuff

    March 12th, 2018 at 4:29 PM ^

    Scheduled one play-in team and lost. 0-7 against top of their league. “Murray State doesn’t coun’t” and “they didn’t do enough winning” are just arbitrary opinions, especially when teams who did far less get in simply because of an abundance of league opportunities against teams that are just assumed to be good.

    TrueBlue2003

    March 12th, 2018 at 4:57 PM ^

    because they did enough with their schedule to not expose themselves to the scrutiny of a weak non-conference schedule (not unlike Michigan and Michigan State).

    The committee doesn't assume anyone to be good or bad.  They actually use unbiased statistical rankings to help them.

    Per those rankings, KSU truly did not have any bad losses.  Again, MTSU did.  They had three.

    Kansas State certainly played with fire with their non-conference scheduling.  But they did enough in conference (10-8) and avoided bad losses to stay out of harms way.  As for "assuming" those teams are good, again they use metrics that remove the need for humans to necessarily assume.  You can argue those metrics all you want, and they're flawed for sure, but they're there and they are unbiased.

    And you're correct, yes, per the metrics Murray State was a Q1 win, so edit my previous comment to "they only won one of the tough non-conference games on their schedule.  They could have been in had they won one more."

    PurpleStuff

    March 12th, 2018 at 5:11 PM ^

    Good and bad. Those rankings have MTSU 20 places higher in RPI and 8 places lower in KenPom. In both MTSU is higher than a bunch of teams that got in. Again, saying they “did enough” by going 11-9 against Big12 is an opinion based on perception of that league, not unbiased statistics. And that is the point. Rationalizations about best win, worst loss, scheduling, valuing RPI/metrics, ignoring them, etc. were used to justify just putting in everybody in the ACC/SEC/Big12. Committee ignored numbers to put in a bunch of 50-60+ RPI teams at the expense of teams with a track record of actual tourney success.

    TrueBlue2003

    March 12th, 2018 at 5:47 PM ^

    1) Conference strengths are actually determined by unbiased metrics.  Kenpom has the big 12 ranked first by a significant margin.  All of the unbiased systems do.  And that doesn't even matter when conferences have uneven scheduling but in the Big 12 where everyone plays everyone twice, it does mean something.  Going 11-9 in that conference is impressive. As proof:

    2) You should familiarize yourself with another metric that is on the team sheets and the committee is using more and more.  Strength of Record is a metric that basically determines how impressive your record is given your schedule, i.e. what is the likelihood an average team had your record with your schedule.  Kansas State is 23rd in that metric. In the country. 

    Again, they had no bad losses.  You like to keep pointing out that they went 0-7 against the best three teams in the Big 12 which they did but that means they went 11-2 against the rest of the conference, which per SoR (not my opinion or anyone elses), is more impressive than going 0-3 against Kansas and 0-2 against both WVU and TTU is unimpressive.

    The committee didn't ignore numbers, they've shifted what they're using to include more metrics (most of which are better than RPI).

    Your assertion that they favored teams in the ACC/SEC/B12 is disproven by the exclusion of Louisville (38th RPI), ND (kenpom 31) and Baylor (kenpom 34).  There is less rhyme or reason to your argument than how the committee included teams this year.

    PurpleStuff

    March 12th, 2018 at 6:09 PM ^

    The committee obviously didn’t give a single fuck about that. ND and Baylor had a massive disparity in KenPom and RPI. Louisville got left out because of the FBI. Only letting in 7-8 teams while Pac12 is a one bid league and 13-5 in B1G is left out doesn’t prove what you think it does.

    TrueBlue2003

    March 12th, 2018 at 6:33 PM ^

    than KSU in every metric out there.  They did give a fuck about it. They did give a fuck about all of it. 23 v 31 is a big difference when there's literally nothing else suggesting you should be in. Nebraska beat one tournament team that one time.  They had a grand total of one Q1 win and two Q2 wins.  They had no argument whatsoever to be in the tournament.

    DoubleB

    March 12th, 2018 at 1:46 PM ^

    I'd like to see mid-majors with very few opportunities to impress out of conference get credit for going 28-5 than see another Power 5 team with a below .500 conference record get yet ANOTHER chance to play. Arizona State finished 8th in a conference that got 3 teams in the tournament.

    KenPom has its merits, but considering it's possession based and the judging criteria is game-based (win-loss) it has its flaws as well.

    J.

    March 12th, 2018 at 1:49 PM ^

    Wins and losses are ignored in KenPom's formula.  All that matters is the efficiencies.

    Being possession-based is a major strength, because it eliminates some of the ridiculousness the raw numbers have when it comes to tempo.

    In reply to by J.

    DoubleB

    March 12th, 2018 at 8:07 PM ^

    that theoretically determines who the best at-large teams are is based on who you played, where you played, and if you won. That's it. That's how the committee attempts to parse these bubble teams.

    KenPom doesn't care about that. It's attempting to find the most efficient teams in the nation on a per-possession basis adjusted on who you played. And while I think that's a very good PREDICTIVE model, it isn't always a REFLECTIVE model of who won and who lost.

    As an example: Oklahoma State beat Oklahoma in 2 of 3 games, was 4-3 against the top of Big XII (KU, WVU, and Tech) versus OU's 2-4, had a better overall conference record (albeit slightly) and each team had one non-conference win against a tournament team. By any objective measure, OSU's resume is better than Oklahoma's. But OU is 9 spots higher in KenPom. OSU didn't blow out their bad non-conference schedule by enough on a per possession basis, yet they still won those games. 

    J.

    March 12th, 2018 at 8:37 PM ^

    Because win/loss record is an extremely noisy stat.  If there are 68 possessions in an average game, using possession data instead of win/loss data is 68x more selective.

    That doesn't mean you should throw out win/loss data entirely -- nor does it mean that you should use a predictive model to select tournament teams -- but it does mean that one should be careful making arguments that depend exclusively on win-loss record.

    I would actually be quite happy with a compromise -- use a reflective model like Elo to select the field and then a predictive model like KenPom to seed it.

    As for OU vs OKST:

    https://crashingthedance.com/teams/OKLA
    https://crashingthedance.com/teams/OKLAST

    OKST with 5 Q1 wins, 3 Q2 wins, 3 Q3 wins, and 8 Q4 wins. (19-14 total)
    OU with 6 Q1 wins, 3 Q2 wins, 3 Q3 wins, and 6 Q4 wins (18-13 total)

    In a weird way, OU's profile is better than OKST because they got 2 wins from sweeping TCU whereas OKST took 2 of 3 from... OU.

    PS -- RPI is terrible, but it's a known quantity.  OKST finished 88th.  If they had played DII schools instead of Houston Baptist and Mississippi Valley State, they would have been 68th.  They probably could have scheduled themselves into the dance.

    Zenogias

    March 12th, 2018 at 2:07 PM ^

    I understand this point of view. I definitely think the tournament would be more *interesting* if you had St. Mary's and MTSU and other better mid-majors over the major conference teams on the bubble. But I'm pretty sure "interesting-ness" isn't the criteria the committee is using. As long as they are supposed to be looking at team quality, that's what I think they should be judged on.

    DoubleB

    March 12th, 2018 at 7:55 PM ^

    I actually happen to think that Middle Tennessee and St. Mary's are better than Syracuse, Oklahoma, and Arizona State. I think teams that finish 28-5 or so in a high-major conference (as C-USA and WCC would be classified) shouldn't be dismissed due to lack of opportunity.

    Where Oklahoma was seeded meant they MIGHT have been able to finish 2-13 down the stretch and 16-15 overall and make the tournament! Oklahoma won one non-conference game over a tournament team this year (Wichita State). Somehow going 8-11 in the Big XII is like navigating the NBA Western Conference.

    Arizona State finished 8th in a conference that got THREE teams in the tournament. 

     

     

    TrueBlue2003

    March 13th, 2018 at 1:27 AM ^

    opportunities to impress out of conference is 100% their fault.  High majors would fall all over themselves to host SMU.  Those are exactly the type of games that are in high demand by the majors that are smart enough to game the RPI: winnable games against teams that will have ridiculously good records.

    They refused to leave home for some reason. The only away game they even played in the non-conference was across the bay in San Jose!  When you're in the West Coast conference and hoping to impress the committee, you have to be willing to travel.

    I applaud the committee for using bubble tiebreakers as incentive for scheduling more entertaining basketball games.  They also punished PSU, Nebraska, Louisville, Ok St. and Baylor for this.  So it's not exclusively a low major problem.  The committee has made it clear: schedule good baskeetball games when you can control it.  I like that they incentivize scheduling better basketball games. 

    maize-blue

    March 12th, 2018 at 12:27 PM ^

    I don't like the potential UNC matchup. An exit in the Sweet 16 would seem like a downer after the momentum the team had built.

    The winner of that game is in the Final Four.

    mGrowOld

    March 12th, 2018 at 12:32 PM ^

    I remember losing to those SOBs seemingly every year (except one).  Got knocked out in 87 & 88 in the Sweet 16 and then again in 93 in the finals.

    Only year we beat them?  1989 and that tournament run ended well so I choose to take that as a sign if we beat them again.

    J.

    March 12th, 2018 at 12:51 PM ^

    Anyone else looking for a nice omen will note that KenPom gave Michigan a 5.1% chance to reach the championship game in 2013 and a 2.1% chance to win it all.  This year's numbers are 5.8% and 2.4%, respectively.

    These numbers look even better if they win the first round game.  In 2013, they had an 86.6% chance to win the opener, meaning that their conditional probability to win the championship was 2.4% if they won the first game -- that game didn't move the needle much since it was such a likely win.  This year, he has Michigan as only a 76.5% favorite -- shockingly low; the next two teams with a lower first-round win probability are 5-seeds OSU and WVU.  However, that means that their KenPom chance to win the title, assuming they beat Montana, is 3.1%.

    #It'sHappening!
    #Math'd