Expert Bracket Definitions & Terminology
We assign each bracket a unique name based on its type, pool size, and other characteristics if they exist. For example, a bracket called RunnerUp-7-250iu indicates the following:
- RunnerUp: The bracket type. (See "Bracket Type" below.)
- 7: The bracket number. This number doesn't mean anything special This is just the seventh RunnerUp bracket that we have published for this pool size and scoring system, and we need to tell them apart somehow.
- 250: This bracket is optimized for a 250-person pool (See "Pool Size" below.)
- i: This bracket is optimized for incremental 1-2-3-4-5-6 scoring by round (See "Scoring" below).
- u: This bracket is optimized or pools which use a seed upset bonus (See "Upset" below.)
A RunnerUp bracket for a standard 1-2-4-8-16-32 points per round scoring system with no upset bonus, on the other hand, will simply be named like this: RunnerUp-7-250
We classify expert brackets into six unique bracket types, defined as follows:
Our top-performing bracket in simulations of a given pool size and scoring system. If we only had one bracket to play in a given pool, this would be it. We publish one Best Bracket per unique combination of scoring system and pool size.
These brackets didn't perform quite as well as our Best Bracket in simulations, but are still worthy alternative choices that feature at least a few different key picks. You can use RunnerUp brackets for a few reasons -- if you simply don't like the picks in our Best Bracket, or if you want to employ a portfolio strategy and play multiple bracket sheets in one pool. In the latter case, you can play the Best Bracket plus one or more RunnerUp brackets.
Brackets that are optimized for pools that employ a "difference in seed" point bonus for picking seed-based upsets. There are a number of variants of this bonus used across the nation, so here's exactly how we define it. If, for example, you pick a 10 seed to win their Round of 32 game, and they end up playing and beating a 2 seed in that game, you would get an upset bonus of 8 points in addition to the standard points awarded for picking a Round of 32 winner. This upset bonus applies to all rounds. It doesn't matter if you picked the 10 seed's opponent correctly or not in your bracket; as long as that 10 seed you picked to advance beats a better seed, you get the bonus points. There is no round multiplier on the bonus points either. For all rounds, you get one point times the seed difference as the bonus amount for picking seed upsets.
These brackets first fix a specific team as the NCAA champion pick, then optimize the rest of the bracket picks given that constraint. As you can see from some of the eROI numbers, it's not a great idea to pick Norfolk State as your NCAA champion, no matter how much you optimize the rest of the picks.
Brackets that assume that our assessment of the strength of a specific conference is too low. Comparing conferences at the end of a season is one of the hardest things for ratings systems to do, since most inter-conference games happen at the very beginning of the season. If you're convinced our Best Brackets are underrating the Big Ten, you can use a BoostBigTen bracket instead.
The opposite of Boost Conference, these brackets assume that our assessment of the strength of a specific conference is too low. If you're convinced our Best Brackets are overrating the Big East, you can use a FadeBigEast bracket instead.
Brackets optimized for competing in a bracket pool where you expect significant bias toward a specific team. In such cases, there is often value in de-emphasizing that team in your bracket. In other words, if you are in a bracket pool full of Duke alums, use the "Duke Bias" bracket to exploit that opportunity.
Which team the bracket has as its NCAA champion, along with their tournament. Hopefully that one was pretty obvious.
The number of people in the hypothetical bracket pool for which the bracket was optimized. Pool size is an incredibly important factor in determining the best bracket picks to make. If you're competing against 1,000 people, the picks you need to make to maximize your odds to come in first place will look much different than if you are in a 10-person pool. If your own pool size is in between two of our published pool sizes, you can first check the picks for the next-largest and next-smallest pool sizes listed and make a few judgment calls, or play both of those brackets!
The scoring system of your pool, defined by points awarded for picking a winner in each round and whether or not a "difference in seed" upset bonus is offered. There are a bazillion bracket pool scoring systems in use out there, but we focus on the two that are by far the most popular: 1-2-4-8-16-32 (one point for each first round game you get right, 32 points for picking the NCAA champion right), and 1-2-3-4-5-6 (in which early round games mean a lot more, relatively). If a bracket is designed for an upset bonus structure too, we append a "UB" to the end of the scoring system value. For example, a bracket with a Scoring value of 1-2-4-8-16-32 UB means that it's for the most common bracket pool scoring system, but with an upset bonus in play.
This stands for expected return on investment, or the percentage return you should expect to earn on your bracket pool entry fee in the long run. This figure assumes your contest has an equal entry fee for everyone competing, as well as a winner-takes-all payout structure where the first place finisher gets 100% of the pot. eROI generally gets higher as pool size increases, however, your odds of actually winning a very large pool in absolute terms are obviously very low. One important note here is that for a variety of reasons, eROI figures for our more specialized brackets (Upset brackets, Boost/Fade Conference brackets, and Team Bias brackets) come with a substantial margin of error on them. For those types of brackets, you can use eROI to tell directionally which brackets look better than others, but don't put a ton of faith in the actual figure.
No surprises here: This percentage denotes the expected chance that you will win a pool of the specified pool size and scoring system if you play the bracket in question. The same caveats as explained in eROI above apply regarding our more specialized brackets, though. For example, brackets such as Boost Conference and Fade Conference effectively assume that you know something about how a teams from a given conference will perform in this year's tournament that most of the rest of the world doesn't. If you're correct, your win odds will be super high, but that's also a risky assumption.