March Madness Predictions & Picks 2013
Ah, March is here, which means it’s time to make your 2013 March Madness picks. We’re been pumping out data-driven March Madness predictions for over ten years now, and in that time we’ve learned a thing or two about what factors most people use to pick their March Madness brackets, and what factors actually lead to better March Madness predictions. Below are 10 things you’ll want to keep in mind when making your predictions for March Madness in 2013.
1. March Madness Picks based on Overall Team Performance
There are many ways to measure the overall performance of a team when preparing to make your 2013 March Madness predictions. Some examples, moving from least to most useful, are win-loss record, average margin of victory, RPI rating, March Madness tournament seed, and predictive power ratings. Yes, that's right; win-loss record is at the bottom of the list when it comes to making March Madness picks. Luck plays a huge role in wins and losses, which is why margin of victory is a better predictor to use when making predictions for the 2013 March Madness tournament. Our power ratings combine margin of victory with information about who a team played, where they played them, and when they played. These are the most important factors when rating a team, and make our power ratings very useful when making March Madness picks.
2. March Madness Predictions based on Recent Play
Some teams you're considering for your 2013 March Madness picks were lousy in November and great in February. Other teams were great in November but lousy in February. If your picks for the 2013 March Madness tournament have a team going far, all else being equal, the former is most likely better than the latter. But don't assign too much importance in your 2013 March Madness predictions to the last two or three games. From a predictive standpoint, playing great in December and very good in February is better than playing poorly in December and great in February. As the Selection Committee says, the full body of work matters. Recent performance isn't the only factor that should affect your March Madness picks.
3. March Madness Picks for 2013 based on Home-Away-Neutral Performance
Some teams in your 2013 March Madness predictions have played very well at home, but poorly on the road. Other March Madness picks seem to perform well at neutral sites. Conventional wisdom says that playing much better at home is a bad sign, and should be avoided in your March Madness predictions, but the data doesn't agree. People like to say there are no home games during the March Madness tournament, and that's true. But there are no true road games, either. For a recent example of a good March Madness pick that played better at home, look at Duke in 2010. They were the best home team in the country, and much worse on the road. Yet they cut down the nets in April, and our March Madness prediction that they would win the tournament was correct.
4. March Madness Predictions for 2013 based on Performance Consistency
Some March Madness tournament teams tend to beat the teams they're supposed to beat, and lose to the teams they're supposed to lose to. Other March Madness teams in 2013 might have beaten some better teams, but lost to some real stinkers. If your 2013 March Madness picks have a high consistency, they're less likely to knock off a much better opponent, but they're also less likely to get upset by a tournament underdog. Our Consistency ratings can help you make your March Madness predictions in 2013 by identify which teams fit the above descriptions. You may want to select double digit seeds ranked low in Consistency to knock off favorites ranked low in Consistency in the first round of your March Madness picks. And your prediction for the 2013 March Madness champion should probably be a team ranked high in Consistency, since they need to play six straight good games.
5. 2013 March Madness Picks based on Overall Odds
A lot people follow March Madness prediction rules like "always pick a 5 seed to beat a 12 seed in the first round." The value of that March Madness pick strategy is debatable, but they don't help matters by picking the 12 seed based on gut instinct. Our data-driven survival odds can help you be smarter about your 2013 March Madness picks by comparing the odds of similar seeds in different regions. Our survival odds can also help you manage risk in your 2013 March Madness bracket prediction, by not picking Final Four teams that are too unlikely, and by focusing your first round March Madness picks on teams that actually have a decent shot at winning.
6. March Madness Picks based on Conference vs. Non-Conference Play
When making your March Madness predictions in 2013, the difference between conference and non-conference play can be safely ignored for most teams. But a few teams rely on unusual defenses (Syracuse's zone) or unique personnel sets (VCU in their Final Four run) that are hard to duplicate in practice. For those potential March Madness picks, non-conference opponents may fare worse against the team, simply because their style of play takes some getting used to. For teams with unique styles, looking for discrepancies between our In-Conference and Non-Conference power ratings may help you when making your 2013 March Madness predictions.
7. 2013 March Madness Predictions based on Close Games
All win-loss records are not created equal, so don't treat them that way when making your picks for March Madness 2013. Some teams dominate their competition and end up with a high average margin of victory, while others squeak by a lot of opponents, without suffering many close losses. Including teams like the latter in your March Madness picks may put you on course for a rude awakening in the March Madness tournament. It's a fact that luck and variation plays a large role in close games. Remember this when making your March Madness prediction in 2013; teams that are on the winning end of a lot of one possession games tend to get over-seeded in the tournament. These are exactly the types of teams that are prime candidates for suffering a first round upset, and getting knocked out a round or two earlier than expected. So you may want to fade them in your March Madness picks.
8. March Madness Predictions based on Good Wins and Bad Losses
Some potential March Madness picks can't seem to break through against good opponents, but handle mediocre opponents with ease. Other teams under consideration in your March Madness predictions rack up good wins, but also suffer bad losses. This is related to the Consistency ratings we discussed earlier. When making your 2013 March Madness predictions, you may want to focus your first round upset March Madness picks on low seeds that have shown they have the ability to pull off a good win, and high seeds that have taken a bad loss or two.
9. 2013 March Madness Picks based on Individual Game Odds
When making your March Madness predictions in 2013, an important consideration when choosing your March Madness picks for the later rounds is a team's overall odds of making each round. But another factor is the odds of winning any individual game. We publish individual game March Madness predictions for every single first round matchup. OK, that's not unusual. But we also publish individual game odds for every single hypothetical March Madness tournament matchup. Using our Matchup Predictor you can see March Madness predictions from six different computer models for any pair of teams, which can be very helpful when filling in the later rounds of your 2013 March Madness predictions.
10. 2013 March Madness Predictions based on Fudge Factors and Intangibles
We think our team stats, ratings, computer models, and March Madness predictions are great, and they are incredibly value for making your 2013 March Madness predictions. But even we have to admit that there are sometimes important factors influencing March Madness picks that our models can't account for. Probably the biggest factor is injuries. If a team loses their star player just before the tournament starts, there's no way to accurately measure what impact that will have on their March Madness prediction. Our models do take Vegas lines into account, which will certainly help, but you also need to throw a dose of subjectivity into the numbers, and make mental adjustments to your March Madness picks for extenuating circumstances.
March Madness Pick Advice to Ignore
There are a few factors that people tend to overrate when making March Madness predictions. Here's a short list, along with why you should ignore these factors when making your March Madness picks.
- RPI Ratings - The RPI rating was not designed to pick March Madness games. It was designed to help decide which teams get into the field, which is why it totally ignores margin of victory. Because of that, it's not very useful for making March Madness predictions. If you're looking for a team rating to rely on when making your 2013 March Madness picks, you're much better off using our Predictive Power Rating.
- Conference Tournament Winners - The media likes to make a big deal about March Madness conference champions being "hot" going into the tournament, but in most cases, that doesn't make those teams good March Madness picks. A three game winning streak does not make a dominant team, so be wary of basing your 2013 March Madness predictions solely on conference tournament play. Connecticut last season is obviously a counterexample of a hot team that was a god March Madness pick, but remember that the Huskies were fantastic in their nonconference schedule, and their Big East tournament run featured five games, not just three. Advancing them far could have been a decent March Madness prediction, because they had a significant body of work in which they played excellent.
- March Madness Seed - #1 seeds are indeed much better teams than #16 seeds. With that said, the March Madness uses an imperfect seeding system to pick the March Madness field. For instance, a #7 seed is not necessarily better than a #10 seed, so you shouldn't automatically pick #7 seeds in your 2013 March Madness prediction. In fact, over the past decade #7 seeds have a winning percentage barely above 0.500 in the first round. Here's a good rule for making 2013 March Madness predictions: scratch out all the seed numbers on your March Madness picks sheet and use our web site as a basis for your March Madness predictions instead!