“Successful people ask better questions, and as a result, they get better answers.” ~ Tony Robbins
Do you want to be successful? Do you yearn for the adulation? It’s time for March Madness – the 2015 NCAA Men’s Basketball Tournament. And you aren’t entering the office pool for fun. You want to win. Or, at least, not lose to the know-it-all three cubes down. Or to your wife. Or, worse, lose to your father-in-law. Actually, even worse, your mother-in-law.
If you want to be successful, you must ask good questions. How often do higher seeds win in the Round of 64? Which coaches have had more tournament success? Is there a trend toward mid-majors in the Final Four? And by asking better questions, you will likely find all the answers you need to ascend atop your March Madness pool.
The NCAA tournament bracket uploaded on the evening of Selection Sunday is a blank canvas. Should you roll up your sleeves and start painting? Should you grab a blindfold and wing it? Or should you do your research, crunch all the numbers and apply careful consideration? Of course not. Who has time to do that?
Instead, I have done the analysis for you. I have asked the nine questions you need answered to fill out your brackets. All of a sudden, the blank canvas turns into paint by numbers. And you now have the color key.
Check out our 2015 NCAA Men's Basketball Tournament Cheat Sheet!
1. HOW DOES EACH SEED PERFORM IN THE ROUND OF 64?
The chart below displays the winning percentage since 1995 of 1-8 seeds in the Round of 64, i.e., over the past 20 tournaments. With the exception of the performance of 5-seeds, the gradually decreasing results over time indicate how well the tournament selection committee does at seeding schools.
These results suggest that you are likely safe in advancing the following:
- All the 1-2 seeds;
- Three or four 3-seeds;
- Three 4-seeds;
- Two or three 6-seeds; and
- Two each of the 5, 7 and 8-seeds
I have also isolated the previous two five-year periods for comparison purposes in case you are speculating that times are changing in the NCAA tournament.
The results from the past five years should have you reconsidering the quantity of low seeds advancing into the Round of 32. For example, the occasional 2-seed falling in the first round has become more prevalent. As well, the 6-seeds are falling victim in the Round of 64 as often as the 5-seeds. The only significant improvement was at the 8-seed – they lost more than they won from 2005-09, but rebounded to a .650 winning percentage from 2010-14.
2. WHICH HIGH SEEDS ARE LIKELY TO REACH THE SWEET 16?
So, you have rolled the dice in the Round of 64 on a few sleepers at high seeds. But which colleges will advance one more round into the Sweet 16? The chart below shows the proportion of high seeds that have reached the Sweet 16 in the past ten NCAA tournaments.
For every four high seeds you advance, approximately three should either be a 10, 11 or 12-seed. But how many high seeds should be pushed forward to begin with? Since 2005, 85 high seeds have won during the Round of 64. That figure suffers a sharp decline in the Round of 32 where only 30.6% of the 85 teams advanced further. Of those high seeds advancing, mid-major programs have had better success (34.0%) in the Round of 32 compared to the majors (26.3%).
To summarize, past tournament history suggest that you should advance approximately 30% of your high seeds from the Round of 32 into the Sweet 16. 75% of those teams advancing should be made up of 10-12 seeds and a slight edge should be given to the mid-majors.
3. HOW DO SEEDS PERFORM AGAINST EACH OTHER FROM THE ROUND OF 32 THROUGH TO THE ELITE 8?
Does a 2 beat a 10? Do 3-seeds typically eliminate 6’s? These are the dilemmas of the Round of 32. To help you out, I have created the following heat map to identify which seeds have the most success playing against other seeds. The heat map is based on data from the Round of 32 through to the Elite 8 over the past ten tournaments. The results are limited to displaying scenarios with a minimum of eight games since the results could be misleading with fewer data points.
The squares with green shading indicate the team has more success against their opponents. The darker the shade signifies the greater the success. The opposite is true for the red and orange squares. A legend is provided categorizing the winning percentages into colors. For example, 3-seeds have beaten 6-seeds in 68% of their pre-Final Four matchups over the past ten tournaments.
The heat map shows that the 1-seeds have a strong likelihood of advancing to the next round unless facing a 2-seed in the Elite 8. The strongest opponents for 1-seeds are the 4-seeds in the Sweet 16. This is similar to 2-seeds who win at least two-thirds of their games with the exception of Sweet 16 matchups against 3-seeds. In evaluating higher seeds, I was surprised at how strong the winning percentages were of 10 and 11-seeds versus 2 and 3-seeds, respectively, in the Round of 32.
4. DO CONFERENCE REGULAR SEASON OR TOURNAMENT CHAMPIONS HAVE SUCCESS DURING MARCH MADNESS?
Does sustained success over an entire regular college basketball season lead to success in the tournament? Or do teams go deeper if they peak heading into March Madness by winning their conference tourney?
Unfortunately, you can’t just look at the aggregate tournament winning percentage of conference regular season or tournament champions. A team’s chances are greatly diminished against a 2-seed from a major conference even though they might be a 15-seed Patriot League champion. What you can do is compare the performance of conference regular season or tournament champs versus that of their peers, i.e., schools with the same seed who failed to win any conference titles. The graph below provides this comparison by showing NCAA tournament results by seed since 1995 and then isolating the results for conference regular season and tournament champions.
The results for conference champions track the general winning percentage of each seed quite well with the greater variances occurring between the 5 and 10-seeds. I have separated the analysis into low and high seeds to better capture the impact.
Taking a closer look at the first 8 seeds, you can observe how conference tournament champions have a distinct advantage at all seeds between 2 and 8, with the exception of a significant drop at the 6-seed. Tournament champs have the biggest impact at the 5, 7 and 8-seeds. Regular season winners track the general results at the low seeds much closer with the exception of the 5-seed. The same analysis can be done with the high seeds, 9 through 16.
Between the 9 and 16-seeds, there is little difference between the results of regular season and conference champions. This is primarily because many of the regular season champions seeded between 9-16 qualify because they also won their mid-major conference tourney. The most noteworthy difference among the higher seeds is for the 10-seed – the success of regular season and tournament winners is about 10% higher. Starting with the 11-seed, you can further observe that the performance of conference champions begin to dip below the general results.
5. HOW MUCH OF AN ADVANTAGE DO SCHOOLS WITH MORE RECENT TOURNAMENT SUCCESS HAVE OVER THEIR OPPONENT?
This question isn’t whether those programs who have tasted success in recent tournaments will have an advantage based on past historical trends. The evidence is in their favor in every round. Rather, the question is how much of advantage they have. And, secondly, when is that advantage at its peak?
The graph below introduces the concept of “senior class” – a team’s current season and the three previous years. For each round since 2005, I have charted and compared the average number of tournament victories for the senior class of the winning and losing teams. For example, the winning team in the Sweet 16 has a senior class that has averaged 6.0 previous tournament victories – 1.5 more than the losing team.
You can see that there is a clear advantage starting the Round of 64 and continuing through to the Sweet 16. Afterwards, an advantage remains, but the additional number of victories by the winning team is merely incremental in comparison.
6. IS THERE A TREND TOWARDS MORE MID-MAJORS REPRESENTED IN THE FINAL FOUR?
It sure feels like times are a-changing. Cinderellas from mid-majors appear to be leaving their mark in the Final Four and not just in the first few rounds. Was it George Mason in 2006 when you first took notice? Or was it Butler in back-to-back Finals that convinced you? Some contend that the trend was real once Virginia Commonwealth emerged in 2011. But is there a trend? The graph below shows the allocation of teams from major and mid-major conferences in the Final Four in five-year increments from the past 30 tournaments.
For greater clarity, I define today’s majors as the following seven conferences: the ACC, American Athletic, the Big 10 and the Big 12, the Big East, the Pac-12 and the SEC. All other conferences are considered mid-majors.
There is a slight trend towards mid-majors improving their allocation, but the Final Four has been represented consistently by schools from major conferences regardless of the time period. Over each five-year period, mid-majors make up 10-20% of those who qualify for the tournament’s final weekend. The dominance of the majors at this point of the tournament is convincing. When making your selections, you will likely want to pencil in three schools from major conferences and then decide whether the fourth should or shouldn’t be.
7. WHICH COACHES HAVE THE MOST (AND THE LEAST) SUCCESS IN THE TOURNAMENT?
Once again this year, 68 coaches will storm the sidelines during March Madness. Some are icons. Some are the new guard. And some are underrated despite perennially leading their school on a trip through the tourney. It’s clear, though, that they all have one thing in common: Each man has one goal – to lead their recruits to a national championship.
The first chart below examines the icons – any coach with a team in this year’s tournament who has coached 30 or more tournament games in his career. The results in the chart below represent each coach’s performance over the past ten seasons.
At the top of the list is Kentucky’s John Caplipari who led the Wildcats to the Finals last year and Roy Williams who always seems to have North Carolina in the conversation. When making those tough calls in your bracket, could a coach’s track record be a deciding factor? If so, make note of career tournament winning percentages of Rick Pitino (Louisville), Bill Self (Kansas) and Tom Izzo (Michigan State). On the other end of the spectrum, Rick Barnes of Texas, Mark Few of Gonzaga and Steve Fisher of San Diego State have had much less success considering the number of trips they’ve made.
The second chart is similar but focuses on coaches in this year’s tournament who have been on the sideline between 10-29 tournament games. In this case, I have included the six highest and the six lowest winning percentages.
Legendary ABA & NBA coach Larry Brown of SMU leads this list but his track record decades ago at Kansas is difficult to translate to today. Arizona’s Sean Miller and Baylor’s Scott Drew have had tremendous tournament success over the past several years and Shaka Smart has surpassed all expectations leading Virginia Commonwealth deep in prior tournaments.
At the top of the list to avoid is Davidson’s longtime coach Bob McKillop who has only been victorious in three of 10 tournament games – all three came during Davidson’s 2008 run to the Regional Final with Seth Curry. Former Winthrop head coach Gregg Marshall has also taken the Wichita State program deep into March Madness but failed to duplicate their 2013 success in last year’s tournament as a 1-seed.
8. HOW DO SEEDS PERFORM AGAINST EACH OTHER IN THE FINAL FOUR AND FINAL?
Excluding the First Four play-in games, it’s not until the Final Four rolls around that the tournament potentially witnesses the same seeds matching up. Therefore, it is worthwhile to re-examine the analysis used to answer Question 3 but instead focus on the entire Final Four weekend. Once again, I used heat maps to compare seed vs. seed performance. In this chart, because of a smaller sample size, I have grouped the seeds into buckets: 1-seeds, 2-4 seeds, and seeds 5 and above.
Over the past decade, 1-seeds have been shoe-ins versus 2-4 seeds – winning in 90% of the matchups. The success extends to the higher seeds as 1-seeds tend to win three quarters of their national semi-final and final showdowns against seeds ranked 5 or higher. 2-4 seeds have similar success versus 5+ seeds winning two-thirds of their games in these latter rounds.
With the exception of last year’s bizarre Final Four performances from 7-seed Connecticut and 8-seed Kentucky, the results above prove that lower seeds have a higher probability of taking care of business once the tournament reaches this stage. In fact, the most striking statistic is that 26 of the past 30 National Champions have either been a 1 or a 2-seed.
9. WHAT SHOULD I PUT DOWN IF MY POOL’S TIEBRAKER IS TOTAL POINTS IN THE FINAL GAME?
Imagine the heartbreak. You hang onto every moment from 67 basketball games hoping that your bracket pays off. Your picks come down to the wire and you realize after the nets are cut down that you are tied with Susan from accounts payable. In that instant, are you comfortable with your tiebreaker?
Typically, March Madness pools will use the total points for both finalists in the championship game as the tiebreak. The graph below shows the distribution of total points in all NCAA tournament games since 1985 (blue) and also isolates the totals over the same period for the Final (orange).
The shape of the curve represents the variability and the dotted line indicates the average. Looking at the average first, you can observe that the Final has historically been higher scoring than tournament games in general. The Final has averaged 145 total points since 1985 compared to 142 for all tournament games.
The shape of the orange curve is taller and slimmer – this indicates less variability compared to the blue curve. Less variability is an indicator that the average is more reliable; therefore, you may not want to venture too far from 145. That being said, the variability of the Final may be lower than the same statistic for all tournament games, but I wouldn’t consider it to have low variability. For example, since 1985, only 26% of Finals have ended with an aggregate score in the 140’s. Further, the average points over the past ten Finals are 135.
LOOKING FOR MORE
You can find a list of the entire 2015 NCAA tournament field in my Cheat Sheet which also summarizes my findings from this analysis.
You may also want to refer to my analysis from last year’s NCAA Men’s Basketball Tournament for which I studied the impact of a school’s seeding, their conference, and the experience of their head basketball coach in the following articles:
- How Does Seeding Affect Success in the NCAA Men’s Basketball Tournament?
- Which Current Conferences Have Had the Most Success in the NCAA Men’s Basketball Tournament?
- How Does the Historical Performance of a Head Coach Affect a Team’s Success in the NCAA Men’s Basketball Tournament?
Bob Sullivan writes periodically for SportingCharts.com and can be followed on Twitter at @mrbobsullivan.