So you want to win that billion-dollar bracket…

As I’ve been crunching the team numbers in anticipation of the bracket announcement (nine teams to go!), that Quicken Loans billion-dollar bracket has been rattling around in my brain. Part of the reason is that the digital marketing company helping QL and Buffett promote the contest has been hounding me to post a completely worthless infographic on how to pick a perfect bracket (“work backwards…pick a 5v12 upset…guard play is key…be wary of 14-16 seeds”—yeesh).

Let’s face it: even if you were 70% correct in all your picks, your odds of getting the bracket right would be about 1 in 580 million. Ain’t happening.

Still…we’re all going to try, aren’t we? And my best advice is this: get crazy within reason. No, do not pick a 16 seed to knock off a one seed. And, despite the last couple years, I’m not picking a 15 to shock another two. But a 3v14 and/or 4v13 upset if fair game—at least in this “shoot the moon” bracket.

One thing you should definitely do is make your bracket look like other brackets from the 29-year, 64-team tourney. Since 1985, there have been an average of 8.7 upsets per dance. I define an upset as any game in which a team seeded four positions or lower surprises a favorite. As you can read in the Insider 2014 Upsets feature, most of these upsets occur in the first two rounds of the dance. There are an average of 4.4 upsets in the opening round, 2.7 in the second round—and only 1.5 thereafter.

If you’re wondering whether you should go with more or fewer than 8.7 upsets, the following chart might help. It shows the number of times over the 29 years that a tourney has seen 3-5 upsets, 6-7 upsets, 8-9 upsets, 10-11 upsets and more than 11 upsets:


It’s worth noting that the last four years have been the most upset-laden in the modern era, with 11 in 2010, 13 in 2011, nine in 2012 and 13 last year. It’s a small sample size, but that’s an average of 11 shockers per dance.

Whether you think this year will have more or less than the average upsets, you should at least make sure that your billion-dollar bracket looks like something the could actually happen—and hope like heck there are no Florida Gulf Coasts lurking out there.

This entry was posted in Basic Concepts, Tourney Trends, Upset Spotting. Bookmark the permalink.

22 Responses to So you want to win that billion-dollar bracket…

  1. Stevie R says:

    I was reviewing my strategies for this year and came upon a question after looking at Final 4 Based on Seed Qualification and Conflict Rules.” In the article from last year it determines that the “Final Four using those statistics and rules would have correctly picked Louisville, Syracuse, and Michigan as Final Four participants (Indiana was the fourth, not Wichita State)

    Looking at the model or bracket for 2014 F4/Champ it showed Louisville vs Ohio State and Kansas vs Indiana with Indiana defeating Louisville in the championship game. Any clarification would be greatly appreciated on why the difference.
    I have been a member of BracketScience for several years and have been very successful using your insight and statistics (I use a combination of the top 3 performing models over the past 2 years to determine my Final Four participants)

    • ptiernan says:

      I tweak the F4/Champ model every year. So that explains it. I’ll be buried from now until late Monday afternoon with number crunching.

      • Ryan Tressler says:

        First thing I am doing is going through and doing the Seed Matchup and upset/toss-up models again, figure out what they see this year, if anything haha . . . always my first step in my try to be perfect bracket I make every year, haha (as opposed to my win my pool bracket, which looks much different, haha)

  2. Ryan Tressler says:

    if you are truly going for the perfect bracket, it might be important to consider that on average, 1.69 out of the 12 teams slotted as a 13/14/15 seed advance past the first round (aka, round of 64, the true first round, haha) . . . if you expect this to be a crazy year, might be worth picking at least two of those teams to pull an upset

    • Ryan Tressler says:

      another interesting thing to consider for a perfect bracket: 26 times in the last 29 years, a 12 or 13 seed has made it to the sweet 16

      • Bullets-and-Blazers!! says:

        Exactly Tressler and now it’s time to pick the 13 seed over 4 seed. Will it be “legendary” Steve Alford and UCLA? How about San Diego St. vs “the sheriff” (7’5” Sim Bhullar) New Mexico St.? How about apprentice/student vs. mentor in Masiello vs. Pitino? One not to pick is Izzo’s squad (forget it, Delaware doesn’t stand a chance).

  3. Paul says:

    anyone know after champ week who still meets the champ criteria?

  4. Paul says:

    Who do you guys think has the best Front court this year?

    • John says:

      New Mexico and Tennessee come to mind. Michigan St. has quality and depth in the front court.

    • Bullets-and-Blazers!! says:

      Tough call. Iowa State has offensive fire power on the frontcourt; Arizona has balance strength on the frontcourt as well; North Carolina too; and then there are some mid-major teams in New Mexico (mind you there are not atheletic, Kirk is pretty slow and Bairstow moves awkwardly at times); San Diego St. where they start four fowards and one guard.

  5. Daniel says:

    Anyone know the best place to get stats until Pete releases the monster spreadsheet?

  6. Bullets-and-Blazers!! says:

    I will admit: I am going for the perfect bracket! I want that billion dollar but thanks to the genius selection committee and their famous mis-seeding teams, it won’t happen. Seriously? Michigan State AND Louisville were given #4 seed? Kansas given #2 seed despite “injury will affect the seeding” rule? Well, I will start a conversation and hopefully Tiernan, Andy and Tressler add info here.

    Tiernan says that previous champs were either 1, 2, or 3 seed. Let’s separate contender/pretender/not enough

    Contender: 1 Florida, 1, Arizona,

    Pretender: 3 Syracuse, 2 Wisconsin, 3 Creighton, 1 Virginia, 2 Villanova, 3 Iowa State (their defense stinks!!!)

    Not enough to make to the Final Four: 2 Kansas, 1 Wichita St., 2 Michigan, 3 Duke

    Oh yeah, we are in for a very interesting Finar Four!!! HA-HO-HA-HO-HA-HO-HA-HA!!!! THE COMMITTEE DID IT AGAIN!!!!!!!!

    • Bullets-and-Blazers!! says:

      Scratch that, 1 Wichita St. may waltz to the Final Four. Which region is a bust (where neither 1, 2, 3 seeds will make it to the Final Four)? I like the East Region to be the bust, but are there any more? I have my eyes on the West (Arizona does not convince me to go to the Final Four, but I could be wrong).

      • Andy says:

        I copied the stats from last year and did my best to replicate the stats for this year so I would have a head start, and I have the Final Four model going with Florida, Arizona, Duke, and Villanova. It’s a compelling Final Four, with all four squads in the top ten for Pythag. Florida, Arizona, and Villanova are solid Final Four contenders, while Duke got in there despite not being a contender because nobody else was a contender.

        One thing about Florida is that while they pass the Final Four model, they actually fail according to the Upset model, due to their low scoring (aided by their slow tempo). No way they miss the Sweet 16, but my eye has been on VCU from the get go. VCU is interesting actually, because UCLA passed the Pretender/Contender for the Sweet 16, while VCU failed. To be fair, VCU’s offense is shaky, but Steve Alford is one of the most reliable underachievers out there. Furthermore, I think I calculated that VCU has too much guard play to pull the upset. The other side of the South is a mess, with Kansas and Syracuse both showing as pretenders and Ohio State as a 6-seed contender, but with the weakest offense among the 6’s and failing the Seed Matchup model against Dayton due to scoring less than 70 points per game. And come on, do you really see New Mexico, Stanford, or Dayton upsetting Florida? I don’t. I will side with that model on this pick, and not the Upset model. Leaning toward Stanford in the Elite Eight though, as they aren’t in the Mountain West (ahem, New Mexico) and have a significantly better Pythag rating than Dayton.

        No surprise on Arizona. Not only that, they also got quite an accommodating draw to go far. Oklahoma State is probably the best hope for them to lose before the Elite Eight….not saying much, huh? Bad luck for the Pokes will keep that from happening. Oklahoma is also the weakest 5 seed, and San Diego State the weakest 4, so we could easily be looking at North Dakota State or New Mexico State as Arizona’s challenger in the Sweet 16. And about the only teams I can give any consideration to for the Elite Eight against the Cats are Creighton and Baylor, and both lose the turnover battle, but Baylor by more (I also trust a team that went to last year’s tournament more, and I’m not 100 percent sold that McDermott is legitimately snakebit), and are thus horrible matchups for Arizona’s top-rated defense per possession. Wisconsin, Oregon, and BYU have either snakebit or reliably underachieving coaches, while Nebraska has the lowest victory margin of any 11 seed. Yeah, can’t pick another team but Arizona here.

        Villanova will have to weather a tough draw because UConn is the best 7 seed. Then, assuming they do so, they could get a decent 6 seed in North Carolina. That said, UConn is not a great rebounding team and North Carolina lacks momentum. Yeah, Villanova lost to Seton Hall, but disregard that for a second and realize the completeness of their team that could lead them for a long run to the Elite Eight. Meanwhile, Virginia is a huge pretender as a 1 seed, likely losing to Michigan State. However, the Spartans did fail the Pretender/Contender model (their assist-to-field goal ratio is too high, above 60%), and get a tough draw potentially in Harvard should both advance. Then again, Tom Izzo always has a way of beating teams like Harvard, and the Crimson had a very weak SOS. But against Villanova? Gotta go with the Wildcats on this one.

        Finally, the Midwest. Kentucky over Wichita State is tempting, but inexperience and too much youth will likely cripple the Cats here, though they should be K-State. I am leery on Louisville’s strength of schedule, but they certainly won’t lose to Manhattan, and shouldn’t lose to stumbling St. Louis, questionable NC State, or inexperienced Xavier. Strength of schedule has to doom Wichita State at some point, and Louisville is better per possession overall, so I can’t imagine Louisville losing to them either. However, bad luck and lack of quality non-conference wins will get to Louisville before the Final Four, too. I am not high on Michigan, but they too won’t lose to Wofford, Texas, or Arizona State. Meanwhile, Duke gets a tough 11-seed draw in Tennessee or Iowa, but neither has the luck or momentum I would like to see for the upset (I’m disregarding UMass as I think they are out quickly to the winner of that game). Michigan vs. Duke will be a shootout, but Duke is slightly better at forcing turnovers and rebounding. Duke vs. Louisville will be an awesome game, as Duke doesn’t have much defense while Louisville has that and talent, but I’m concerned that Louisville is not prepared enough to take that next step, even after last year. You really have to prove it in the non-conference, and they went 0-2 there against quality opponents, who are 6 and 8 seeds on top of it and Pythag rankings of 17 and 26. Duke wins this region.

        • Bullets-and-Blazers!! says:

          You don’t think Mercer has a shot against them? I have my eyes on Mercer for a while and I think they can pull the upset. Albany had size but their big players were pretty raw yet they only lost by 12. It was Albany’s guard that got them close. Mercer has great guard play and they have size. But, I do see Duke avoiding it too. Syracuse vs. Western Michigan is a very interesting draw. Back in 2010, Ohio stats didn’t wow anyone but Georgetown itself had some issue inspite of three McDonald All-American in their roster. Syracuse lost to Boston College, out of any team, and they struggled down the stretch. I am very skeptical with Villanova because of their lack of perimeter defense and I do not trust Jay Wright. As for Manhattan over Louisville, I see it happening. They are my 13 over 4 team (the student will surpass the master; Louisville lost to Memphis twice and the AAC isn’t really a powerhouse conference). New Mexico State over San Diego St. I do not see it (last march, St. Louis destroyed NMSU in the tourney.

          • Andy says:

            Mercer is barely in the top 100 on the Pythag standings. I usually don’t consider a team rated that low seriously for the upset. They also are below the top 100 in both offense and defense. I think it could be a close game, as I think Mercer is otherwise a good team, and I remember Duke winning by one point at home against Vermont. And Duke’s stats look like Missouri’s from a couple years ago when they lost to Norfolk State. However, they have played a much tougher strength of schedule, and Coach K is far more trustworthy as a coach than Frank Haith ever was that year. Decent game, but Duke wins. I would consider it a bit more if Mercer was playing Iowa State, I think.

          • Andy says:

            As for the others, can’t go with Manhattan over Louisville. Even with Louisville’s shaky SOS, a 20+ margin of victory has me sold on a Sweet 16 run, especially with St. Louis, NC State, and Xavier in their way (hardly worthy competition). Also I hear you on New Mexico State, but I think they do pass the upset model while San Diego State fails it, so it could happen. That said, they do have a problem winning in the tourney, and more likely to me for a 13 over 4 is Tulsa over UCLA. Conference USA has been pegged all year as potentially producing a giant killer, and I really do not trust Steve Alford. I am miffed that UCLA passed some Sweet 16 contender models while VCU failed them, when VCU has the better coach and better Pythag rating.

            I also understand what you are saying about Villanova, and a few people are questioning why I would take them all the way to the Final Four. He did also last overachieve in the dance in 2009, and has only overachieved three times out of ten appearances. However, this is still the best of the 2 seeds, and not only passed the Final Four model, but also the upset rule (then again, so did both UConn and St. Joseph’s). I think the push for Nova in their first game will be that they have experience overall that the other two teams do not have. St. Joseph’s is also a weak 10 seed according to Ken Pomeroy. I also can’t imagine Iowa State in the Sweet 16 with their weak defense and weak rebounding, and Providence is a weak 11 seed with a very low victory margin. North Carolina is much more compelling, but with two straight losses coming in, I can’t take them past the Sweet 16. One thing too about Roy Williams that is very unknown is that he actually has only once reached the Elite Eight seeded lower than 2, as a 3 seed way back in 1991. Further, for as weak as Villanova’s perimeter defense is, North Carolina doesn’t take that many threes to begin with anyway, and is thus a great matchup for Villanova if they get them. Therefore, after analyzing everything, I have no choice but to take Villanova to the Elite Eight for now. Michigan State is another matter, and Tom Izzo is due and got a great draw to go that far, so I will take another look at that matchup. I do know however, that they have a very high assist-to-field goal ratio and I think much of their scoring is from guards. They did fail some models based off of this that suggest they could lose before the Sweet Sixteen, but I’m not buying it with Izzo. Couple that with Virginia as a bogus 1 seed in my opinion (not that they didn’t deserve it, but they are just so weak as a 1!), and they really have no business not making it to the Elite Eight.

            One thing I agree on all accounts, is that three models pointing to Villanova as the championship is not telling the whole story. Indeed, they aren’t tough enough to do that based on their blowout losses. However, I’m not just going to take Michigan State because “everyone else is.”

          • Andy says:

            After reviewing Villanova and their draw, I still feel they have a good chance of making it to the Elite Eight. However, from a coaching standpoint, it is hard to overcome a 30% overachievement rate against a wizard like Tom Izzo. Michigan State is also a top 10 Pythag team, and is rebuilding its momentum. I wonder if their guard play and assist-to-field goal ratio could be a function of the injuries they suffered. Villanova has luck on its side, but I have kind of wondered all year if they have too much of a good thing. Michigan State’s is much more neutral. On top of it, I would probably go with Florida in the final over Villanova, and I’m not really that big on such a scenario like I am with Michigan State. I still think the Gators get there, though. Just a really easy draw for them, with the possible exception of VCU.

            I do concern about Michigan State winning the championship against Arizona, however. Sean Miller will have his team ready, and he simply doesn’t lose as the higher seed. I also wonder if they could end up with not one but two All-Americans. I know Aaron Gordon is pretty much a lock, but I wonder about Nick Johnson, too. I’m also concerned that with so much guard play and lack of post play, that Michigan State might not match up well with Gordon or with Kaleb Tarczewski, one of the better big men this year, even though he isn’t likely to be an All-American.

          • Bullets-and-Blazers!! says:

            Andy I can’t go with Tulsa over UCLA because UCLA is the PAC-12 champion (Big six, except for Missouri in 2012 which they were handed the Big 12 crown, always advance to the next round). The deliberate seeding of #4 and pairing them with Mansiello (whom knows Pitino) is more than enough to convince me that Pitino may very well throw the game away. And a 13 seed vs 12 seed (5th seed St. Louis) further convince me that they are a bust, regardless of what their stats shows. I highly doubt we will see Wichita vs Louisville rematch. Remember what I wrote about back-to-back Final Four appearance and three out of the four Final Four from last year are at the same region Wichita St., Louisville, and Michigan. Michigan is incomplete, Louisville aready won the championship, so that leads us to Wichita state. Kentucky to upse them? Ehh, Julius Randle might not play to his potential and blow the game to Kansas State so he can enter the draft uninjured.

            Note: I am not using only stats, these players are young kids playing ball. Anything can happen.

  7. Dave says:

    Intial thoughts for me:

    South- correct me if I’m wrong but in the seed quite for 1 v 8 matchups Florida qualifies as a victim and Colorado qualifies as a giant killer. Hard not to pick Florida overall here but I can’t get myself to pick them to win this region. For some random reason I think Dayton has a major shot at pulling a deep tourney run. Great rivalry game against OSU should fire them up for the rest of the tournament. I do think that the winner of the Dayton Syracuse game, if it happens, will go to the Elite Eight. As for Kansas… they better watch out for EKU but I think they’ll find a way to get past them. I can’t quite figure out how UNM and Stanford will fare against each other. I think both have a shot at beating Kansas.

    East- this region could get crazy in my opinion. First off, I have GW beating Virginia. I hate doing this as it makes it easier to pick MSU to the final four but I have a gut feeling GW could pull ofd the upset. Its also a great value pick over Memphis. I have Harvard over Cincinnati as (I think) Harvard in the seed quite qualifies for the upset and Cincinnati also as a victim. MSU… someone please convince me there’s a better team in this region… Everyone will be picking them to the final four. I’m afraid there’s no other team to pick be process of elimination. For UNC Providence I’m not quite sure but leaning towards Providence. I think either will lose to Iowa State. For the bottom quadrant, I think the winner of the St Joe’s UCONN game will take down Villanova. In my opinion, Villanova is the most vulnerable team in the tournament. They were blown out by Creighton twice and suffered the agony of an epic buzzer beater in their conference tournament. For the elite eight, I guess I’m forced to pick MSU and then in the bottom half either Iowa St. StJoe’s or UCONN.

    West- I see a few upsets here however nothing major. I think Arizona, despite being a top seed, will be a great value pick to reach the sweet 16. A lot of people will be picking OSU to defeat Arizona. I don’t think it’ll happen. Not quite sure what to do in the next quadrant as I think OK, NDSU, SDSU, and NMSU would all lose to Arizona. The upset in this region would be, for me, a Oregon v Baylor sweet 16 matchup. Both are hot and Baylor will be playing Creighton in Texas. Wisconsin matches up terribly with Oregon as they are complete opposites. The toughest past for Oregon will be going into Milwaukee. I think Baylor reaches the elite eight and Arizona takes this region (although still considering Baylor)

    Midwest- the strongest region by far, I see very few upsets here. Wichita St, like Arizona, will be a great pick to reach at the very least the sweet 16. Many people will pick Kentucky. Louisville will clearly tech the sweet 16 also and will face Wich St. in a rematch of las year’s final four… I’m leaning again towards Wich St. for the great value. In the bottom half Umass is a weak 6 and Tennessee looks like a nice upset pick. Does anyone know if our brackets must be submitted before or after the play in games? For me, I’d pick UMass if Iowa wins but Tennessee if Iowa loses. Either way I think they will all lose to Michigan as Michigan will take this region.

    Obviously an incomplete bracket but I’m fairly confident in a few of my picks. If only Michigan won today I would’ve pegged them easily to win it all… but they didn’t. Thoughts?

Leave a Reply

Your email address will not be published. Required fields are marked *

* Copy This Password *

* Type Or Paste Password Here *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>