College Basketball

You can tell, because I became enamored at the prospect at something happening that’s never happened before in the national championship game, something that would doubtless send ratings through the roof. I’m running myself ragged trying to finish up classes. Honestly there aren’t that many teams I’m enamored of in this year’s tournament, and many of the ones I am enamored of I have going down. This bracket basically predicts a repeat of when Duke won the national championship a couple years ago basically by being the last team left standing when the carnage cleared. I’ve actually done something I’ve never done before: submit multiple brackets.

I don’t know if you’ve heard, but truTV will be airing a special “Hardcore Brackets” show after the selection show on Selection Sunday. And on this show, for the first time ever, we will learn the actual order that the NCAA ranks all 68 teams in the tournament.

I know a lot of you like to measure just how accurate you are each year, so I would hope that you recognize the new opportunity this presents you. As such, I call for as many of you as possible to release your own S-Curve rankings when publishing your final bracket if you do not already do so.

Refer to this post if you don’t know what this is about or to catch up on the rules.

This year, four conferences produced multiple bids to the NCAA Tournament: the MWC, A-10, CAA, and C-USA. These conferences are guaranteed one spot each in the Mid-Major Conference.

Five teams reached the Sweet 16, and for the first time since I started doing the MMC, two of them came from the same conference, the Mountain West (both lost in the Sweet 16). Of the other three, Butler did not come from a multi-bid conference, while VCU and Richmond did. Neither team from Conference USA won their first game, but Memphis did not have to play in the “First Four”, won the conference tournament, and swept UAB in the regular season. According to the link at the top of this post, BYU’s 2-1 record against San Diego State trumps SDSU’s win over the Cougars in the finals of the conference tournament.

This leaves three spots in the MMC to be determined by my discretion, with no conference restrictions.

Without further ado, the eight members of the 2010 Mid-Major Conference:

Butler (Horizon League)
VCU (Colonial Athletic Association)
Richmond (Atlantic 10)
BYU (Mountain West Conference)
Memphis (Conference USA)
Gonzaga (West Coast Conference)
Princeton (Ivy League)
Wichita State (Missouri Valley Conference)

A lack of mid-major success in the NCAAs (very few multi-bid conferences, very few single-bid conference teams winning tourney games – basically Gonzaga and Morehead State, which falls under the Northwestern State rule) means I not only picked a team in the NIT final four, I almost picked another NIT team in College of Charleston, ahead of Princeton. Then I remembered how good Princeton and Harvard were. Wichita State was maybe a fringe contender at best for an at-large, but Indiana State and Missouri State didn’t make good cases for themselves with the way they crapped out of their respective tournaments.

My experience with Bracket Ladder got me thinking about criticisms that could be made against my rules. VCU simultaneously is an argument against my Sweet 16 auto bid rule – so you’re mediocre(ly good) all season and catch fire at the right time? – and an example of why I have it: no one remembers that VCU only barely got into the tournament now that they’re in the Final Four! A more problematic case is giving Memphis an auto bid solely because UAB got a bid they might not have been deserving of, but the multi-bid-conference rule is more at the core of the MMC; it’s intended to reflect the best conferences. Had they not received an auto bid to the MMC, Memphis might have received a discretionary pick anyway.

Well, that was fun, but I’m never doing it again.

Over the last two months or so of the college basketball season, I engaged in a project I called Bracket Ladder – attempting to show how meaningful the college basketball regular season really is through my own attempt at “bracketology”. I knew it was probably a bad idea to try to balance such a project with my schoolwork, but I didn’t realize just how much of my time it would monopolize. It regularly took me all day to create a new ladder, by which point it would already be out of date. (Is there a reason CBSSports.com’s RPI page, the only freely available page of its kind I know of, doesn’t update until late in the morning the following day, as opposed to, say, 1 AM PT at the latest?) By the end it was taking me two days – and I’d barely even crossed over past the tip of the bubble – largely because the tedium of doing the same repetitive comparing work for two days was starting to wear on me. The result: Despite intending to go daily during Championship Week, I pretty much decided to up and quit after putting out a Ladder Tuesday night.

All that, and I didn’t even show what I had intended to show. My original plans for Bracket Ladder involved not just the NCAA Tournament, but coverage of every team contending for the NIT, CBI, and CIT, to show that all of them are good teams in their own way, comprising still less than half of Division I, a smaller percentage than go to the NBA or NHL playoffs despite playing fewer games per team. By showing how “good” can be a relative term, I would show how even bubble teams are really among the elite squads in the country, not to show that expanding the NCAA Tournament further wouldn’t be a disaster, but to show the opposite: that the regular season is plenty meaningful and to counteract the “regular season is meaningless already” mentality behind the recent push for a 96-team NCAA Tournament. (I’m worried that the ultimate motivation for turning the first and second rounds into the “second” and “third” rounds may be to set the stage for an eventual 96-team expansion.)

There’s a part of me that regrets not getting further than the tip of the bubble (not only for not showing what I wanted to show, but for not finding out if there’s a pecking order between the CBI and CIT), and a part of me that wants to do it again next year just to make good on that, but then I realize I can’t even imagine the amount of work that would have been required by tripling the number of teams I would have had to compare (assuming all the auto bids are within or close to the top 140 teams). But even to the limited extent I was able to do what I intended, it doesn’t look good for that premise, as I found plenty negative to say just about the teams in the NCAAs. (Then again, the fact that I was able to find bad things to say about 1-seeds, and good things to say about teams on the wrong side of the bubble, probably suggests that as a whole, a longer ladder would have largely succeeded in showing “good” to be a relative term.)

Whether or not I would have shown what I wanted to show, though, I still think the concept of the Bracket Ladder is still incredibly useful. College basketball’s biggest problem is the lack of a true national “standings”. The polls extend to the top 25 only, have no bearing on NCAA Tournament seeding and don’t always reflect potential tournament seeding. Most “bracketologists” release their findings as a bracket, which is meaningless until the real bracket comes out on Selection Sunday, and the seeds can’t be used to tease out a rough order of teams because they reflect bracketing principles, including moving teams up or down a seed line as necessary. The only alternatives tend to focus on the bubble, or whether or not a team is getting in or out of the NCAAs at all, not seeding within it, and tends to be treated as radically separate from the bracket despite being two sides of the same coin – and they don’t always do a good job with relative standing, often showing three gradations of teams at most. Extending past the bubble into the NIT field, let alone the CBI or CIT fields, is extremely uncommon and subject to more severe versions of the same problems.

Having some sort of reference of this kind would help me figure out what’s at stake for every team in every game (assuming they’re in contention for a postseason tournament). Personally, I think the NCAA Tournament selection committee should embrace more transparency, which they’re slowly being dragged kicking and screaming to. Slowly, they’ve adopted releasing the order of the #1 seeds, then the RPI throughout the season, and now with the “First Four” the last four teams to make the field. But the controversy surrounding the inclusion of VCU and UAB and the exclusion of Colorado, and the tournament committee chair’s inability to explain those moves, suggests they have a long way to go. The argument that the committee doesn’t want to offend fans of included or excluded schools is starting to no longer hold water. If the committee released their full ranking of not just the at-large teams in the field, but some number of teams that were under consideration at the end but wound up on the wrong side of the cutline, it might go far to help teams figure out what they need to do to improve their chances of getting in, and it might help improve the Selection Committee’s work as well.

(It probably says a lot that my own ladder wouldn’t necessarily have disagreed with the selection committee; the last ladder had UAB – and several other C-USA teams – in the field (and two teams on the bubble, Marshall and UCF, that didn’t even make the NIT) and Colorado out. Once the good wins and bad losses of teams I was comparing no longer involved teams I had placed on the ladder, I was left to work with RPI, and my habit was to favor whoever had both the best wins and least bad losses, and if that wasn’t the same team – generally regardless of how good or bad those wins and losses were – I looked at the “index numbers” – strength of schedule, road/neutral record, out-of-conference record, and record against RPI Top 50 teams – completely equally, and even threw out numbers where one team’s wins and losses were both greater than the other. Incidentially, a funny thing I found out during this process: If the Selection Committee were really conference-blind as they claim, it would actually help teams in conferences with a lot of bids, since they play each other so much. By the end, I had 10 of the Big East’s 11 bids on the top five seed lines.)

Because of this deficiency in college basketball, I still believe in the concept of the Bracket Ladder, though I now suspect it would take a team of people to carry it out to the extent I intended (presumably, a team more versed in college basketball than I am). I still consider the colored bar on the right side of the team name to be the most important part of the ladder. Until Championship Week, it’s largely meaningless and several colors are missing because of the uncertainty and density of games during that span, breaking a lot of the symbolism – “Green”, the color for teams whose seed ceiling is 5 or less, didn’t appear until Old Dominion locked up the CAA’s auto bid – and I would consider simply having a single “green” color for NCAA tournament locks until then, but I still think that the “blue” and higher colors are important to show there are still things worth playing for even within the field, and that the seed ranges and colors would still be an important resource during Championship Week, so you know what even the teams already in the tournament are still playing for besides pride. Certainly it would be useful for me.

(I would use progressively darker shades of red to progress from “NIT Lock” – the same shade of red as “Probably out” of the NCAAs – to black for “NIT Probably Out”, the same color as “CBI/CIT Lock” or “CBI/CIT Probably In”, with a gray color for the CBI/CIT bubble and white for “CBI/CIT Probably Out”. All NCAA, NIT, and in the case of the Great West Conference, CIT auto bids would be integrated into the ladder with a different shade of gray for their color in the place they would be ranked in comparison to the other teams on the ladder.)

One last thing: I intended to eventually introduce another concept as part of the Bracket Ladder, “Recent Win Percentage”, an attempt to accomodate and exploit the committee’s decision to no longer consider any particular number of games down the stretch when evaluating teams. The idea was to average your winning percentage in your last game, your winning percentage in your last two games, your last three games, your last four games, and so on. Though it could be useful on its own, it’s mostly useful by contrast with the regular win percentage, but by the time I got around to calculating it, the Ladder was taking long enough already.

Much of the uncertainty surrounding the seed ranges of teams on the ladder in the past reflected the uncertainty of conference tournaments – namely, who you would face in the conference tournament. All the conference tournament brackets are now set, so we can begin to determine solid seed floors – and seed ceilings, for that matter. All the seed ranges have now been recalibrated to reflect the conference tournament brackets, and we have some new developments as a result, most notably our first “medals”.

(Why do each of the top four seed lines have its own color corresponding to that seed being the floor? Once you get outside the at-larges, the differences between teams go up dramatically, so on the other end of the bracket – the top four seed lines – there’s a lot more competition to get the worst opponent possible. The committee doesn’t make seed adjustments for the top four seed lines and there’s little reason to do so for the bottom four, but that doesn’t mean a one-to-one comparison between the best teams playing the worst teams, so seed line matters a lot more.)

I’ve finally begin to create a bracket and talk about tourney sites. Some caveats: We (or at least I) know next to nothing how the teams will be paired up for the “First Four” games, other than the last four at-larges will be paired with each other and the last four auto bids will be paired with each other, and we also know next to nothing how the NCAA will try to keep the HBCU association happy. With the old play-in game, the NCAA always made sure at least one team did not come from one of the HBCU conferences (the MEAC and SWAC), to leave open the possibility of both HBCU conference champions making the main field. What will they do now with an additional play-in game? Continue to keep an HBCU team out of the First Four? Put the two HBCU teams in different play-in games to keep the possibility of both teams making the main field? Put the HBCU teams in the same play-in game to guarantee one makes the main field? My guess is the first of the three, but we’re starting to push the limits of that strategy working (we’re talking a 15 seed in a 64-team field), so I’m not holding myself to any particular strategy.

Note that the teams out of the tournament are restricted to just the “first four out” for today only. I hope I can do enough on Wednesday to extend the cutline all the way to the first four NIT seeds.

This edition of the Bracket Ladder is complete through the games of March 7, 2011. This means it does not include any of Tuesday’s games, including the Connecticut-DePaul game.

How to read the chart: Teams are listed in order of my assessment of their strength based on the criteria established by the selection committee. The large gray number to the left is the team’s seed in the NCAA Tournament if the teams were seeded strictly according to the list order. Teams may receive a higher or lower seed because of bracketing principles. If a seed has an “f” superscript, that team would play in one of the “First Four” games in Dayton on the Tuesday or Wednesday after Selection Sunday before playing games against teams in the main bracket. The code at the right side of each team name represents the team’s conference and a running count of the number of teams that conference has in all tournaments. The row beneath the team name packs in a whole bunch of information. In order: The team’s record is on the far left in bold. RPI: Rating Percentage Index rank. SOS: Strength of Schedule rank. R/N: Record in road and neutral-site games. OOC: Record in games outside the conference. RPI TXX: Record against teams in the RPI Top 50 or 100. Wv≥: Number of wins against teams listed seven spots behind them or higher on the ladder. Lv≤: Number of losses against teams listed seven spots ahead of them or worse on the laddera. The colored bar at the far right side of the team name is the most important element, containing most of the information you need to know. It is color-coded to reflect where each team is in the pecking order and what they have to play for, as follows:

Read moreBracket Ladder for March 8, 2011

Conference tournaments are starting to get into gear, so better late than never to add the remaining 19 or so auto bids to the ladder. (I don’t actually have write-ups for them because this is late enough as it is.) I used to marvel at the committee’s ability to sort out fairly weak teams no one had heard about, but going in to trying it myself I was actually expecting it to go a bit easier than sorting out the auto bids, since under normal circumstances these teams would be quite spread out on the ladder. Teams listed for auto bids for conferences whose auto bids haven’t been determined are those teams that would have the highest position on the bracket ladder if I extended the ladder that far, not counting conference tournament games. The use of red for the top few teams in this new group is not intended to reflect on where those teams stand in relation to the bubble; it is a very vague guesstimation of how far the bubble extends.

With this addition, we can finally begin to create a bracket (look for a link on Twitter shortly after this goes up) and talk about tourney sites, so a lot of the descriptions for teams on the main ladder have undergone substantial revision even if their relative situation hasn’t much. Some caveats: We (or at least I) know next to nothing how the teams will be paired up for the “First Four” games, other than the last four at-larges will be paired with each other and the last four auto bids will be paired with each other, and we also know next to nothing how the NCAA will try to keep the HBCU association happy. With the old play-in game, the NCAA always made sure at least one team did not come from one of the HBCU conferences (the MEAC and SWAC), to leave open the possibility of both HBCU conference champions making the main field. What will they do now with an additional play-in game? Continue to keep an HBCU team out of the First Four? Put the two HBCU teams in different play-in games to keep the possibility of both teams making the main field? Put the HBCU teams in the same play-in game to guarantee one makes the main field? My guess is the first of the three, but we’re starting to push the limits of that strategy working (we’re talking a 15 seed in a 64-team field), so I’m not holding myself to any particular strategy.

I’ve also made an attempt to break the improbable gridlock at the tip of the bubble, but it’s still not what you expect. I still believe in Marshall and Miami more than most people, I’m still not quite a believer in Washington or Michigan, and I still can’t slide Minnesota or Memphis all the way out of the tournament. But Ole Miss does fall well out of the tournament after losing to lowly Auburn, and apologies to Michael Wilbon, but for all their consistency Northwestern doesn’t have the depth of wins I’d like for the lofty position I had them at before and has way too weak a road record and record against good teams – Minnesota is barely anything at this point. Updates for Saturday’s games coming Monday.

This edition of the Bracket Ladder is complete through the games of March 4, 2011. This means it does not include any of Saturday’s games, including the Duke-North Carolina game.

How to read the chart: Teams are listed in order of my assessment of their strength based on the criteria established by the selection committee. The large gray number to the left is the team’s seed in the NCAA Tournament if the teams were seeded strictly according to the list order. Teams may receive a higher or lower seed because of bracketing principles. If a seed has an “f” superscript, that team would play in one of the “First Four” games in Dayton on the Tuesday or Wednesday after Selection Sunday before playing games against teams in the main bracket. The code at the right side of each team name represents the team’s conference and a running count of the number of teams that conference has in all tournaments. The row beneath the team name packs in a whole bunch of information. In order: The team’s record is on the far left in bold. RPI: Rating Percentage Index rank. SOS: Strength of Schedule rank. R/N: Record in road and neutral-site games. OOC: Record in games outside the conference. RPI TXX: Record against teams in the RPI Top 50 or 100. Wv≥: Number of wins against teams listed seven spots behind them or higher on the ladder. Lv≤: Number of losses against teams listed seven spots ahead of them or worse on the laddera. The colored bar at the far right side of the team name is the most important element, containing most of the information you need to know. It is color-coded to reflect where each team is in the pecking order and what they have to play for, as follows:

Read moreBracket Ladder for March 5, 2011

Take a look at the list of colors below. You’ll see a veritable kaleidoscope of colors mentioned: gold, silver, bronze, purple, blue, green, yellow, orange, and red. Many of these colors were intended to keep to a theme; the Olympic medals were obvious, but the bubble colors were also intended to resemble a “traffic light” through green (go to the tournament), yellow (caution), the in-between orange, and red (stop, you’re not getting in).

Instead, it’s been more like blue, yellow, orange. I just tend to leave my options so wide-open that seed ranges are really broad fairly late into the season. I fully expected that teams wouldn’t lock up medal positions until close to championship week, but we only got our first tourney lock with a seed ceiling that wasn’t 1 a week or so ago! I intended the blue area to represent a “4-seed bubble”, or at least representing something worth fighting for other than just getting in to the tournament, and envisioned most of the tourney locks falling in the green range, but it seems blue is becoming the true color representing “lock”. (Maybe I should have made green “probably in”, yellow “barely in”, and orange “barely out”.) It doesn’t help that on Saturday, teams with seeds as high as 6 were being listed as “probably in”, and bubble colors trump colors that require a seed range because making the tournament is still more important than seed inside the tournament (and what’s the seed floor for a team that could miss the dance entirely?). With a better handle on what the committee is looking for, I could probably tighten the seed ranges and maybe get some teams in the green, and much of the uncertainty in seed ranges has to do with the inherent uncertainty of conference tournaments. Expect seed ranges to close fast over the next few ladders; you’ll see this already in the capsules for Old Dominion, George Mason, and Gonzaga, whose regular seasons are already basically over. I fully intend to put up ladders daily during Championship Week starting Monday, when I’ll do my conference championship seed range recalibration (CCSRR for short).

In the name of keeping the time spent on composition brief (I started working on this after 2 PM PT!), this is not going to be as robust a comparison as what you’ll see on Friday. I intend to take advantage of having relatively less work to do on Friday, so don’t be surprised by seemingly unjustified, “upon further review” moves on Friday.

This edition of the Bracket Ladder is complete through the games of March 1, 2011. This means it does not include any of Wednesday’s games, including the Connecticut-West Virginia game.

How to read the chart: Teams are listed in order of my assessment of their strength based on the criteria established by the selection committee. The large gray number to the left is the team’s seed in the NCAA Tournament if the teams were seeded strictly according to the list order. Teams may receive a higher or lower seed because of bracketing principles. If a seed has an “f” superscript, that team would play in one of the “First Four” games in Dayton on the Tuesday or Wednesday after Selection Sunday before playing games against teams in the main bracket. The code at the right side of each team name represents the team’s conference and a running count of the number of teams that conference has in all tournaments. The row beneath the team name packs in a whole bunch of information. In order: The team’s record is on the far left in bold. RPI: Rating Percentage Index rank. SOS: Strength of Schedule rank. R/N: Record in road and neutral-site games. OOC: Record in games outside the conference. RPI TXX: Record against teams in the RPI Top 50 or 100. Wv≥: Number of wins against teams listed seven spots behind them or higher on the ladder. Lv≤: Number of losses against teams listed seven spots ahead of them or worse on the laddera. The colored bar at the far right side of the team name is the most important element, containing most of the information you need to know. It is color-coded to reflect where each team is in the pecking order and what they have to play for, as follows:

Read moreBracket Ladder for March 2, 2011

So yeah, this is really late because I’m actually starting to get tired of the whole enterprise. But I will press on for you all! Next two ladders will be out Tuesday and Friday.

Three developments on this ladder compared with Monday. First, St. John’s rockets up the board again and turns the Big East’s Big Eight into a Big Nine. There just isn’t enough of a case to be made against them, and frankly, the Big East has an unfair advantage in the seeding since they have so many teams on the top few seed lines. Second, Connecticut’s loss to Marquette finally gives me an excuse to dislodge them from the top seed line. Third, we’re finally starting to see those ugly profiles I mentioned the last two ladders. On Monday, you’ll see the rest of the auto bids, an actual bracket, and the start of talk about tourney sites.

This edition of the Bracket Ladder is complete through the games of February 25, 2011. This means it does not include any of Saturday’s games, including the BYU-San Diego State game.

How to read the chart: Teams are listed in order of my assessment of their strength based on the criteria established by the selection committee. The large gray number to the left is the team’s seed in the NCAA Tournament if the teams were seeded strictly according to the list order. Teams may receive a higher or lower seed because of bracketing principles. If a seed has an “f” superscript, that team would play in one of the “First Four” games in Dayton on the Tuesday or Wednesday after Selection Sunday before playing games against teams in the main bracket. The code at the right side of each team name represents the team’s conference and a running count of the number of teams that conference has in all tournaments. The row beneath the team name packs in a whole bunch of information. In order: The team’s record is on the far left in bold. RPI: Rating Percentage Index rank. SOS: Strength of Schedule rank. R/N: Record in road and neutral-site games. OOC: Record in games outside the conference. RPI TXX: Record against teams in the RPI Top 50 or 100. Wv≥: Number of wins against teams listed seven spots behind them or higher on the ladder. Lv≤: Number of losses against teams listed seven spots ahead of them or worse on the laddera. The colored bar at the far right side of the team name is the most important element, containing most of the information you need to know. It is color-coded to reflect where each team is in the pecking order and what they have to play for, as follows:

Read moreBracket Ladder for February 26, 2011