Tag: Bracket Master

Leap Day Major Update

Bracket Master is being updated today with its first functionality additions of the year. This release has been in the works since November and primarily comprises ‘under-the-hood’ improvements, but there are several new features included.

The app now includes several new well-regarded team rating metrics. Also, our exclusive Scrutinous Score has been refined and given more visibility in the app. We believe the enhanced Scrutinous Score and Scrutinous Game Odds have significantly improved our (already very successful) tournament game picks and simulation.

A Compare Teams function has been added to the top-level menu. This allows you to view side-by-side statistics of any two teams from any season. This feature was previously available through the Picks Page to contrast teams in a scheduled matchup, but this new option gives much more flexibility.

A Settings page has been added to the top-level menu. This allows you to set a Dark Theme or Light Theme display. Additionally, the ability to update the app’s data is provided. Clicking on the ‘Check for Data Update’ button will query the internet for a data file refresh. This is a database file provided by Scrutinous, so an update won’t always be available, but this function should allow us to deliver more frequent data refreshes to you. When an update is available via your app store, you likely will not be able to update using this option.

All of the team and conference logos have been refreshed and hopefully improved.  Some of the images don’t look great while using Dark Mode, so we’ll try to revise them over time.

This Bracket Master version 1.4.1 update is essentially a full re-write of the app. The new code base should improve performance, polish the app’s appearance, better support the delivery of new features, and sustain the app over the long term. Stay tuned for more big enhancements coming this year.

If you encounter a bug, please describe it on our Report a Bug forum. We welcome suggestions for new functionality on our Request a Feature forum.

The Cinderella Score: PTP

Bracket Master’s latest new feature is the Performance Page.  This page allows you to examine the tournament performance of Teams and Conferences compared to expectations.  The exclusive PTP© score presented here evaluates tournament advancement relative to the designated seed number.  PTP© is an initialism of Performance to Par.

PTP defines a seed’s expected performance par by assuming that the higher seed should always win, i.e., game results are determined by ‘chalk’.  If one works through a bracket picking the higher seed in every game, the following results unfold:

  • 1 seeds are meant to reach the Final Four
  • 2 seeds are presumed to reach the Elite Eight
  • 3s & 4s should make the Sweet Sixteen
  • 4s through 8s should win one game into the Round of 32
  • 9s through 16 aren’t expected to win a game.

PTP assigns a score based on these expectations.  For each game a team advances past their seed number’s expected round the team earns one point; each round the team fall short they lose one point.  For example, in 2022 the championship Kansas team (#1 seed) earned two points for winning two games beyond reaching the Final Four.  Saint Peter’s (#15 seed) earned a +3 for winning three games; their first round opponent, Kentucky (#2 seed), gets a -3 for not making their forecasted Elite Eight.

Accumulating a large positive score is challenging for high-major teams that consistently receive top seed bids because of how hard it is to advance deep in the tournament.  To help account for this, we display the Average Seed of each team alongside their PTP.  A lower (better) Avg Seed means the team was anticipated to win multiple games.  A positive PTP score over time with a low Average Seed is particularly impressive.

The challenge for highly seeded teams is contrasted by mid-major Cinderellas such as the Loyola Ramblers.  Their 2018 run as an eleven seed banked themselves +4 PTP.

We feel that the PTP listing of Conferences is particularly useful.  It lessens the bias of small sample sizes to illuminate the leagues that produce title contenders and dark horses, and those that have perhaps become over-rated for a span.

The PTP score helps to remind of the biggest shocks in tournament history, such as DePaul’s collapses as a #1 seed from 1980 to 1984, and Butler’s run to the finals in 2010 and 2011.  The DePaul years are described in the following article: The Bonus: March 14, 1981: When the NCAA tournament became Madness – Sports Illustrated.  Butler is said to be the smallest university to play in the championship game since the 1985 64-team field.  Their 2010 tournament is detailed here: Butler’s incredible 2010 NCAA run, remembered by Coach K, Tom Izzo, Frank Martin and Jim Boeheim | NCAA.com

The Performance Page featuring the PTP score is part of the BracketMaster 1.3.6 release of February 27, 2023.

Introducing the BEST

Bracket Master’s first new feature of 2023 is the Trends Page.  This page allows you to examine the consolidated tournament achievements of Teams and Conferences over a user-specified period of time.  The exclusive BEST© score presented here intends to identify the top entities over an extended period of time by tallying tournament victories, weighted by their advancement in the championship.  BEST is an acronym for BracketMaster Elite Season Trend.  The calculation of BEST is based on the following terms:

BEST© Achievement Points

  • 16 => Won Championship
  • 12 => Lost in Finals
  • 9 => Lost in Final Four
  • 6 => Lost in Elite Eight
  • 4 => Lost in Sweet Sixteen
  • 2 => Lost in Round of 32
  • 1 => Lost in Round of 64/First Four

This scoring system seeks to reward repeated deep runs without overemphasizing championship victories.  Tournaments in years prior to the 64-plus-team field (introduced in 1985) award the same number of points for championships, final fours, etc; those years just  don’t provide points for early, partial rounds.  For example, 1984 champ Georgetown earned 16 points, and first full-round (Round of 32) loser UTEP earned 2 points for making the tournament, but not winning any games.  Similarly, the 1939 tournament with just an eight team field awards 16 points for a championship and 6 points for a first game loss.

BEST also provides an optional weighting factor in order to emphasize more recent performance over that of the distant past.  Seasons of increasing time into the past have their achievement points reduced by a shrinking percentage factor.  BEST uses a 3rd degree polynomial to calculate the weighting factor, with more recent seasons slowly reducing in percentage, seasons after 20 years more rapidly falling, and then leveling out after 50 years in the past.  In terms of specific points, the weighting equation provides a 94% multiplier at 10 years, sliding to 67% at 30 years, 50% at 50 years, and bottoms out at 35% at 80 years.  The most recent five tournaments are always weighted at 100%.  The ability to disable this factor eases comparisons between different spans of time.

BEST can be calculated for the entire history of the college basketball tournament, but it also allows one to examine interesting questions, such as: who was the top team of the 1980s, or: what has been the most successful conference of the 21st century.

Thanks go out to user “Brewer” for suggesting the framework of this rating.  You can suggest your own idea for a new feature on our forum: scrutinous.net/forums/forum/request-a-feature/.

The Trends Page featuring the BEST score is part of the BracketMaster App 1.3.3 release of January 17, 2023.

Bracket Master Tournament Simulation

The 2020 Basketball Tournament was cancelled, but now you can simulate the championship using Bracket Master!  You can simulate a full championship tournament using any season from 1985 to the present.  The Simulation uses an exclusive methodology for rating teams and determining their odds to win a particular game.  Try to get your favorite team to win the championship!

The BracketMaster simulation is based on our exclusive “Scrutinous Score” methodology.  A Scrutinous Score is calculated for each team using advanced metrics from the season’s game results.  Using the Scrutinous Scores of a game’s two teams, we utilize a unique calculation to determine the odds for the matchup.  We simulate each game in the tournament, team-by-team, round-by-round.  Scores of games are produced by the distinct tendencies of the opposing teams.

Better, more highly rated teams will win more simulations than lesser teams, but upsets can, and will, occur.  Our extensive study of the tournament has allowed us to develop a mathematic calculation for the circumstances and frequency of these upsets.

For seasons prior to 2002, Scrutinous Scores are determined using a shorthand calculation that heavily weighs the team’s tournament seed. This demarcation is due to the lack of availability of advanced college basketball metrics prior to 2002.

The Bracket Master simulation is available starting with version 1.1.0.0726 of the app.  The Scrutinous Score game odds calculations can be viewed on both the Simulation page and the  Picks page of the app — Select the “Scrutinous Odds” option from the “Display Mode” drop down menu.

2020 Tournament Cancelled

The 2020 NCAA Men’s and Women’s Basketball Tournaments have been cancelled due to concerns about the spreading COVID-19 health threat.  The NCAA will not release an official 2020 tournament bracket, but Bracket Master has published a projected field that includes: the automatic bids that were decided, the top seeded teams in conferences that did complete their tournament, and a projection of at-large berths.

This result is very disappointing for college basketball fans, but must be incredibly frustrating for several teams that were experiencing historically successful seasons.  In particular we will call out the University of Dayton Flyers and the San Diego State Aztecs along with their consensus All-American players Obi Toppin (Dayton) and Malachi Flynn (SD State).  It’s a shame we won’t get to see these teams compete for a national title.