The AAO Guide to a Winning Bracket

350 or so days a year, our audience is the serious, or at least regular, sports bettor. The kind of person who's looking for an edge, the kind of person who scoffs when some asks them to explain what the -180 and +150 numbers mean on the big fancy boards in Vegas. However, for two events a year the sports betting world opens up to pretty much everyone. One of those events is the Super Bowl. We blew that one, only making one play and failing to get some "fun" prop bet tips up on the blog. Sorry. We're gonna make it up to you on the other event, March Madness.

Today's post is a guide to your office pool. We're not making any picks here, although we will discuss a few teams in the course of the discussion. Instead, we'll give you some rules and guidelines for your office pool. And so without further ado ...

Rule #1, The Golden Rule: Know the Scoring System

Most people know this already, but it can't be emphasized enough. Your picks in a bracket that rewards all winning picks the same way should be completely different from a bracket that gives you double points for upsets, and those should be completely different from those that multiply the points in each round by the team's seed. Obviously, a bracket that doesn't give you any bonus for upsets should have a good deal of chalk, with one general exception that we'll get to in Rule #2.

Let's start with the "straight" pools, the ones that give you no bonus for upsets. According to this incredibly handy guide produced by the Washington Post, since the tournament expanded in 1985, #1 seeds have a 1.000 winning percentage in Round 1, an .875 winning percentage in Round 2, and most surprisingly to me, an .821 winning percentage in the Sweet 16- I was thinking it would be far lower, maybe in the .700 to .750 range. So if you are in a small pool with no upset bonuses, in the past it has paid to go with the #1 seeds to the Elite Eight. Of course, there are some ways to evaluate the best picks that are far better and more accurate than seeds. We'll get to that in Rule #3. And history only tells us so much, as we'll discuss in Rule #4. This is what we call "foreshadowing." But generally speaking, taking favorites for the most part is a solid strategy in these types of pools.

What about those that give you double value for certain upsets? There your best bet, at least for Round 1, is to go to the Vegas money lines. If the potential upset pick is +200 or better on the money line, they're generally a value play for double points. If you're in a larger pool (again, we'll get to that later), you might want to go as high as +250 or more if there's something else you like about the team. Beyond Round 1, your best bet is probably to use the odds for advancement found at basketballprospectus.com. They've got their essential previews up for the East and Midwest today, look for the South and West tomorrow. These previews,while not perfect, are the best tool you'll find to gain an edge. We'll discuss this more in Rule #3.

And for those pools that multiply the total points by the seed? Upsets are the way to go here, but not overwhelmingly so. Generally speaking, if you look at the data from this useful Washington Post link, the seed multiplication matched up pretty closely with the likelihod of advancement. For example, #1 seeds are twice as likely to advance to the Final Four as #2 seeds (42 vs 21 since 1985), a little better than three times as likely to advance as a #3 (42 vs. 12), and a little better than four times as likely as a #4 (42 vs 9). Beyond that, the data at the Final Four level becomes insignificant. Thankfully, Basketball Prospectus once again provides a guide to figure out whether there's value in the higher seeds. By way of example, let's look at the South and Midwest, the two previews available already. The BP numbers suggest that in a multiplier pool, you would be out of your mind to pick the #1 seed to advance to the Final Four. UCLA and Duke provide far better value, when you multiply their odds of advancing to the final four by their higher seeds. The other lower seeds provide value that is generally slightly less than Pitt's value here. Turning to the Midwest, you can see that only Pomeroy favorite West Virginia provides better value than the top seeded Cardinals. So if you like our reasoning with these "multiplier" pools, but can't shake the need to pick some chalk, this is the place to take a #1. I can tell you before they come out that UNC and UConn will not provide similar value in this sort of pool.


Rule #2- Know the Size of Your Pool

Our discussion here will be much briefer. The rule is simple- the smaller the pool, the more favorites you can take. In a fifteen person pool, you can win by simply hitting on two or three of the final four like everyone else, being one of the three or four people to pick the champion correctly, and using basic tools like the Pomeroy numbers and the Vegas money lines to find a couple sneaky "upsets" that people who think every 8 is better than the 9 and every 7 is better than a 10 will miss. In a larger pool, you are far more likely to win if, say, you are the only person who correctly has UCLA coming out of the East than you are if you are one of the 75% of people who have Pitt coming out of the East. In a huge pool, picking Pitt gets you virtually nowhere relative to the field; picking UCLA gets you everywhere ... if you think UCLA have more than a snowball's chance, they are one of many 3 of higher seeds that is worth a shot in those 100+ person pools. For more guidance, the folks at BP should give you some help. In these large pools, it probably serves you well to consider the conventional wisdom, as espoused by the talking heads on ESPN and the "experts" at your local paper. If what these people say about a particular team is at odds with what Pomeroy, Sagarin and others have to say, that might be a good way to pick up some points relative to what the masses will do. Which brings us to Rule #3 ...

Rule #3: The Statisticians and Bookies Are All Smarter Than the NCAA Committee, Jay Bilas, Digger Phelps and Andy Katz Put Together

This one's also really simple. When you fill out your bracket, keep a close eye on teams that are "underdogs" in the brackets but are the better team according to Pomeroy, Sagarin, and the books. Those are the best place to figure out who is more likely to win. If one of them disagrees with the seeds, it's a great place to get the odds in your favor to pick up points on Joe Chalklover, my made-up pool participant who always loves the lower numbered seeds except for his one "upset pick" that he saw mentioned in the local paper yesterday. In my mind, Joe wears short sleeve dress shirts with ties, keeps his ID badge prominently displayed, and has dandruff.

Some experts should be drowned out even more than others, or even faded. Digger Phelps is a great example. His pick of a #1 seed and a #4 seed to meet in the Elite 8 this year is becoming legend. UNC fans all fondly remember 2005, when Digger loudly and repeatedly informed anyone who would listen that UNC would never win the title because UConn had blocked 16 Tarheel shots in a February game, failing every time to mention that UNC had nevertheless won the game, played in Hartford. The lesson here: Digger should be ignored at all costs. You might get dumber by osmosis if you listen to him.

Rule #4: History Means Almost Nothing

You have no doubt heard, or will hear, most of the following things in the runup to the tournament, and probably some other ones about tournament history as well:

- A 12 always beats a 5!
- Duke used to be great, but they've been upset in the tournament the last four years running.
- Gonzaga always gets a lot of hype and then disappoints in the tournament.
- The elite programs always "step up" at tournament time
- Experience is key- seniors win championships.

And so on.

It is important to always remember that when the teams take the floor, it's just a bunch of guys and their coaching staff trying to score more points than the other bunch of guys and their coaching staff. So what about our historical facts above? All true, and all irrelevant:

- A 12 always beats a 5 because there are four 5-12 games every year, and the 5 seeds aren't that much better than the 12 seeds. This year is a great example. 12th seeded Wisconsin is actually better than 5th seeded Florida State according to Pomeroy (although as we've discussed here on other occasions, the Pomeroy system always overrates Wisconsin substantially). The Utah-Arizona game is a pick 'em according to the books. And so on. Again, don't consider the seeds, consider the teams. No need to force a 12 vs 5 upset because "one always happens" if you don't like any of them after reading the other Rules here- although this year, there are several to like.

- Duke has been pathetic in the tournament lately. Embarrassing. Awful. Frankly, they are a disgrace to the university and the alumni they represent, and the University should probably just admit the basketball team's uselessness and move on to things the school actually excels at, like golf and fashion and musicals and whathaveyou. However, we don't see what the miserable peformances of chokers like JJ Redick and DeMarcus Nelson have to do with how Kyle Singler and Eliot Williams will do this year.

(N.B.: the tone of the above paragraph may have been influenced by the writer's three years in Chapel Hill)

- Same principle applies to Gonzaga. Adam Morrison's failings in the past do not affect how this year's team will do. This year's team is good, and they deserve your attention. If you, like Digger in 2005, are not a UNC believer this year, Gonzaga is worth considering as an upset pick in the Sweet Sixteen. Consider the other Rules here before deciding if you think it's worthwhile to take a chance on the Zags in 2009. Just don't let their past performances influence your thinking.

- In 2006, George Mason beat UConn in the Elite Eight, surviving an overtime that 99% of the viewing populations assumed would just crush the mid-major Patriots. In 2007 an Ohio State team reached its first Final Four in eight years (more than that if you abide by the sanctions that removed the visit from the records). In 2008, Memphis reached only it's second Final Four ever (I know, I was surprised too!). Sure, the elite teams make the final four all the time, but it's because of the talent underneath the jerseys, not the names on the front of the jerseys.

- Winning teams almost always have seniors in the rotation because 95% (wild guess) of Division I teams have seniors in the rotation. You know why? Because on average, 22 year olds are better at basketball than 19 year olds. That doesn't mean young teams with great talent should be written off. Ask Carmelo Anthony how important experience is to winning a championship.


Do you have your own thoughts, hints, systems, etc. for office bracket pools? Feel free to share them with us in the comments.