Potential improvements to invite matching?
 

I was thinking about release figures and software and all that, and if your chapter has a return rate lower than in previous years, you can get burned. You are supposed to invite enough girls each round that the number of girls you have at each party equals the number of girls remaining in recruitment divided by the number of chapters.

So, let's say there are 200 girls in recruitment, ten chapters on campus and 5 parties this round. You want 20 girls at each party, for a total of 100. If you are expecting a return rate of 100%, you issue 100 invites. If you are expecting a return rate of 50%, you issue 200 invites. 66%, 150 invites, etc.

The problem comes in where you are expecting 66% return, and only get 50%. Now you have 75 women attending, where you'd like to have 100.

So, two crazy suggestions, let me know what you think:

1) If a chapter's total number of women returning is not enough to fill up the number of parties - 1, they only hold the number - 1. In the above example, if you had fewer than 80 attending, you would only hold 4 parties instead of 5. This helps the parties look fuller, and gives the chapter a break.

2) Allow a second invitation list below the first one, so if not enough women accept, it goes to the second group, as in final bid matching. In the above example, you could submit 150 invitations, plus up to 50 others you wouldn't mind inviting if you weren't limited by the number of invitations. If only 75 women accept from the first group, the software looks at the second list, and if any of those women do not have a full schedule, they are added to your party list.

I know that to do this right, you would technically have to rank the second list, but chapters could do it totally randomly and still come out better than if they didn't have a second list. I suspect in most cases, too, the second list would not be everyone else, but may be short enough to rank.

I am certain that software could handle these suggestions with little additional work for the greek life office, so what do you guys think the pros and cons are?

They both look good to me but I don't have any direct experience with it like you do.

About the second option, is it better from the PNM perspective than in the old days of no release figures because it would only invite "extra" girls if the chapter would be aware that they actually needed to look at giving them bids?

(Remember that one of the benefits of release figures was to keep groups from stringing girls along until right before pref? If everybody gets a back up list for parties, what's to keep it from being about stringing people along?)

I think they are good suggestions, but I wonder if the problem could be solved somewhat by simply changing the release rate for the next event at that chapter? Does it not work that way?

The only time where it is actually a problem that your return rates are lower is for bid day. I understand that return rates will fluctuate a little from year to year, but I think something catastrophic would have to happen to lower them to the point where it would make it very difficult for an otherwise healthy chapter to make quota, even if they have to snap bid a few women.

I think, in some situations, having really empty parties could be killer. Also, if you up the number of invites for the next party, it is possible that you could be inviting a woman to the next round who did not attend this one. That can also be killer.

I see what you mean.

I kind of forgot about systems with a smaller number of groups. I'm used to systems where the drop off between the 2nd to last round and pref cuts the number of parties a girl can go to in half.

At a campus like that, you could just not release many and have almost double the girls you would need. (Assuming of course that you'd really want to pref them and that the smaller party size the round before didn't turn most of them off.)

But at a campus where there are only four groups to begin with and a PNM is cutting from three to two or something before pref, there's no way to make up the numbers really.

I think this could be good. Only panhellenic and the chapter would know how many parties they were hosting. The PNMs wouldn't know. The parties would be full and the chapters would only be looking at PNMs they were interested in.

Are you all going to be able to do it this year at Auburn with the extra parties you built into the schedule? It seems like it makes it easier to vary and not make it obvious to the PNMs since all chapters will have at least some off parties, right?

Do PNMs notice party size and come to the conclusion that a smaller party = less desirable chapter? If anything, I would think that PNMs might notice a smaller chapter size and conclude that the chapter is less desirable. We all know that tent talk happens. If ABC is the "top" chapter on campus, they will have smaller parties because they will have to release more women with the RFM. Tent talk will sway PNMs more that party size will.

I was pretty oblivious to the number of sisters in each chapter during recruitment and the party size. If party size appears to be an issue, condensing the parties for certain chapters may be a good idea.

I think you are misunderstanding the RFM. The idea is not to make the top chapters have fewer women at their parties, it is to make them release more women because the women they do keep are more likely to show up.

I think there are PNM's who would assume fewer PNM's = less desirable chapter.

This quite often takes place. And the converse is also true. If some chapters are releasing heavily early in the process, their second round parties may appear quite small - even though they are a historically strong recruiting chapter. It's really all about the chapter's preference - Panhellenic should accommodate them. If you want fewer parties with greater people in each, I say why not!

In RFM, this is called a Flex List. It is used only with Priority ranking, though, rather than Accept/Regret. (Which is another "pro" for Priority.) In A/R chapters don't know their return rates until after all invites have been issued and either accepted or regretted. Chapters are given a certain number to flex from both their keep and release lists. So if a chapter is doing uncharacteristically bad, they get to bring back more women from their release list (the ones they have indicated on Flex that would have been invited if they could). If they are doing uncharacteristically well, some women can get pulled from the keep list (the ones they have indicated on Flex that they would release if they had to.)

You are correct. ICS handles it quite nicely!

The campus I advise does this. There is what is called a "flex list." Each chapter has a different size flex list, depending on their average return over the past 3 years and it works both ways. Example: A chapter has a flex list of 20. This means that they must choose, in order of their preference, 20 women that they did not invite back that they would be willing to invite to keep their numbers "ideal." Additionally, they must choose a list of 20 women from their invite list that they are willing to release if their return rate is higher than expected.

For obvious reasons, there is no "flex list" requirement before preference round.

Sorry, I must have been typing at the same time.

The flex list sounds great!

What determines what system a campus uses? Is there expensive software that the system would have bought that determines it, or does every campus have the choice of choosing priority ranking and flex list over accept/regret every year?

I agree with Violetpretty that chapter size probably matters more than party size, but a smaller chapter who also find themselves with sparsely attended parties kind of faces a double problem in terms of impressions, I suspect. I think a chapter would be better off having fewer parties and having those parties seem fuller and livelier.

We actually did #1 at my campus, and while it made for a better party, trust me that the PNMS do know. All they have to do is compare party times with their friends to figure out that everyone who went to ABC went at 3:00 p.m. Even if they're not thinking that way, somehow the word gets out.

I think that the reducing the number of parties idea would better at larger schools where there are more PNM's or in earlier rounds. It would be far less noticable if XYX only had 9 parties instead of 10 or even 4 instead of 5. At smaller schools, it is very noticable. I know it's happened a few times at my college and while it does make the chapter members feel better because they X larger parties, as a Rho Chi, I know that the PNM's definately knew what had happened which defeated the purpose...

Does it completely defeat the purpose? I'm really asking, not being rhetorical. I have no idea.

I don't think you are ever going to be able to fool the PNMs about which groups most people consider more desirable, but if you made the experience of being at that chapter more like being at other chapters that round, wouldn't that help the chapter in the eyes of PNMs who didn't purely go by tent talk or what other PNMs thought. I know they are few and far between anyway, but a girl who goes to a sparsely attended party with lower overall energy has an "objectively" different experience at that chapter than a girl who although she knows the chapter has lower returns attends a full party with energy comparable to the others she attends.

If you are asking what determines Priority over A/R, that is up to campus preference. An that preference is usually determined by history. If they've always done A/R, it is harder to convince them to go to Priority. Technically you don't have to have software for either - it can be done by hand. But it is very time consuming to match invites - even for schools that have only 4 or 5 chapters. There is currently a (slight) trend toward Priority, and those that are switching to Priority over A/R are doing so in order to utilize Flex and because it is easier for the RFM Specialist to consider current year chapter performance when working on invitation matching.

To my knowledge, there are three options out there for recruitment software - D&D, Innova, and ICS. I've never used Innova, but as far as being user-friendly, ICS is fabulous. It is all online, so chapters can submit lists at any time from anywhere rather than having to hand carry a list or a disk to the Greek office. Same for receiving invitation lists.

Technically you could do the Flex if you were matching invites by hand, but it would be a lot more back-and-forth and time. Luckily there aren't many campuses still doing that. I don't recall the cost of ICS right offhand, but I do know that several smaller systems have started using it without breaking the bank.

Thank you. I was mainly wondering if campuses basically decided what to use and paid for it each year (maybe through a contract with a recruitment management service?) or if it was an expensive one time purchase that then drove the decision for several years.

Ahhh. I see what you are saying. I haven't been on the purchasing end myself, but I THINK that for ICS, it is a year-to-year thing. A campus Greek Advisor might know more details. :)

We used the ICS flex list last year and it worked well. My only qualm is that after voting on so many girls, the sisters get tired when it comes to pick the flex list and it's kind of time-consuming to rank the flex list and you may be cutting it close to when you have to turn in your votes. I think your panhellenic council has to decide if the flex list approach is best for your recruitment. If many of the chapters have had issues with fluctuating return rates over the last few years, it may be a good idea to switch to the flex list system.

One more thing: for our preference round, we actually requested one MORE party because we wanted there to be fewer PNMs in the room. It made it easier to match our sisters with the girls for conversations!

Okay, I'm bumping an old thread with another math question. My understanding is that, at a large recruitment using RFM, bid matching is run several times with different quotas, and the one that results in the highest number of women placed is used. That, is, they put 85 into the computer, see what happens, put 86 into the computer, see what happens, etc.

The question is, when there are QA's, what difference does it make? Do the ultimate Deciders of Quota look at how many women match to their first choice under each scenario, or just how many match at all? Do they consider how many chapters make quota under each scenario, too?

I was thinking about this as a multi-objective optimization problem. One objective is to maximize parity, and the other is to maximize PNM happiness. Trying to place the maximum number of women prior to QA's doesn't seem to accomplish either one.

They look at all of it. It's not an either/or thing. So, it could go either way at any given time. The goal is to place the most women in the highest placement possible. So they try to determine what that is.

Okay, my question was mainly around whether they looked at WHERE women placed.

I'm always interested in the pieces done by hand, rather than automated, and why/how they are done by hand. In this case, automation would be an Integer Program (IP) something like:

Maximize "Total PNM happiness"
by changing the variable "quota"
subject to "all chapters make quota"

Basically, you'd come up with some sort of measure of PNM happiness, like 3 points for a woman placed in her first choice, 2 for a woman placed in her second choice, and one for a woman placed in her third choice. Then each time bid matching was run for a different value of quota, a value for Total PNM happiness could be calculated, and there'd be a "best" value of quota.

I'm not sure that "all chapters make quota" would be a hard constraint, though, as we know that won't happen at all on some campuses. Rather, I'd adjust the objective function to call it something like "CPH Strength" and define it as "Total PNM happiness - missed quota penalty," i.e. for each chapter not making quota, we'd subtract some quantity from the total PNM happiness.

Of course, you can't *really* just hand it over to the computer...who decides if a PNM in her first choice is three times happier than in her third choice, or four times? Maybe the points should be 4-2-1 or 6-3-1 instead of 3-2-1.

Even if someone were to write the program and hand it over to the computer during bid matching, the fact remains that the design would require a huge amount of up-front subjective judgment calls. In any case, I think it's fun to think about, and a good example of why real-world modeling is very difficult.

I think I'd make missed quota a huge penalty, because that way, if a chapter was NEVER going to make quota, it would apply to every scenario, and wouldn't make a difference, but if a chapter makes quota under some scenarios and not others, it would automatically avoid the latter. But then again, does it matter if a chapter misses quota by one instead of making quota? Then why penalize by number of chapters missing quota rather than the number of women they are below quota?

At a school that has reached parity (more or less), I think weighting QA's to rushee preference would be an excellent idea (while still trying to retain something resembling parity). But for schools where there are 1 or 2 well below the rest, the sad but truth is those chapters are going to get the emphasis. As far as I know, the weighting metric is determined differently at each school, so this may actually be how it works. I have not been invited behind the curtain as yet, so I don't know how it really works at one or many schools.

Other than that, you made my head hurt.

Oh! I hadn't even gotten into placing QA's yet...I was still working on determining quota. But yes, that's a good point, QA placement is another thing done by hand.

I'm not a computer expert nor a math whiz. That being said, whether it is hand matching or computer matching, it always matches the woman to the chapter not the chapter to the woman. One of our members here posted the latest version of the NPC Green Book (MOI) a day or so ago and I strongly recommend that all of you take a look at it.(It's a sticky near the top of the page). Most of your concerns will be addressed. They have a method for adjusting the number of parties. The Flex List adds women if need be. The MOI lists things a computer program should do and duplicating the hand process is one of them. For those of you who have never done hand matching, the PNMs name is called with her first choice. If it is a match on that group's first list (or remaining women available which equals quota), then she is matched to that group. If she isn't on that list, then her card is set aside until all are gone thru and then they start over to see if her number has come up. And this is repeated over and over until there is gridlock - no one is matched after going all the way thru the remaining cards one full time. Then the method to break gridlock is implemented. The computer programs must adhere to this as well.

I'm not talking about the matching itself. I'm talking about the fact that the matching is run more than once, with several different quotas, and then the "best" one is chosen.

So, there are two things that are still done more or less by hand, choosing quota and placing QA's. I'm just talking through why that stuff isn't automated. Either the MOI has rules for it, in which case it can be automated, or these things are really subjective subject to the GA's whim.

My understanding is that IS automated.

It is automated. The computer places the QAs based on matching the girl to the group in the same way as quota. If the numbers look skewed, they can run it with a different quota. But it's nothing but a math formula. How many times will 10 chapters go into 287 women with the smallest remainder. But stop trying to make it something it isn't. And the school's assigned RFM specialist has to sign off on the final match anyway. So she can override the FSA if she thinks something hinky is going on.

Thanks for the info. My intention is not to imply that anyone is intentionally screwing it up or going in with an agenda, I have just heard many different things about how QA's are placed. The only reason for the bump is that I enjoy getting into the really nitty gritty details of what the computer can and cannot do vs. what requires some kind of subjective human judgement.

In hand bid matching the QAs were placed in the exact same way - only with a cap of 5% of quota. If you will look at the sticky with the MOI/Green Book, there is a whole section on what a computer program must do. It's very detailed.