The 2008 and 2007 NCAA tournament will be remembered, as much as anything else, for their uncommon predictability. The 2007 dance featured only three upsets, easily the fewest of the 24-year 64-team era. And 2008 marked the only tournament since 1985 to have all top seeds reach the Final Four.
What do two consecutive unpredictably predictable tourneys portend for 2009? My guess is a return to maddening unpredictability. And that means that the value of picking upsets in your bracket will be more important. Maybe not George Mason important, but pretty darned important.
Let's face it: The timing is right for another tourney of double-digit upsets. It's very likely that your bracket pool will be won by the person who can identify the right favorites to fall and long shots to advance. Of course, Cinderella-spotting is tricky business. Settle on the wrong high-seeded victim and your bracket could collapse in the first weekend. Good thing that picking the right underdogs and paper tigers isn't all guesswork. The upset victors and victims of the modern era have shared common attributes. When you know what they are, it's a lot easier to sniff out the upsets. Let's examine the factors that correlate with upsets. Right after Selection Sunday, I'll post a blog that identifies the 2009 dark horses and vulnerable powerhouses. (For a list of all the Cinderellas in the modern era, check out the historical brackets at www.bracketscience.com/history.)
When is a win an upset?
Not every game in which a lower-seeded team knocks off a higher seed is an upset. Nobody's going to fit a glass slipper on a ninth seed that beats an eight in Round 1. (Actually, nine seeds are 52-44 against their higher-seeded opponents.) It's only when you get a gap of at least four seed positions between opponents that a game has upset potential.
Surprisingly, two-thirds of tourney games meet this condition. Of the 1,512 games that have been played in the last 24 years, 1,006 of them have pitted longshots against favorites -- and the underdog has won about 20 percent of the time. That's an average of 8.5 upsets per tourney, or roughly one in every seven games. (Now you know why 2007's total of three upsets was so unusual.) This chart shows the round-by-round results of upset games in the 64-team era.
Because of the way the brackets are set up, most of the upset matchups occur in the first three rounds of the tourney. Of the 1,344 games played in the first three rounds, 965, or 72 percent, have been upset matchups. Picking these upsets correctly -- or at least minimizing the number of victims you advance -- is essential to building a winning bracket.
While first-round upsets grab most of the attention, the upsets in the second and third rounds do the most damage to your bracket. Only 19 of the 100 opening-round upset victims are teams seeded one, two or three. More important, the top three seeds are a dominating 269-19 (.934) against first-round darkhorses. Only about one in 15 top-three seeded teams will fall victim to an upset in Round 1. To put it another way, fewer than one top-three seeds will lose per tourney. The last victim was third-seeded Iowa, which was humbled by No. 14 Northwestern State in 2006.
It's a different story in the second round, where the top three seeds are just 163-52 against Cinderellas. That's a solid .758 winning percentage, but it's nowhere near the lockdown .934 rate of the opening round. Instead of only one in 15 teams losing, the top three seeds lose one in every four games -- and between two and three per year. They also make up 78 percent (52 of 67) of the second-round upsets. In 2008, there were two high-seeded upset victims: Second-seeded Duke lost to seven seed West Virginia, and two seed Georgetown lost to 10th-seeded Davidson.
The Sweet 16 isn't as treacherous for the top three seeds as Round 2, but it isn't a first-round cakewalk either. Of the 27 third-round upsets of the modern era, 17 (or 63 percent) have involved teams seeded one, two or three. But they do have a .832 winning rate against longshots (84-17). In fact, upsets among the top three seeds are so rare that fewer than one occurs per tourney. Last year, there were five long shots (Michigan State, West Virginia, Davidson, Western Kentucky and Villanova) -- and only Davidson was able to spring an upset.
Upset prediction value: Balancing forecasting accuracy with frequency
Before we take a deeper dive into the anatomy of upsets, you need to ask yourself: What kind of an upset spotter do you want to be -- accurate or prolific? You could create an elaborate rule that has never failed to predict an upset, but the rule would almost certainly apply to just a handful of games. By the same token, you could pick every single upset -- if you're willing to be wrong 80 percent of the time -- and end up dead last in your pool.
In determining which factors have the biggest influence on upset prediction, you have to balance how much the rule increases the odds of picking an upset (accuracy) with how many upsets it describes (frequency). Here's an example: Long shots have won 17.2 percent of the 576 games in which an upset could happen in Round 1. These giant killers are rarely 15 or 16 seeds. By eliminating these teams from consideration, you improve your chances of picking an upset by 44 percent -- a 25 percent winning rate (91-277) vs. 17.2 percent. Just as important, 96 of 100 first-round Cinderellas satisfy this rule. By multiplying the increase (44 percent) above the typical upset rate by the percentage of upsets described (96 percent) we arrive at the upset prediction value (UPV), for the "15- and 16-seed exclusion" rule: 42.2. When you evaluate rules by their UPV, you're better able to compare their relative worth in predicting upsets.
Six sorta simple guidelines for Cinderella spotting
If the deadline for finishing your bracket is approaching, you might not have time to dig into the nitty-gritty of upset probabilities. Still, it's worth knowing a few general guidelines to boost your odds of picking an upset. Remember these six rules and you'll dramatically improve your odds of spotting an upset.
1. Don't pick any longshots lower than a 13 seed: Sure, the bottom dwelling seeds do spring an occasional upset, but it's at a much lower rate than higher-seeded teams. These long shots are 21-255 (.076) against opponents with a seed position at least four rungs higher than them -- hardly worth risking your bracket on.
2. Never pick a top seed to be an upset victim: You don't have to look any further back than UConn's loss to 11th-seeded George Mason in 2006 to know that top seeds can be toppled. But it's not worth predicting. In the 269 games they've played against prospective Cinderellas, one seeds have shattered the slipper 246 times (23-246). That's a 92 percent success rate -- much too strong to bet against.
3. In the opening round, look for senior-led, guard-dominant long shots with a strong starting unit and solid records that handily beat opponents -- but aren't playing favorites too close to home Got all that? OK, so maybe this isn't such a simple guideline ... but it could help point you to the best 2009 underdog candidates. Restrict your choice of first-round Cinderellas to teams that:
• Have a winning record better than .600 and have won more than three of their past 10 games.
• Get more than 25 percent of their points from guards.
• Get more than two-thirds of their points from their top five scorers.
• Score more than 71 and beat teams by more than six points a game.
• Have at least one senior starter.
• Aren't playing a favorite within 300 miles of their campus.
Easy, eh? These teams are a respectable 46-54 in the opening round.
4. In the second round, only consider Big Six Cinderellas with an experienced coach that aren't limping into the dance: Here's a handy guideline: Just avoid mid-major and small conference dark horses, coaches making their first tourney appearances and teams that have won fewer than four of their past 10 games. All other longshots are 25-22 against favored seeds.
5. In the Sweet 16, never pick a two seed to be victimized -- and only pick senior-led, backcourt dominant long shots: You're already steering clear of top seeds. And after the first two rounds, it no longer pays to pick two seeds as victims. They're a dominant 25-6 against underdog six and 11 seeds. By ignoring upset matchups involving the top two seeds -- and focusing on teams that get more than 35 percent of their points from guards and have at least one senior starter -- you'd achieve a 15-14 record in identifying Cinderellas.
6. Don't pick any upsets after the Sweet 16 round: From the Elite Eight on, Cinderellas have a 9-32 (.220) upset record. At first blush, that might seem pretty good. But when you consider that just two teams (Villanova in 1985 and Kansas in 1988) account for a 5-0 performance, you're looking at a more daunting upset rate of 11 percent. If you think you can peg the next Villanova or Kansas, go for it -- but such a team hasn't come along in 20 years.
So how strong of an upset predictor would you be if you observed these guidelines? How does a 49 percent accuracy rate grab you? In upset matchups meeting all the above conditions, the underdog is a reasonable 86-90 (.489) against the favored seed. That's 142 percent better than the typical 20.2 percent upset rate and it describes 42 percent of the upsets (86 of 203). That translates into a strong UPV of 60.1.
Round-by-round rules for predicting upsets
If you're willing to examine potential upset squads more carefully, you can boost your upset-spotting success rate to the point where you're right more often than you're wrong. Here's what you need to do in each round:
Round 1: Take high-scoring 11 seeds, frontcourt-dominant 12 seeds with tourney-tested coaches, and 13 seeds with experienced coaches that soundly beat opponents. Even if you're going to examine the opening round more closely, it doesn't pay to pick a Cinderella lower than a 13 seed. That restricts your choices to 4 vs. 13, 5 vs. 12 and 6 vs. 11 matchups. In those games, look more favorably on underdogs with these attributes:
• Take 11 seeds that score at least 73 points a game and beat their opponents by seven or more points. These teams are 22-19.
• Take 12 seeds with coaches who've been to the tourney at least once before, have a winning percentage higher than .700 and get at least 55 percent of their points from the frontcourt. These 12 seeds are 12-5.
• Take 13 seeds with coaches who've been to the tourney at least once before that have won between seven and nine of their past 10 pre-tourney games and score over 10 percent more points than they allow. These teams are 8-4.
By applying these three rules, you'd be 42-28, or 60 percent accurate, in predicting first-round upsets. That's 246 percent better than the typical first-round upset rate of 17.6 percent, and it describes 42 percent (42 of 100) of the upsets.
If you used these rules last year, you would've gone 1-1 in picking upsets. Both Kansas State and Baylor had the credentials to be 11 seed Cinderellas; only Kansas State sprung an upset, beating USC in the celebrated Beasley-vs.-Mayo match-up.
Upset percentage increase: 245.6 | Upsets described: 42 percent | UPV: 103.2
Round 2: Pick experienced or battled-tested seven through 12 seeds -- but avoid ninth-seeded underdogs. With all apologies to ninth-seeded UAB, which upset top-seeded Kentucky in 2004, there's no reason to pick a nine seed Cinderella in Round 2. They're an abysmal 3-49 against the big dogs. And 13th- through 15th-seeded teams should be ignored as well, since they've only posted a 6-33 record. That leaves seven, eight, 10, 11 and 12 seeds. Here are the upset rules for these seeds in their matchups:
• In 1 vs. 8 games, take battle-scarred eight seeds with tourney experience. Eighth-seeded squads that: 1. went to the tourney the previous year, 2. have a coach who has made at least three trips to the dance, 3. have less than a .700 record, 4. beat teams by eight or fewer points a game and 5. have a pre-tourney losing streak -- they're 7-7. Every other eight seed is 2-28. Last season, no eight seeds met the criteria -- and top seeds won both matchups.
• In 2 vs. 7 games, take seven seeds from the Big Six conferences (ACC, Big East, Big Ten, Big 12, SEC and Pac 10) that beat opponents by at least six points a game. These teams are 9-8. All other seven seeds are 8-34. Last year, both West Virginia and Miami had the traits to spring an upset. The Mountaineers stunned Duke while the Hurricanes narrowly lost to Texas 75-72.
• In 2 vs. 10 games, take 10 seeds with at least four junior or senior starters. They're 10-5 against their favored opponents, while all other 10th-seeded teams are 5-13. Last year, Davidson upset Georgetown -- and they had the sufficient team age to qualify as a Cinderella.
• In 3 vs. 11 games, take 11 seeds that have won fewer than nine of their last 10 pre-tourney games and have a winning record either below .675 or above .750. Teams satisfying these conditions are 8-4. All other 11th-seeded squads are 0-15. Kansas State had the numbers to be a second-seeded giant killer but got overwhelmed by Wisconsin.
• In 4 vs. 12 games, take 12 seeds coming into the tourney on either a losing streak or a single-game winning streak that beat opponents by fewer than 10 points a game. They're 6-1 against four seeds. The rest of the 12 seeds are 4-13.
If you applied all these rules, then you'd accurately predict 34 upsets in 58 games for a 59 percent success rate. That's 134 percent better than the typical 25 percent second round upset rate, and it describes over half (34 of 67) of the upsets. Last year, you would've identified the only two upsets in round two -- the West Virginia and Davidson stunners. Of course, you also would've been wrong on Miami and Kansas State.
Upset percentage increase: 134.5 | Upsets described: 50.7 percent | UPV: 68.2
Sweet 16: Avoid one and two seed victims; pick seventh- through eleventh- seeded teams that aren't struggling or guard-dependent. Seeding is the key indicator of Sweet 16 Cinderellas. Avoid underdogs that face one or two seeds and any long shot seeded lower than 12. All other potential upset pullers are 16-21. If you further restricted your upset choices to teams that have won at least five of their last 10 pre-tourney games and got less than 80 percent of their points from guards, then you'd improve your prediction success to 16-13 (.552). That's 147 percent better than the typical Sweet 16 upset rate -- and it describes 59 percent of the upsets. Last year, both Davidson and West Virginia had the numbers to pull upsets. Davidson surprised Wisconsin, while the Mountaineers fell three points short of Xavier.
Upset percentage increase: 147.3 | Upsets described: 59.3 percent | UPV: 87.3
Elite Eight: Pick high-scoring, backcourt-dominant Big Six conference teams. Picking upsets in the Elite Eight is risky business. Only five of 28 longshots have snuck up on favorites. Eight seed Villanova beat two seed North Carolina in 1985; 11th-seeded LSU knocked off top-seeded Kentucky the following year; Rick Pitino's sixth-seeded Providence Friars beat top seed Georgetown the next year; Michigan's sixth-seeded Fab Five beat one seed Ohio State in 1992; and George Mason was the most recent team to spring an Elite Eight upset when it beat No. 1 UConn. Other than George Mason, the remaining four Cinderellas came from a Big Six conference, scored more than 70 points a game and got at least 40 percent of their points from guards. Elite Eight underdogs with these characteristics are 4-4; the rest are 1-20. A word of warning, though: the sample size on Elite Eight upsets is small. Last year, Davidson didn't qualify as an Elite Eight Cinderella -- and wound up losing to Kansas.
Upset percentage increase: 190.0 | Upsets described: 80.0 percent | UPV: 152.0
Final Four and championship: Pick tourney-tested long shots led by coaching legends. Two teams account for the four upsets that have occurred in the Final Four and championship rounds. Eighth-seeded Villanova beat second-seeded Memphis State, then shocked Georgetown on its way to the championship in 1985. And sixth-seeded Kansas toppled second-seeded Duke and top-seeded Oklahoma to win the tourney in 1988. The other eight teams seeded fifth or lower have all lost in the last two rounds of the tournament. What did Villanova and Kansas have that they didn't? Simply put, experienced coaches with strong programs. Both schools had legendary coaches with at least seven tourney appearances (Rollie Massimino and Larry Brown) who had taken their teams to the Big Dance at least four consecutive years, longer than the impact of any single class.
Upset percentage increase: 200.0 | Upsets described: 100.0 percent | UPV: 200.0
Putting all the upset rules together
If you followed all of the rules listed above for each round of the tourney, then you would've picked upsets correctly in 96 of 165 games, for a 58.2 percent success rate. That's a 188 percent improvement over the typical upset rate of 20.2 percent. And it describes 47.3 percent of the 203 upsets that have occurred in the 64-team era. That works out to an impressive 88.9 UPV -- nearly 50 percent better than the UPV achieved from the six basic guidelines. Let's face it: If you could identify 96 Cinderellas (four per tourney) with 60 percent accuracy, you'd be the envy of your fellow bracketeers, whether you won your pool or not.
The value of indentifying high-seeded upset victims
While it might be a nice ego boost to be an expert Cinderella spotter, it's probably not going to help you build a winning bracket. If your main goal is to climb to the top of your pool, it's more important to know the upset victims to avoid than the perpetrators to advance. Just as Cinderellas possess common traits, there are definite qualities that first-, second- and third-seeded victims possess. Steer clear of the following high-seeded squads and you'll stand a better chance of nailing the Final Four:
Top-seeded underachievers: 27 of 96 one seeds have failed to win three games and reach the Elite Eight. What separates these disappointing top seeds from their more reliable counterparts? If a one seed possesses any of the following attributes, consider penciling them in for a pre-mature tourney exit:
• They've lost two or more consecutive games entering the tourney.
• Their coach is making his first tourney appearance.
• They didn't go to the previous year's dance.
• They score fewer than 80 points and allow more than 65 points a game.
• Their scoring margin is less than 10 points a game.
• They get less than 33 percent of their points from guards.
• They have an older starting unit, with an average class of 3.4 (freshman=1, sophomore=2, junior=3, senior=4)
Top seeds that possess any one of these traits are just 17-18 (49 percent proficient) in getting to the Elite Eight. All other one seeds are 85 percent proficient (52-9) in winning their first three games.
Second-seeded underachievers: 36 of 96 two seeds have failed to reach the Sweet 16. How can you spot the paper tigers? Second-seeded squads coming to the dance with two or more consecutive losses that average scoring fewer than 80 points, allow more than 60 points and win by a margin of less than 12 points a game are just 14-18 (.438) in reaching the Sweet 16. All other two seeds are 46-18 (.718) in winning their first two games.
Third-seeded underachievers: Half of the 96 three seeds never see the second weekend of the tourney. Here's an easy way to spot a third-seeded squad that won't reach the Sweet 16: they average fewer than 80 points and win by fewer than 10. These hapless three seeds are just 26 percent (6-17) proficient in winning their first two games. Other three seeds are a more reliable 42-31 (.575) in sticking around until the second weekend of the dance.
Peter Tiernan has been using stats to analyze March Madness for 19 years. His insights into the NCAA basketball tournament can help you build a better bracket. E-mail him at bracketscience@comcast.net or visit bracketscience.com.
