FanNation - The Republic of Sport

So, you don’t or haven’t followed college basketball, but would like to participate in a bracket contest and last year coin flipping didn’t get you very far? Well, you don’t need to be an expert to do well, as any seasoned “bracketor” will tell you. Dick Vitale can’t even get it 100% right. The more knowledge of the basketball season the better, but you also need to understand statistics. The truth is only one thing determines who wins: the score. Not rank, nor record, nor reputation will give you a win, only more points than your opponent. So it’s not a question of whose better, but more who will play better. Upsets are not uncommon and there are always a few surprises in the bracket. However, there are ways to predict or feel out your way through the bracket, even if you don’t really know what you’re doing yet.

So from a man who enjoys statistics in his free time and without further ado…

Here are some tips for making your picks:

Do the Research
The last thing you want to do (if you want to do well) is go into the tournament blind. Take the time to do some sports article reading, talk to people who follow the game, watch sports highlights, etc. The more you know the better. It’s kind of like you went the whole semester without doing any work and now you have to cram before the final. Are you going to know every detail? No. But it’s important to extract as much useful info as you can.

Here’s are some things to look for:

Who are the best teams? Make sure you know who’s been good all season.

Check out all the top teams’ records. This doesn’t just mean wins and losses. Was it a difficult schedule? Who did they lose to and was it close? What top teams did they play and how well did they play? How dominant were the scores?

March 2008 example: Though Memphis and Kansas are both #1 seeds, Memphis had a harder schedule and better record. Memphis’ only loss came to Tennessee, a very good team, while Kansas lost to Kansas St., Oklahoma St., and Texas. Texas is the only acceptable loss in that list.

Who’s overrated? There are plenty of teams that are overrated each year, it’s important to figure that one out.

March 2008 example: Duke

Who’s underrated? Some teams have harder schedules than others and receive a lower seed because of a worse record.

March 2008 example: USC

Look at the history of the tournament. Some teams are just usually good picks. This is a combination of things: being underrated, experience, and/or luck

March 2008 example: Winthrop; they’re always in it and produce upsets in years past

Negatives and Positives; A team could have been great all year, but if their top player gets hurt right before the tournament, they probably won’t go deep. It’s not just injuries, look at the teams’ morale. Has anything happened recently that might make them not perform as well as they have? Similarly, maybe a team has had some recent pick up in their game and are better than they were most of the season. A player coming back from injury is another example of something that may alter the expected flow of the tournament.

March 2008 example: Indiana. Yes, they were good and perhaps an earlier favorite to go deep, but after the Kelvin Sampson mess it’s questionable how well this team with hold together.

Miscellaneous: It’s important to keep your ear and mind open to other peoples’ opinions, especially if you didn’t follow the season closely. It’s sports reporters’ jobs to analyze this stuff, they’re great references. Talking and listening to people who follow the game is very important in helping you with your decision making

Start with a conservative bracket
When first filling out your bracket, don’t make a bunch of upsets. Start with something easier and not much thought. Think less about the early rounds and more about later rounds. Who do you think will go to the final four, who’s going to the finals, and who’s going to win. These are the money rounds, one wrong pick and you lose a lot of points. Work on getting an idea of the outcome of the tournament, not all the small details. You start really getting into it later. If you get a game wrong in Round 1 it’s only 1 point but if you get a game wrong in the final four it’s 16 points. And remember, you can change your mind later.

It’s also important to take the time to evaluate what teams have harder draws. A team is more likely to get to the final four if their quarter of the bracket is easy.

March 2008 example: UCLA has arguably the easiest bracket and Memphis the hardest

Utilize helpful online brackets
Whether you’re participating in the ESPN bracket, a bracket pool at your work, or just one between a few friends, use an automated bracket online. The more info the bracket gives you the better. The Seth Davis Challenge is a very useful bracket. It gives you the percentage of how many people are making the same pick, quick access to team information (e.g. season record and scores), and it’s very easy to change picks and experiment. The most important thing it does, however, is the percentage. This gives you a good idea what other people are doing.

Reading Trends
This is an example with numbers taken from the Seth Davis Challenge bracket. The percentage given in this particular brackets means it is the percentage of people who are picking that team to win that round. For example, let’s say in the second round you have Clemson (5) playing Vanderbilt (4) and lets also say that 54% of participants have picked Clemson to win that round, 36% picked Vanderbilt, and the remaining 10% picked a different team. The populous favors Clemson over Vanderbilt, even though Vanderbilt is the higher seed. This is a difficult pick, and if you’ve seen Who Wants to be a Millionaire, you know that on (legitimately) harder questions, the audience isn’t always right. Your gut is telling you to go with Vanderbilt. In this scenario, playing with statistics can help out. It’s important to try to analyze what others who picked Vanderbilt did. Let’s say in the first round, 91% of people picked Vanderbilt to win. So if you divide the current round % by the previous round % (36/91) you find that about 40% of people who picked Vanderbilt to win the first round again picked them to win the second. We’ll call this a loyalty percentage. Similarly, let’s say 77% of people picked Clemson to win the first, so then doing the same calculation discussed before, we find that 70% of people who picked Clemson to win the first round also picked them to win the second. It’s also important to analyze people who also have Clemson vs. Vanderbilt in this round and their picks. To figure out how many people in this category picked Vanderbilt, divide Vanderbilt’s second round percent by the sum of both Clemson and Vanderbilt’s second round percent (36/[36+54]). This shows that 40% of people with the same match up, picked Vanderbilt to win, and thus 60% picked Clemson to win.

What do these numbers mean? This is not an obvious game like the 1^st seed playing the 16^th seed, where you’ll find 99.9999% of people picking the 1^st seed. You would normally expect this 4^th seed vs. 5^th seed to be around the 50-50 mark, something like 52-48 biased towards the higer seed. 60-40 in favor of Clemson with 70% loyalty is a little on the high range in this instance.  An upset is normally not the most popular pick and not this dominant. Now you have to ask why Clemson is the fan favorite. Well, the most likely possibility is that Seth Davis said he thought Vanderbilt was overrated. People doing the Seth Davis challenge probably read what he says, so a lot are taking his advice. This is where you would look more into Vanderbilt and Clemson and decide yourself if you feel Vanderbilt is overrated. Either way, this is a safe pick to go against the grain and chose Vanderbilt. If you found out more, like one of Vanderbilt’s better players was injured, then you’d understand why people would be likely to pick them to beat a 13^th seed but not a 5^th. In this case, make your own decisions and opinions, don’t make a pick strictly on the fact it’s what a sports writer said. Watching what the crowd does is just as important as figuring out why they do it.

This is also a good method of watching pick trends. Unless someone is an overwhelming favorite, you’re going to find numbers lingering around 50-50. If you don’t, it’s your job to figure out why. If you find that 52% of people picked team X to win the previous round, but 3% of those same people picked that team to win the current round, it’s a good tell tale of what pick to make, because no one has faith in that team. Continuing after that 52% to 3% drop off, what if your numbers started showing that 85-95% of people who picked team X the previous round, picked them in the current round for each of the following rounds (i.e. a consistent high loyalty percent after a small loyalty percent)? I would guess these are people who straight up picked that team to do well without thinking about it; fans of team X would be responsible for this trend most likely. Most brackets won’t show this though because their percentages don’t go into decimal places, when it gets small enough it’ll show 0%. However, it’s possible for a team with a large following to cause this kind of trend to be visible.

Don’t be loyal, be logical
I went to UMBC, they made it into the tournament for the first time since…well, ever this year. Am I picking them over Georgetown? As much as I want to have UMBC winning the tournament, I don’t honestly believe it’s going to happen. This is the case of logic vs. loyalty. If you want to do well, don’t let what you want to happen get in the way what you predict will happen. Keep an open mind. Your team probably isn’t the greatest team ever and people will understand if you don’t have them as champions. And if you do pick them to lose a game, but they win it, that’s great. You’ll be happier about the win than disappointed about the pick.

Don’t be a gunslinger
Picking lots of crazy upsets is not going to get you anywhere. So you may have one right, but you’ve completely destroyed your chances of doing well. A performance like GMU in 2006 is completely unpredictable. Picking a double digit seed to go to the final four is reserved for fans of the team and X game athletes. Make sure you really think that upset could happen before you pencil it in. Don’t go looking for upsets either. If something stands out to you, investigate. If not, don’t be afraid to be a conservative. Remember: it’s better to get one wrong, than only get one right.

Go with your gut
Instincts are usually a good building block. If you feel strongly about a pick, then pick it. This also means, once you seriously made a pick, don’t change it unless you have good reason. Doubting yourself is not good reason. It’s similar to what they say about a multiple choice test, always go with your first answer unless you have definite proof you’re wrong.

There’s exceptions to every rule
Statistically, statistics can be wrong. It’s called one in a million for a reason, else it’d be zero. It’s your job to interpret the numbers. You make the pick, not the stats. Sometimes even the most overwhelming odds are toppled, and we’ve seen it in big instances twice this year already: Djokovic defeated Federer in the semi-finals in the 2008 Australian Open and the Giants beat the Patriots in Superbowl XLII. All the experts knew Federer would win another slam and all the experts knew the Patriots were going undefeated. Well, they were wrong.

Statistically analyzing each game, you would expect all 4 top seeds to make the Final 4, but how many times does that happen? It’s actually against the odds that something won’t go against the odds. Wrap your mind around that one. Abnormalities are a part of nature and this tournament is very organic. Injuries, hot streaks, teams solidifying, teams falling apart, coaching, crowds, anything you can think of are factors and you can’t predict them all. The truth is something that seemed very unlikely was probably more likely to happen than what people were able foresee. There are a lot of factors that can’t be detected, but are still factors nonetheless.
It is important to keep in mind this natural appearance of anomalies. They’re going to happen, you just don’t know when or why yet. Some are more foreseeable than others. It’s your job to try to detect the ingredients for one of these against-the-odds occurrences.