Looking back one year ago, here's some more valuable March Madness Bracket strategies for the fan willing to go the extra mile and do some additional hoops research.

Originally published March 12, 2007. 

Every fan likes to think of himself as an expert on NCAA basketball, and bracket contests give fans a battlefield to test this knowledge.

Unfortunately, while the most brilliant followers of college hoops might know the most trivial details about entire conferences, they often do not see 'the big picture' involved in bracket picking, and have virtually no chance at all to win.

March Madness brackets, like everything else in life, rewards those who do some research before pulling the trigger. If you are willing to spend a little extra time before filling out your bracket, you can greatly improve your chances of winning a bracket. The extra hour you spend will be well worth it as you watch the Final 4.

In a normal bracket contest, you pick the winner of every game. Winners in the round of 64 are worth one point each, the Round of 32 are worth two each, and continue doubling until the championship game that is worth 32 points. You make all your picks before the first game is played; so if you believed Florida International would win it all and they lose in the first round, you cannot win points for any games where you picked Florida International, even though they didn’t play in the game.

When you fill out your bracket, you have one main goal: maximize how many points you will win. You can best accomplish this by accurately evaluating each team, and understanding the effect that having several tough games has on a team advancing. The easiest way to use this approach is to begin with matchups in the round of 64, and go forward one round at a time.

Before picking any games, research a little about any team that you might pick to win in the third round or later. There is no point spending a lot of time studying a team ranked 16th in its regional, or even 12th unless you believe something about a team warrants additional study.

You should typically limit yourself to the top 10 teams in each regional. Start with a typical power rating such as the Sagarin ratings which are easy to get on the internet or newspaper. Next, look at the teams’ results from games this year. Were their results in the last 10 games of the season either significantly better or significantly worse than during the early portion of the season? A problem with power ratings is that those ratings assume that a team plays at one level for the whole season. Different coaches emphasize different parts of the season. Some will try hard to win every game, while others are willing to risk losing games early in the season to teach the team, or find superior rotations.

If a team is peaking in its last 10 games, power ratings will not reflect this and will underestimate that team. Note that the public also tends to underestimate such teams, since it focuses on a team’s win-loss record. If you were evaluating a team that struggled early at 12-8, and then won the remainder of its games, this would warrant a large positive adjustment (perhaps 8 points). Lesser improvements warrant lesser adjustments.

Another area to consider is the starting lineups and substitution rotations used by each team. Did the last 5-10 games use the same starters and subs? If they did, that team will play at a similar level to those games. If the rotation changed, it is helpful to know why. Was there an injury or player suspension? Did the new player(s) perform at the same level as his predecessor? A late-season rotation change is usually a negative for a team, especially if it involved the loss of a point-guard with no experienced backup. In those circumstances, you should downgrade a team.

When selecting teams for the first round, you have an additional piece of information: the betting spreads. Millions of dollars are bet in Las Vegas and elsewhere, making these betting lines reasonable accurate predictors of what will happen in the actual game. It is fairly common to see a lower seed picked to win a game, especially in the #8- #9 and #7-#10 matchups. You can look at these odds in your local newspaper, an on-line newspaper like USAToday.com, or on SportsbookReview.com’s live lines. If you know the spread, you can estimate the chances of either team winning. If a team is favored by three points, it has about a 60% chance to win. At 7, it is about 75% likely.

In the first round you should simply stick to the team that you think will win each match. It would be difficult to suggest that a #12 seed or worse is expected to win. While huge upsets do happen, your bracket will suffer if you keep selecting large underdogs to advance. If you want to win your bracket, never select a #12 team or lower to win a game.

While strong teams are likely to advance many rounds, each round poses additional risk that they will get knocked out. Some teams will have much easier paths than others due to seeding in a bracket. In some situations, it might be easier for a  #4 or #5 seed to advance than a #2 or #3 seed. If you can identify seeds that have an easier route to the top, you can further enhance your chances of winning. You can identify easier bracket routes using a 'strike system.'

After you select your winners for each round, you want to gauge the risk that each team you picked has been eliminated in any round, including earlier rounds you predicted the team to win. Look at the first-round matchups and give a team three strikes if it played in a game that it has a 50/50 chance to win. If a team is favored by 2-4 points in a matchup, give it two strikes. Give them one strike if they are favored by 5-9 points. At a spread of 10 or more, disregard the chance of a huge upset. That team could be upset, but try to focus on events that are likely to occur.

Once you have strikes for each team advancing from the first round, start your second-round picks. Using power-rankings and any information you know about the individual teams, try to set a spread for that game. If one team has more strikes (and has a greater chance of already being eliminated), adjust the spread in that game by three points for each extra strike.

With each round played, most of the surviving teams will continue to accumulate strikes. As you get to the later rounds, there will be big disparities in the cumulative chances of getting upset. The payoff for considering strikes can really be felt around the round of the Elite 8 and Final 4. Most of these matches could be won by either team, so you are often better off just picking the team that had the easiest route there. If you notice that one of the regionals is much tougher than the other three, you will often decide this regional won’t advance out of the Final 4, even if it has the best team.

When you get to the Final 4 and Championship game, keep using this methodology. In addition, you might want to see whether these teams play earlier in the season Are the teams playing similarly now to how they were then? Don’t spend too much time on this analysis though, as there is a good moderate chance these teams will not be playing to advance, in which case your extra efforts were in vain.

The most important part of the bracket contest occurs after you submit it: enjoy the games! March Madness is one of the most exciting stretches of the whole year. Whether you are up against 10 people at your office or 1000 people in an online competition, your Bracket Sheet makes every game interesting. With discipline and a little luck, your card will be a contender to the end.

©2007 Elihu Feustel
Check out Elihu's updated bracket discussion at THIS LINK.