Double-bye option

[Phil]

One of the Val Ackerman proposals is to switch to a double-bye format (similar to the way the old Big East Tournament worked.) I did some mathematical modeling for Val Ackerman to demonstrate the impact of the switch. I've copied below a revised version of my letter to her. (The table doesn't show up here, see this Google Doc to see the revised letter with the table).

I would love to have some feedback, both on the concept itself, and on how best to make the case.
__________________________________________________________________

Dear Val Ackermann,

As discussed, I promised to do some mathematical modeling to estimate the impact of a double-bye format compared to the traditional format. I have some preliminary conclusions.

The main differences:

1. Under a double-bye format, the chance that the lowest seeds can advance is materially improved
2. Under a double-bye format, the margins of victory are likely to be materially smaller, thus reducing the number of games that are blow-outs, sometimes by halftime

As a reminder, the traditional format means 64 teams are seeded with 16 teams in each of four regions. Six rounds are played, with the first round matchups between 1-16, 2-15...and 8-9. In a double bye tournament seven rounds are needed. Teams are still seeded in four regions, 1 through 16. In the first round, only the bottom 32 teams play. The matchups are 9-16, 10-15,...12-13. In the second round, the winners are matched up against the second 16 teams, that is, seeds 4-8. In the third round, the top 16 teams who have earned a bye in the first two rounds, now are included. At this point, 32 teams remain, and the process follows a traditional five round format with 32 teams.

Impact of switching to a double-bye format:

• Under the traditional format (TF), the chance that a 16 seed will advance is under 1 in 10,000. Given four such matchups in any year, we could expect to wait about 3500 years before seeing a 16 seed advance (at present levels of parity)[1]
• In contrast, under a double-bye format (DBF), we can expect to see a 16 seed advance every five years or so.
• Under TF, we can expect a 15 seed to advance every seven years or so.
• Under DBF, we can expect one almost every year.
• Under TF, we expect the 16 seeds to lose by almost 40 points. While some will be closer, some will have even higher margins.
• Under DBF, the 16 seed is still expected to lose, but by 16 points or so, which will lead to a loss in most cases, but means somegames will be competitive late in the game.
• Under DBF, the 15 seed expected margin drops from 18 to 8, which means many games will have single digit margins late in the game.



I’ll summarize these results in a table:
[See link to see the table]

These are dramatic differences. There are differences that apply to the other seeds as well, but the differences are smaller.

I’ve made a number of key assumptions to produce these results. I think they are reasonable, for the purposes of illustrating how dramatic the differences are, but the assumptions should be refined before using these numbers more formally. In particular, one of the key variables is the standard deviations of margins. I haven’t found a solid source for this value, but some sources suggest a value of about 10. I’ve used this value, but would like to find a better study to support whatever value is best. A related assumption, almost certainly imperfect, is stationarity, that is, the standard deviation can be treated as a constant. It is likely to vary based upon the seeding pairs, however, the broad conclusions are unlikely to be materially changed with refinement of this assumption.

While I do have some statistical experience—I am an actuary and former member of the Board of Directors of the Casualty Actuarial Society—I do want to run my results by someone with more extensive experience. I have already received some preliminary comments from Stuart Klugman, who is not only a statistical expert, he literally wrote the book.[2] He has agreed to help refine the model, if needed.

I hope you will find these results helpful, and if so, I will undertake to improve the model so that the assumptions are more rigorous.

Sincerely,
Stephen W. Philbrick


[1] We all know that a 16 seed has advanced. But we also know that this was a seeding anomaly. Stanford earned a one seed, in no small part due to the contributions of All-Americans Nygaard and Folkl. Both of the players were injured after the regular season ended, but before the NCAA tournament and were not able to play. Had they been uninjured, they likely would have won. While injuries are a part of the game, the loss of two All-American between the end of the regular season and the beginning of the tournament is an extremely unlikely event.
[2] SeeLoss Models: From Data to Decisions, by Stuart A. Klugman , Harry H. Panjer , and Gordon E. Willmot

 

[UcMiami]

Phil - really interesting letter. Small note ... in [1] you have 'in no part part' when I believe you meant to write 'in no small part'.
I can't see your table, but assume it deals with the 9/16 and 10/15 games. I think an additional analysis of the whole of the first two rounds vs. an historical analysis of the last 10 years of the current format vs. the statistical model for the TF would be very illuminating and supportive of the DBF change. Lots more number crunching, but if the idea is to support a DBF change, showing the change in competitiveness in the first two rounds would be very supportive. and doing the statistical model and comparing it to the actual ten year history would help to support your model (or point out where the adjustment(s) need to be made.
I would assume that your model would show that the DBF would result in both a dramatic change in point spread through the two rounds, but also a significantly higher rate of 'upsets' in the 32 games played in those rounds than exist in the 32 first round games played each year in the historic record and I suspect in the whole 63 games of all 6 rounds.
I hadn't looked at the proposed tournament changes up until your post. I sort of dislike the idea of changing the format as it messes with historical record trends, but reading this makes me think it could be very exciting. Assuming the location for the first two rounds would be the 9 - 12 seeds home courts, it could also spread the excitement and 'buy in' from a much broader fan base. Getting a 9 or 10 seed would actually be more highly coveted than getting an 8 or 7 as it would bring home court and presumably two wins to a schools NCAA record.

 

[Phil]

Thanks for your observations. The typo has been fixed.

I can update the numbers for the other options. In addition, I should estimate the total number of upsets. That will be a bit more work, but may be work the effort.

Your observation about the 9 and 10 seeds is worth emphasis. This cuts both ways. On the positive side, it means getting a 9 seed is better than getting an 8 seed, with likely home court advantage, and a good chance at an NCAA win. On the flip side, as was pointed out to me by a colleague who was reviewing the proposal, it means that a human committee, with human foibles, will be making a distinction between the 8 and 9 seeds. While they have always done this, under the TF with pre-determined sites, getting an 8 versus a 9 is a non-event, and I can image that the committee doesn't dwell long on the distinction. Under DBF, the distinction is important, and we might not feel that the committee can do it well enough.

 

[Wbbfan1]

Phil

I wondering if the #1 seeds getting a two game bye is an unfair advantage for them. No Travel, more rest from the end of the season. Although in some cases a #1 seed could end up with having a month off before they play their first tournament game and that may not be an advantage. I don't like the idea of one team having to play 4 games to reach the final four if they continue to win and and a #1 seeded team only having to play two games to reach final four. It doesn't pass the smell test regardless of what Stats may say.

 

[Phil]

Phil

I wondering if the #1 seeds getting a two game bye is an unfair advantage for them. No Travel, more rest from the end of the season. Although in some cases a #1 seed could end up with having a month off before they play their first tournament game and that may not be an advantage. I don't like the idea of one team having to play 4 games to reach the final four if they continue to win and and a #1 seeded team only having to play two games to reach final four. It doesn't pass the smell test regardless of what Stats may say.

This is exactly why I posted this, so I could identify the potential objections, and see if there is an adequate response.

One of the things I have yet to do is to lay out a possible sequence of dates. The DBF requires an extra round. My hope is that I can have the last six rounds match the current six round schedule, and squeeze in the first round earlier.

At present, the women's selection day is Monday, to avoid conflict with Selection Sunday for the men. I'd like to see if it could be moved to the other side, making it Selection Saturday for the women. (This might mean some conferences need to end a day earlier than in the past–I'm not sure whether this is a minor issue or a big deal) If this could be done, then the announcement could occur on Saturday, with the first rounds played on Tuesday and Wednesday, then pick up with the second round on Saturday and Sunday, just as now. (One complication is that the move of the final two rounds from Sunday-Tuesday to Friday-Sunday is causing some thinking about whether other dates have to be moved. At present, I'll assume the answer is no).

While this would mean that the top four seeds would be idle for two games, it only changes the day of the first game from Saturday/Sunday to Monday/Tuesday. In terms of time off, it only means two days, hardly worth worrying about.

The other concern is more meaningful. The top seeds get two more days of rest, and then have to play a team who has played once or twice in the last few days. On the other hand, there is a lot of debate, especially in the pro playoffs, where the differential can be several days, about whether the extra time off is a benefit or a hindrance. It is a subject amenable to statistical analysis, and probably has been done, but I'm not familiar with such a study. Worth reading if it exists, but I'll be surprised if it turns out to be a very material factor.

 

[Phil]

Phil

I don't like the idea of one team having to play 4 games to reach the final four if they continue to win and and a #1 seeded team only having to play two games to reach final four. It doesn't pass the smell test regardless of what Stats may say.

Under a Double-bye format, the number of games need to reach the Final Four for a 1 seed is three, not two.

It s correct, that teams seeded 9-16 would have to win five to reach the Final four. However, only one team seeded 9-16 (Arkansas 1998, a 9 seed) has ever made it to the Final Four, so it isn't like this is going to be a common hurdle.

In 1998 Arkansas won four games to make it to the Final Four, where they lost to Tennessee. If that year had been played under the DBF, they would have had to beat one of the sixteen seeds before going on to win four more games. I don't think beating one of the 16 seeds would have been much of a problem, and would not have tired them out. In addition, while all credit to Arkansas for making it to the Final Four, in terms of analysis, that feat might deserve an asterisk. For the 9 seed to make it to the Final Four, they first have to beat the 8 seed, not a major accomplishment, then the winner of the 1-16 game. In all other situations, that means beating the 1 seed in the region. However, this is the year that Harvard beat Stanford, so Arkansas had the easiest path ever to the next round. Again, especially if there are any Arkansas fans present, I don't want to diminish the accomplishments, but in the context of determining how often a 9-16 seed makes the Final Four, this single instance is a special case.

 

[Alydar]

The double-bye is an artificial maneuver to hide mediocrity and fool people into believing that entertaining BB is being played in the first 2 rounds. The way I see this, the problem is that there aren't 64 worthy teams. The reality is that those proposed first two rounds will only further convince casual fans that wcbb is boring and poorly played. After watching E Oshkosh U beat S Podunk U in a game with 50 fouls and 50 TO's many wouldn't even bother to watch the next weekend. Instead of reducing the number of games the best teams play, they should be maximizing those games. I suggest that the field be reduced to 32, 2 rounds later take the E8 and have a double elimination tourney. There would be more quality to market and many more great matchups.

 

[Phil]

The double-bye is an artificial maneuver to hide mediocrity and fool people into believing that entertaining BB is being played in the first 2 rounds. The way I see this, the problem is that there aren't 64 worthy teams. The reality is that those proposed first two rounds will only further convince casual fans that wcbb is boring and poorly played. After watching E Oshkosh U beat S Podunk U in a game with 50 fouls and 50 TO's many wouldn't even bother to watch the next weekend. Instead of reducing the number of games the best teams play, they should be maximizing those games. I suggest that the field be reduced to 32, 2 rounds later take the E8 and have a double elimination tourney. There would be more quality to market and many more great matchups.

I don't disagree that the NCAA tournament includes some teams that don't belong there, if by belonging there you mean meet some hurdle of competence. Obviously, that depends on how you define the hurdle, but I don't think many would disagree that there is quite a gulf between the level of the play of the 1 seeds and the 16 seeds, a much wider gulf than in the men's game. In fact Val Ackerman suggested investigating whether the field should be cut to 48 or 52.

However, getting from this observation to a workable alternative is tougher than one might think.

For starters, remember that the history of the NCAA Tournament is that it consist of conference champions, supplemented by some at-large teams (I haven't reviewed all NCAA sports, does anyone know if this is universal?) This is how it has worked for the Women's Tournament since the inception in 1982. It has been true of the mens's tournament since their inception. This is in contrast to the NIT, which selected top teams, not necessarily conference winners. The NIT on the men's side was the more prestigious for some time (and predated the NCAA by one year) but over time, the NCAA Tournament became the premier tournament, and the NIT became known as Not In Tournament.

While any rule can be changed, in fact, that's why we are having this discussion, some rules have enormous history built in, and will not be changed without very strong reasons. I suggest that the concept of inviting all conference champions, plus some other teams is such a rule, and is highly unlikely to be changed.

If you believe that, two inconvenient facts rear their heads. First, while most would agree that some of the team invited to the Tournament do not "belong" in some sense of the word, most of the teams not meeting this mythical hurdle are conference champions. So if the sole goal is to limit the invitees to "deserving" teams, one has to repeal the historic rule that all conference champions are invited.

A second inconvenient fact is that there are 31 conferences. While this means that a 48 team Tournament might be plausible, it all but eliminates a 32 team field. If one keeps the rule that conference champions are invited, then a 32 team field means only one at-large team.

This doesn't mean nothing can be done. Perhaps a 48 team field is plausible. With 48 teams, one can have a single bye format. One might also create a hybrid tournament, where conferences are ranked, and perhaps thr weakest 20 conferences are paired up, play a one game playoff, and send ten representatives to the real tournament, or male that the first round, and follow with five traditional rounds.

However,
There is another thread discussing the proposals in general. I started a new thread becasue I wanted a discussion of the merits of a double-bye option. If you don't think we should even invite 64 teams, I have some sympathy, but I'd prefer to discuss that in a separate thread.

 

[Wbbfan1]

I thought of a double elimination concept, but dismissed it. Last year it would have unfairly punished Louisville because if Louisville had to play Baylor again, we know what the outcome would have been. Don't think the NCAA would ever agree with a losing team going to another region to avoid repeat match ups. It would increase travel cost and the NCAA is looking for ways to reduce cost.

 

[Alydar]

For starters, remember that the history of the NCAA Tournament is that it consist of conference champions, supplemented by some at-large teams (I haven't reviewed all NCAA sports, does anyone know if this is universal?) This is how it has worked for the Women's Tournament since the inception in 1982. It has been true of the mens's tournament since their inception. This is in contrast to the NIT, which selected top teams, not necessarily conference winners. The NIT on the men's side was the more prestigious for some time (and predated the NCAA by one year) but over time, the NCAA Tournament became the premier tournament, and the NIT became known as Not In Tournament.

While any rule can be changed, in fact, that's why we are having this discussion, some rules have enormous history built in, and will not be changed without very strong reasons. I suggest that the concept of inviting all conference champions, plus some other teams is such a rule, and is highly unlikely to be changed.

If you believe that, two inconvenient facts rear their heads. First, while most would agree that some of the team invited to the Tournament do not "belong" in some sense of the word, most of the teams not meeting this mythical hurdle are conference champions. So if the sole goal is to limit the invitees to "deserving" teams, one has to repeal the historic rule that all conference champions are invited.

A second inconvenient fact is that there are 31 conferences. While this means that a 48 team Tournament might be plausible, it all but eliminates a 32 team field. If one keeps the rule that conference champions are invited, then a 32 team field means only one at-large team.

This doesn't mean nothing can be done. Perhaps a 48 team field is plausible. With 48 teams, one can have a single bye format. One might also create a hybrid tournament, where conferences are ranked, and perhaps thr weakest 20 conferences are paired up, play a one game playoff, and send ten representatives to the real tournament, or male that the first round, and follow with five traditional rounds.

However,
There is another thread discussing the proposals in general. I started a new thread becasue I wanted a discussion of the merits of a double-bye option. If you don't think we should even invite 64 teams, I have some sympathy, but I'd prefer to discuss that in a separate thread.

You make some valid points. The first 2 rounds of the tournament would be more competitive for sure, but do you actually think any fans would watch the games other than friends and family? This is the one time of year when wcbb is in the public eye and you want the face of the sport for the first weekend to be a game between the 37th & 64th best teams in the country? I am a fan of the women's game but I can hardly watch those games. By the time the top 16 teams get to play most of the casual fans will be turned off. Sorry, but there's nothing entertaining about 52-45 games where both teams combine for more TO's than baskets. This is not "March Madness". it's March mediocrity. Another point is that if they had done that this year UConn would have had about 3 weeks off before their first game. Having no games for the better part of a month is not the way to keep people interested or teams sharp.

 

[Husky1964]

Old timers ( like me) may recall the Mens Tourn. in the years before 1970. UCLA had to win one or 2 games to get to the FF. The Big Dance had 21 teams when we played DOOK in 1964. Of course the women did not get the 64 team Dance until the 80s.

 

[Alydar]

Old timers ( like me) may recall the Mens Tourn. in the years before 1970. UCLA had to win one or 2 games to get to the FF. The Big Dance had 21 teams when we played DOOK in 1964. Of course the women did not get the 64 team Dance until the 80s.

I remember those days. The teams also played in their own regions so UCLA would beat Long Beach State and Montana to reach the FF. Sixteen teams were allowed in so the point in a post above that all conference champs got in is incorrect. The field was expanded due to the massive interest from the public. The women expanded without that support, doing it only because the men did.

Right now nearly 20% of the teams in wcbb make the ncaa field. That's too many considering the available talent. Half that would be better. But I'm not holding my breath.

 

[Phil]

Old timers ( like me) may recall the Mens Tourn. in the years before 1970. UCLA had to win one or 2 games to get to the FF. The Big Dance had 21 teams when we played DOOK in 1964. Of course the women did not get the 64 team Dance until the 80s.

64 teams started in 1994.

See This section for the details of the number of teams by year

But yes, your point is well take. For example in 1968, there were 23 teams in the tournament, and UCLA had to win two to get to the final Four.

 

[UcMiami]

You make some valid points. The first 2 rounds of the tournament would be more competitive for sure, but do you actually think any fans would watch the games other than friends and family? This is the one time of year when wcbb is in the public eye and you want the face of the sport for the first weekend to be a game between the 37th & 64th best teams in the country? I am a fan of the women's game but I can hardly watch those games. By the time the top 16 teams get to play most of the casual fans will be turned off. Sorry, but there's nothing entertaining about 52-45 games where both teams combine for more TO's than baskets. This is not "March Madness". it's March mediocrity. Another point is that if they had done that this year UConn would have had about 3 weeks off before their first game. Having no games for the better part of a month is not the way to keep people interested or teams sharp.

I think you have a chance of developing a following for the schools seeded 9-12 by giving them a home town tournament game to promote with a real chance they can win the game. And while the skill level will not be the same for those 16 teams as for the top 16 seeds, the games will be much more competitive than the typical 1 through 4 seed vs. 16 - 13 seed games so the fan experience may be greater. I do not think you would find that the 'national' audience was very interested in those games and ideally ESPN would release the rights to them to local broadcasters, but that is not that different from a high percentage of the current games.
As for the issue of 5 vs. 3 games to the final four - I don't see that as that big an issue. There was some question about the BE tournament when it changed but I don't think anyone complained once it began (except the coach who only brought two shirts!) And that format was consecutive days - the format we are discussing would include days off between all three games in the first 'weekend' - not that strenuous.
I agree that cutting the tournament down from 64 teams would be very tough to do - you need overwhelming buy in from the schools to make any change and eliminating the automatic bids for all conferences would be unacceptable to a large number. And once you have given berths to the lesser conferences who make up the 13 - 16 seeds, cutting down the at large bids is really tough - the teams going missing will be the 5th 6th and 7th BE/SEC/ACC/B12 teams who most years play some pretty good ball. Not sure what the lowest seeded at large bid has been in history but I doubt they have ever been worse than an 12 seed.

 

[Icebear]

You make some valid points. The first 2 rounds of the tournament would be more competitive for sure, but do you actually think any fans would watch the games other than friends and family?

Bingo. I believe it will create more expenses than it will pay for.

 

[Tonyc]

Getting more teams in the tourney is good in that it gets more fans from other schools excited and creates interest which is good for the game. Yes they're not UConn but they all have to start somewhere.

Should UConn get a game off? I don't like that. Yes UConn has been ranked in the top 4 for a long time and gets an easy first game or two. But what happens to a 4 or 5 seed that advances and has to play a 2 or 1 seed. They could be sharper then a team sitting out a game.

Conference tourneys are different. The NCAAs are for the NC and I frankly don't like watching UConn blow out a 16th ranked team but not everybody else blows them out. For that matter UConn beat up on everybody this past season.

If the NCAA decides to try it, I don't think it could help the better teams sitting out. That first game against a weaker team is a good tune up. JMO

 

[DobbsRover2]

Indeed, one question would be, is getting a one or two round bye an advantage? In NY the highest class football division for eastern NY has a rotating bye among three sections over a three year span in the state playoffs, and none of the sections want the bye because their champs usually lose, 12 times in 17 years. In the MLB it also seems that teams that had 1-game playoff like the Mets in 2000 had the advantage of being revved up to win.

Although the current system has its charm, it would be better if the regular season conference winners were made the automatic tournament goers, while the tournament winners just got a nice trophy. Sure, the excitement at the conference tourneys will go down, but the quality of team at the NCAA tourney will improve.

Generally it seems that there is a lot of teeth-gnashing over the attendance and quality of the first and sometimes 2nd round games of the NCAA tourney, so I can't see that split tier system would make a horrible difference, and maybe it would allow a placement of more geographically similar schools for the bottom 32 so that attendance and interest might be improved.

I don't love the 4 games to an NC route since it does mess with historical stats, but it is the de facto situation at the moment and does put more value on the regular season. The big concern taken from another proposal for the "4 games for the top 16" approach would be the question of whether allowing an elite team like UConn to host the regionals would basically be rolling out the red carpet to another NC. On the second thought, that's not really a big problem.

 

[UcMiami]

Dobbs - not sure the bye/double bye is that big a deal in WCBB because the teams with one bye (5-8 seeds) would be playing their first game in the same time slot they currently do and they would be matching up against basically the same opponent as in the current format. The teams with a double bye would be playing two days later than normal but again against the presumably same opponent as the second round game currently and a 1 or 2 seed shouldn't lose to a 7,8 or worse opponent if they have been properly seeded. Might be a little tougher on the 3 and 4 seeds but those games have always been pretty competitive. the format would be
Round 1 - 9 v 16, 10 v 15, 11 v 14, and 12 v 13
Round 2 - 5 v 12 or 13, 6 v 11 or 14, 7 v 10 or 15, and 8 v 9 or 16 (really the same or worse than they would currently face.)
Round 3 - 1 v 8 or worse, 2 v 7 or worse, 3 v 6 or worse, and 4 v 5 or worse.
Rounds 4 - 7 are exactly the same as current opponents.
I agree there is a little rust being knocked of in the first games of the tournament currently, but not that much. There doesn't seem to be quite the same 'momentum' issue in basketball as in football, and you can practice full speed in basketball - football you don't want to be beating up on your team physically so it is harder to stay 'game ready'. And with the large number of teams still involved by those round three games, the top end is still way more skilled than the bottom end. I think a few 4 seeds would probably prefer being 5 seeds, but ...

 

[Icebear]

They should consider the attendance at the Div 2 and 3 tournaments as an indicator. It is most often miserable. These games would be lucky to draw 500-1000 fans. You have now taken those teams further away from what will likely be seen as the "real" dance.

For many of these low placement teams the reward of going to the tournament is the once in a lifetime chance to play a UCONN TN, Stanford, ND, etc.

 

[Phil]

Bingo. I believe it will create more expenses than it will pay for.

You brought up an interesting potential concern.

While I haven't thought through whether the revenues would likely be higher or lower, I'm intrigued about your supposition that expenses will be higher.

One large component of expenses is the cost of running the games themselves, arena fees, referee costs, score keepers and the like. Under the DBF and the traditional format, the number of games is the same, 63, so I'm not seeing a major difference there.

Another large component of costs is the travel costs. It wasn't immediately obvious to me whether there was more, less, or about the same travel under the options.

I created a spreadsheet to lay out the options. For numbering convenience, because the last four rounds are identical under either option, I kept the numbering of the six round option as 1-6, and for the double-bye, seven round option, I numbed them 0-6, so that Rounds 3,4,5 and 6 refer to the Regional semis, Regional finals, Final Four and Championship games in either option. The additional round in the DBF is round 0


Interestingly, the number of trips is almost exactly the same. Under the DBF, it will always be 68 trips (unless they decide to allow teams to host regionals, for the moment, I am assuming not.) For the traditional format, the number of trips depends on how many predetermined sites has a local host. I used 2013 as an example, and all but one did, so the number of trips needed is 69.

If you assumed that the increase in expenses was due to more travel needs, that doesn't appear to be the case.

 

[stwainfan]

I do think some changes are needed. The way it is now the 1 vs 16 seeds. The one seed basically gets a bye in first round anyway. I do think that when a team gets beat they should be out.

 

[UcMiami]

They should consider the attendance at the Div 2 and 3 tournaments as an indicator. It is most often miserable. These games would be lucky to draw 500-1000 fans. You have now taken those teams further away from what will likely be seen as the "real" dance.

For many of these low placement teams the reward of going to the tournament is the once in a lifetime chance to play a UCONN TN, Stanford, ND, etc.

But I think they would now get the chance to actually win a tournament game before being crushed in the wake of one of the legendary programs.
I also think that the further you get away from the top 10 in WCBB the less the disparity in skill and the harder it is to properly rank teams so having in effect two 'play-in rounds' to select the 16 teams to be added for the 'real tournament' would not be a bad idea. I would expect the upsets to be quite numerous in those first two rounds which in general should address one of the 'complaints' of WCBB being too predictable. As for attendance ... some limited potential schools do a lot better than more storied programs in D1 on that score, and while the lead time for games would be quite short, with a little bit of effort you might improve the overall attendance for that first weekend. CT does pretty good with first weekend games (at a heavy premium on pricing) but a lot of places that currently host do not. Spreading the wealth down the food chain to some lesser programs might beef up the draw. And having two more competitive games out of the three might also improve the attendance prospects. It might be great for the players on a 16 seed to play Uconn and great for some of the Uconn fans, but for the rest of the fans seeing Uconn play good basketball does not mitigate the agony of watching an 80 point laugher.
Part of the problem with the tournament (and this is true for the men as well though no one would ever admit it) is that the minor conference don't really belong on a guaranteed basis and with the guarantee being to the tournament champ and not the regular season champ you add in the possibility of a fluke team that beat a much better team on a one and done basis. I don't see them setting up a BCS kind of system where the only guarantees are for power conferences (and certainly hope no one ever considers that!) And a team like Louisville who provided 'the greatest upset in the history of the women's tournament' would have ended up in a bubble position if you went to a 48 team slate with conference champs.

 

[DobbsRover2]

Dobbs - not sure the bye/double bye is that big a deal in WCBB because the teams with one bye (5-8 seeds) would be playing their first game in the same time slot they currently do and they would be matching up against basically the same opponent as in the current format.

It's not necessarily the same as in HS football where a team has two weeks to get in the wrong mindset, but on the other hand, I wouldn't love being especially a 3 or 4 seed team facing usually a 6 and 5 seed team that has had a good tune-up. Also guessing that those first game upsets for even top 4 seeds like the 1 and 2 seed teams will increase dramatically beyond what you would expect if an 8/9/15/16 whatever team was matched against them in the same rest period. A one-day rest sounds perfect grounds for breeding upsets.

Back in the early days of the WCBB tourney in the 1986-1993 years, the NCAA moved from a 32 team format during 1982-1985 to a 40-team format from 1986-1988 and 48 team format in 1989-1993. That meant there were 32 times when a #1 seed met a #8/9 in its first game, and their record was 29-3. Since then during the 64-team era, the #1 seed has gone 76-3 in those matchups after both teams have gotten a game in. So that's 47 more games the #1 won for the same amount of losses, a pretty significant difference. The #3-4 seeds won at least 6 of 8 of their first game contests in four of the five years in the 1989-1993 era, but in 1990 they only achieved a 4-4 split, and overall they were 31-9, and they were the host teams.

Not saying that a little more chaos and surprises are a bad thing, but we do need to accept the consequences and be prepared for troglodytes going all wacky brained because WCBB isn't as safe and predictable for them anymore.

 

[UcMiami]

Back in the early days of the WCBB tourney in the 1986-1993 years, the NCAA moved from a 32 team format during 1982-1985 to a 40-team format from 1986-1988 and 48 team format in 1989-1993. That meant there were 32 times when a #1 seed met a #8/9 in its first game, and their record was 29-3. Since then during the 64-team era, the #1 seed has gone 76-3 in those matchups after both teams have gotten a game in. So that's 47 more games the #1 won for the same amount of losses, a pretty significant difference. The #3-4 seeds won at least 6 of 8 of their first game contests in four of the five years in the 1989-1993 era, but in 1990 they only achieved a 4-4 split, and overall they were 31-9, and they were the host teams.

Thanks for this research - interesting, but I suspect the changes to the game, the improvement in talent and coaching at the top, and the small number of upsets in both cases could cause the disparities - not sure when the shot clock and 3 pointer were introduced, but those two changes alone could have an huge impact on 8 seed success rates. And I guess I would be OK with a little more uncertainty. I kind of feel a 1 who loses to an 8 at any time in the year or in any situation during the tournament didn't deserve the 1 seed.
I agree the 3/4 seeds might be in bigger trouble.

 

[DobbsRover2]

WCBB got the 3-pointer in 1988, so the perimeter shot would have been in effect for three of the 8 years of #1/#2 seed byes. It's possible that a new weapon could have had an effect on some of the upsets, as some elite teams may not have adjusted fast enough in the early years. On the other hand, 3-point shooting can be a bit flukey and could also cause the downfall of even the top teams under special circumstances, i.e. Baylor vs. a #5 Louisville this year.

And looking at the 1 vs. 8/9 stats another way, the upsets during the last 20 years have occurred at a rate of less than 1 every 26 games, which would seem about right, especially considering that many of those games were won on home courts. So about every 6 1/2 years, a #1 seed on average has gone down in the 2nd round. In the 1986-1993 period, they occurred at a rate of 1 every 9.3 games, or more than once every 3 years (1 per 2.67), which sounds way way too often. (Consider that the ESPN pundits mainly picked all four #1 seeds this year in the FF.) I wouldn't downplay the power of the bad-bye effect. When one team has won and is fired up and another is trying to shake off the rust of a long lay-off, bad things can happen to the favored team. And if it's a 4/5 scenario, I might rather be the 5 unless the top seeds get the host role for that one game, though from an attendance perspective that's not really optimal.