A Radical Idea for MIT Dining

From ThePlaz.com

(Redirected from Radical idea for MIT Dining)
Jump to: navigation, search

So why do all of the 4 dorms need to offer breakfast? Let’s say 30% of the students want breakfast. That’s pretty much one dorm’s worth. Why not move all those students to one dorm????? The only flaw in this is the UA rhetoric that students should not choose dorm based on culture or dining – but I think that they already do. I chose Baker because it has a dining hall. People in EC chose EC because it did not have dining. I think that is perfectly fine – it is an essential part of the choice of where you want to live – it’s an essential part of the culture of the dorm. So why don’t we just let people shuffle around to what they want? I think that people would be far happier if they were able to explicitly choose which dining plan they want.

The Housing Office should put out a mandatory choice of what option people want – like they do with the freshmen survey. Responses would be mandatory and binding (see below):

  • I want to live in a Breakfast, lunch and, dinner dorm at $5,000/year
  • I want to live in a Breakfast and Dinner dorm $3,800/year
  • I want to live in a Dinner dorm $1,800/year
  • I want to live in a cook for self dorm $0/year.

Then HDAG would look at the results. If ~300 people chose option one, then one dorm with ~300 beds would become the B,L,D dorm. (At that dorm, the selected meals would be mandatory.) The students would help select which building that would be, taking into account the size of people that chose that option, and where a disproportionate share of the votes for that option came from. (Perhaps restrict the choice of buildings to only affect the 4 dorms with house dining now. Perhaps exempt McCormick, idk. But in making the choice of what to offer where aim to affect the minimum number of people. Actually that last option sounds best. Just calculate how many people would have to move based on what you offer where, and maximize that. EC, say, almost everyone would want to continue to self cook, so making EC self cook would minimize moves.) They would then do this for each level of service, starting up new house dining or shutting down house dining halls even, if need be. Whatever makes the least people move.

Then once HDAG decides which dorm offers which service, there would be a form of the readjustment lotto. You could only choose the dorms in the category you picked (in order to make the initial choice important; or maybe this is unnecessary, you can choose any dorm [in which case it is the normal readjustment lotto]), plus maybe (I have not thought this through) your current dorm, by which you would be accepting whatever service they offer. There would be no guarantee they could move you in, but since the housing office knew approx. how many student wanted each option there would be approx. the right number of beds under each option.

It is good for the housing office because they know very well how many people demand each option, so they can vary the number of beds in each option to demand. And then the students can choose which actual dorm to live in among the ones which offer what you want. Students would be for the most part happy because they got the service they wanted. They may have to move around, but at least they would be very likely to have the dining service they want. Culture at the new dorms would readjust and might become stronger, because almost everyone there wanted that meal plan. Cultures would also become more distinct because there would be more variation between the dorms. And its all based on the assumption that dorm culture depends on dining. And since we would target minimizing students who move, determining what service goes where is a mathematical problem, hopefully devoid of arguments of survey and representation bias.

Thoughts? Please forward. Permission to republish granted.

Contents

Question

(Question asked how select what each dorm offers?)

Well the whole point is that you don’t prespecify which dorm has what. The sole goal is the MINIMIZE THE NUMBER OF MOVES. Since we are minimizing the number of moves, what is chosen for a building is what most of the people WHO ARE IN THAT BUILDING NOW want (except in extreme cases). If your current building “voted” to have something you don’t want, you could choose to stay and accept it, or choose among the dorms which offer your selected option.

Since the moves algorithm would not apply to W1 and W1 is really the only place that offers lunch, it might end up being the lunch place. If, however, only ~50 people in all of MIT want the all you can eat lunch, that tells us that perhaps AYCE lunch would not be a good choice anywhere, even at W1.

We should also split hot breakfast and grab and go breakfast into 2 distinct options. There would still be an option to hold just dinner (not even grab and go breakfast) and a no-food at all option.

(Also I am operating under the assumption that the cost of a dining hall is mostly fixed. I’m guessing 80% of the cost is there if they serve 1 meal or 300. This is why we want to reduce the service that is offered to only what people want.)

Defending the Plan

(Another email response; this time in response to an East Campus resident)

At its core all it is a binding survey asking people what they want with actual prices and then allowing them to move around (which they already can do). The only new thing is that the housing office will now vary the number of beds under each system based on demand – not just going with the default of all 4 dorms B,D.

EC people will see no change whatsoever. The algorithm takes care of that – so no need to argue for a special exemption. If you go to MIT, you should be able to see that the choice which would minimize moves is to continue making EC cook for themselves.

The other “new” thing is the radical truth that students pick dorms based on dining options – maybe not cost – but options of what they can buy. The thing is, students must pay for these options. Mandatory is important because I estimate dining is about 80% fixed cost – 1 meal or 300/night. So what you want to do is to concentrate the people who demand dinner into a single place – we kinda do it already – people in Baker chose a meal plan - but it would make people far happier than paying $2,000 for a plan they don’t want. If tons of Baker people vote no dining, than it would minimize moves for Baker dining to close. If tons of Baker people wanted B,D we would have that. And a minority does not want it, moving out is not so bad. Problem solved – minimize cost and maximize happiness.

It is true that the reason the admins is doing this is for admissions. Dining was my biggest concern coming in – and like (the EC author) it all worked out (except I did not learn to cook). But now MIT can go and say, we got a B,L,D dorm, a B,D dorm, a D dorm, and plenty of no meals dorm.

I think what a lot of people don’t think about is the COST of providing all these options in many places across campus. By aggregating the breakfast eaters and providing breakfast to them, and them alone, we are minimizing cost. And they are likely to share a common culture, stronger, perhaps, than some of the current West Campus houses. And MIT now offers that option to the people who want it, just like they offer Baker Dining to those that want it.

Clarifications

I think there has been some misunderstanding of my proposal. Let me go through some scenarios, because I think they will help explain the proposal:

Let’s take EC. The vast majority (say 95%) of people in EC would vote to continue not having a dining hall (unless there are some closet dining fans). Since it would be move minimizing (5% would need to move to be happy vs 95%) to not offer a dining hall at EC, Pritchett would stay closed. The 95% who voted for no dining would stay in EC and be happy, no issues. The 5% who voted for another plan could choose to stay in EC (even though they would not get their dining choice) OR they would have a virtually guaranteed spot in a dorm with the dining program they want. They could then choose among the dorms that offer the option that they want – according to the other factors of culture.

(I have not decided if they could choose among any dorm, ie ones other than their current one or all of the dorms with the dining level they desired. The downside is that by changing their minds they are taking away a bed in a dining level from someone who voted for it, which would not be fair, since space was allocated for them at their desired dining level. Now since students can already choose to stay in their current dorm and thus no bed is guaranteed, this may not be as big an issue. As I am thinking about this, I think that we should allow people to move anywhere (just like we already do).)

Then let’s say 300 students in the dorm system voted for hot breakfast and dinner. The dorm which would require the least students to move would be offer this plan. The least number of moves would most likely be in whatever dorm had the strongest demand for hot breakfast and dinner. Let’s say it is Baker. I (as a Baker resident) can now choose to stay in Baker and buy into breakfast, or I can move elsewhere with a spot open for me in one of the dorms that has my dining option. I choose which new dorm that is based on culture. However, there would at least be room for me somewhere, so I would be able to move out and not pay for something I don’t want. This is because the number of beds available at MIT under each dining level would roughly match the demand for each dining level. Under the HDAG plan many students said that they would move out. But where would they go? And who would fill their empty beds? It is pretty clear that almost no one would be able to move out because MIT had vastly over supplied the number of hot breakfast + dinner beds.

If only 50 students voted for hot breakfast we would have a decision to make. We could decide that the demand was too small and those students would have to settle with cold breakfast. Or, if the administration felt obligated to offer hot breakfast we could squeeze it in somewhere (minimizing moves) and force other people their eat that hot breakfast. Or we could offer the students to aggregate in a certain dorm and then MIT could do optional hot breakfast, eating the loss to meet student demand. Again the student would have a choice: stay where I am and take the program that makes the majority of the people in my dorm happy or move to a dorm that fulfills my dining wants.

If the plans the Institute can offer without the subsidy are too expensive, then many students who currently live in a dorm with a dining hall choose to not have a dining hall. Then to match the number of beds with no dining hall to demand, one of the locations would close. Again it would be the location that would be move minimizing and again, students could choose to stay or move to a dorm with the dining level they like.

Some have pointed out that it’s not dining that defines a dorm’s culture. Well fine. If you want to stay in your dorm, you can stay. You would be subjected, however, to the dining level which minimized the number of other people who would have to move out of your dorm. I think that is perfectly fair. You either have a virtually guaranteed spot in the dining level of your choice, or you can accept the dining level which will minimize the number of other people who would want to move out of your dorm. If you want the dining level your current dorm picked, awesome, there would be no change. And if you would prefer your desired dining level above your current community, you have a spot for you in one of the dorms with your desired dining level. You could then choose the particular dorm at that level based on all the other factors of culture.

I also realized that under this system we might not even need mandatory. If all of the people who want breakfast would live in one dorm, then the system would do better. The problem is that I estimate ~70-80% of the cost of a house dining meal is fixed. Those costs are the same if they serve 5 meals that night or 500. That is why mandatory makes some sense. But the basic idea is that if you want breakfast, you need to pay for breakfast. Aggregate the breakfast eaters together (well what I am proposing is to offer them the choice of keeping their current dorm vs having a spot in a dorm that has breakfast) This is much better than having every option everywhere were students have to pay for it if they eat it or not. Having breakfast everywhere is not fair either because everyone is paying the cost of offering breakfast in 4 places, if they want it or not. Also, right now everyone else who buys food on campus subsidizes House Dining to the tune of ~$600,000/year. I don’t think this is fair. Why should you, EC student, be paying for my house dining when you buy pizza at Stata?

My plan minimizes the cost of the service be providing it only in the places people want. It minimizes the people paying for service they do NOT want. It then gives students a choice of their desired dining level OR their current community (whose dining level was chosen so that it minimizes the total # of people who would have to move out of all dorms). It strengthens west campus culture by increasing the distinction between the dorms. And lastly if you agree with the majority of your house on dining level, then you are happy right where you are. Perfect!

Further Clarification

Also one clarification: this is all about house dining, and how much people want to support it. House Dining is very expensive to operate, so we want to minimize where we offer it to where students want to pay for it. Dining levels again:

  • Hot Breakfast, Dinner $3,800/year (the HDAG proposal)
  • Grab and Go Breakfast, Dinner $3,600/year (estimating numbers based on what I heard from HDAG members)
  • Dinner $1,800/year (similar to what Baker, Simmons, next, McCormick have now, but more fiscally sustainable)
  • No House Dining (the current level in EC, Senior House, MacGreggor, etc)


I think what I am trying to get at is: allow the current students in each dorm to select what dining level should be offered. Then:

  • if the dorm’s desired dining level is what you desired – perfect, you can stay, just like always
  • if the dorm selected a different dining level, such that it minimized the total number of people on campus who would have to move to make people happy
    • you can either decide to stay in your current dorm and accept the new dining level
    • or you have a virtually guaranteed spot in a dorm with the dining level of your choice. Which one that is, is up to you and a lottery.

The constraint is cost. We can’t offer everything everywhere. It’s way too expensive and you have a lot of people paying for stuff they don’t want. (The biggest complaint I hear from Baker people). That is why we want to give people a choice: the community they have now (whose dining level was picked to minimize the total number of people on campus who have to move) OR a virtual guaranteed sport in the dining level of your choice. Offering breakfast everywhere since 15% of people in each dorm want it is wrong. If you want breakfast you should decide: stay where I am for all the other culture reasons and give up dining hall breakfast or MIT has a spot for me in a place that has dining hall breakfast.

Another item which needs to be ironed out is how to allow people who don't live in a dorm with house dining to visit the current dorms. The administration had said before that they would want to make it fiscally difficult to opt out of the plan and then cherry pick what days to visit at the cash price. I don't know how they jack up the price of $8 breakfast any higher. Of course the dining halls make incremental revenue from visitors. However, if the visitors cherry pick the days they want to visit, while avoiding paying the fixed cost behind the operation that does not balance out to fiscally sustainable. With the current $500/student subsidy we have been getting house dining "on the cheap" below the cost of its operation. Everyone likes a subsidy and does not want it to disappear. But if MIT wants to get rid of it, we need to add $500 to the current plan's House Dining Membership to make it sustainable. With mandatory at least you are getting something for your up-front $1,100 (plus ~ $600 to include the actual meals – about what students spend now) – and it would incentive people to visit as often as possible. As I said, my plan optimizes even without mandatory. But what plan can we make that is sustainable, without the subsidy? The more people who eat there, the more you spread the cost around, the lower the cost per meal will be. Spreading the cost over students who don't want the dining level (hot breakfast, for example) is what the admin wants now, and I find deeply unfair.

A Slightly Less Radical Idea on Dining

Sorry for one more email.

I’ve had this plan proposed to me by David Lawrence (dlaw). It’s a slight modification – but I think it is better for a choice point of view. I just don’t know if the cost works out. (Apologies in advanced to David if I misrepresent his plan)

Each student would sign up for a dining plan (or none at all) they desire:

  • Hot Breakfast, Dinner $3,800/year (the HDAG proposal)
  • Grab and Go Breakfast, Dinner $3,600/year (estimating numbers based on what I heard from HDAG members)
  • Dinner $1,800/year (similar to what Baker, Simmons, next, McCormick have now, but at the fiscally sustainable price)
  • No Dining Plan (the current level students have in EC, Senior House, MacGreggor, etc) See below for not choosing a plan and then paying cash

HDAG would decide which house dining service was offered where. Again the decision would try to minimize the total number on students who would have to move to be happy with their dining plan (BUT THEY DON’T NEED TO ACTUALLY MOVE AND THEY DON’T NEED TO GET THE PLAN THEIR HOUSE OFFERS; KEEP READING! The algorithm just tries to offer their desired plan in their current home.) It would also minimize cost by minimizing the locations in which service would be offered to about match population (ie 300 people want breakfast, one dorm). What would be offered where would be announced.

You would be signed up for your desired dining plan, no matter where you live. You would have the choice of moving to the dorm which has a dining hall at your desired dining level or staying right where you are. In any event, you have selected your meal plan. Let’s say I am a Baker resident who wanted dinner only. But the algorithm said that it would be move minimizing to have Baker be the breakfast dorm. I could stay in Baker and only pay for and only get the dinner. Meanwhile any student who wanted breakfast, would be signed up for the breakfast plan at Baker (and another dorm if demand is there). They could decide to try to move into Baker or stay where they are and just eat breakfast and dinner in Baker (or wherever else those meals are offered) as part of the meal plan they selected.

Living choices and meal plan choices would be totally separate. On first glance, that seems suboptimal. - Especially in the administrator’s eyes, I’m guessing, who do connect living community with dining community. - But, if I understand correctly, that is what all of the critics of the plan say that is what they want.

Another item which needs to be ironed out is how to allow people who don't live in a dorm with house dining to visit the current dorms. The administration had said before that they would want to make it fiscally difficult to opt out of the plan and then cherry pick what days to visit at the cash price. I don't know how they plan on jacking up the price of $8 breakfast any higher….. Of course the dining halls make incremental revenue from visitors. However, if the visitors cherry pick the days they want to visit, while avoiding paying the fixed cost behind the operation that does not balance out to fiscally sustainable.

The flaw is the cost of running such a fancy operation. Currently dinner at Baker “costs” about $12. This is currently paid $3-4 of the HDM fee, $3-4 at the cash register, and $3-4 in subsidy. Not many people would pay $12 cash to eat dinner at Baker – what a meal currently costs. The only way to lower that cost is to spread it over more people by encourage more people to attend. Since most of the costs are fixed, we want to jam as many people into Baker as possible. $12 cash which would be the discretionary price without the subsidy will not do that. In fact less people would buy at $12, forcing them to raise the prices, rinse, repeat, death spiral. The house dining locations are just not competitive at complete discretionary. By that logic, they should be closed. But you will have trouble finding a Baker resident who wants dining to close. That is the challenge. You need to pay for the service you want.

Anyway, I’m not convinced that this plan will get enough people to sign up at the price that the plan costs. Everyone in Baker wants the subsidized price we have now. But if the subsidy is going away – we will need to pay more. Period. With the current $500/student subsidy we have been getting house dining "on the cheap" below the cost of its operation. Everyone likes a subsidy and does not want it to disappear. But if MIT needs to get rid of it, we need to add $500 to the current plan's House Dining Membership to make it sustainable. But do students want to pay $1,100 for nothing? Now with a pre-paid plan at least you are getting something for your up-front $1,100 (and you pay ~ $600 more to include the actual meals all you can eat– a bit more than what students spend now) – and it would incentive people to visit as often as possible – lowering per meal cost for everyone. The more people who eat there, the more you spread the cost around, the lower the cost per meal per person will be. Spreading the cost over students who don't want the dining level (hot breakfast, for example) is what the admin wants now, and I find that deeply unfair. And at least David’s plan and my plan minimize cost by only adding enough service that is demanded – not the plush one size fits all monster HDAG published.

You simply cannot just have the dining hall there and buy $5 meals 3 nights a week. That does not add up. At that rate they need to close. If you want the option of house dining, you need to pay more. Period. And you can either pay ahead of time or a very high $12 per meal cost. Neither option is good; Baker students, at least, want house dining. No one would likely do the $12 per meal cost, so you are left with some sort of pre-paid plan. If you want the option to be available, you need to pay the cost. You cannot do this without some sort of plan. [I should point out that at least in all of this you are not paying for services you do not want – which is exactly what the admin wants us to do under HDAG. I wrote a 5 page opinion against http://wiki.theplaz.com/File:Dining_Letter_10-14-2010.docx] Minimize cost by only offering service people want. Under David’s plan you can choose your dining plan and you can choose your living group separately. But if you want house dining, only a prepaid plan will work.

Plus, could someone please tell me if either my plan or David’s plan is not optimal – then what is all fucking world is! Something that gives students what they demand and where the costs are minimized and directed at the people who want that service level. Something where the costs of the services student want are reasonable.

Lastly, remember that if you are a student in a dorm without a dining hall, and you do not want a dining hall, and you never plan on visiting House Dining – nothing will change under my plan or David’s.

Leads to more choice

(Again sorry for spamming, but I think we are having a lively discussion on this issue)

The vast majority of people who wrote me trying to preserve the choice of culture independently of dining were from non-dining dorms. They argued that you should not choose your dorm based on culture – but then they say that the really hope no dining hall will open in their dorm. THEY BASICALLY SAID THEY LIKE THEIR CURRENT DORM BECAUSE IT DOES NOT HAVE DINING. And under my plan and David’s plan that stays.

Look freshmen select a dorm based on a variety of factors: culture, facilities, room type, and, dining options, and cost. P(C, F, D, $, etc) [Preference is a function of culture, facilities, dining, cost etc] Each student’s preference function is different. Yes cost is important. So far it has been more or less the same, so cost has not been an issue for the vast majority of freshmen. But cost is about to become a far bigger variable. My plan says is that if one variable changes then we should allow students to rerun their preference function. Perhaps that would result in no change for them (since dining option and/or $$$ is not important to them) . Perhaps when the student reruns their personal preference function it would lead to a change in where they want to live. Lots of students said that at the Emergency UA Senate meeting they would want to move out under the HDAG plan because of cost. THERE MUST BE SOMEWHERE FOR THEM TO GO! Right? If so many people want to move out of hot breakfast and dining $3,800/year than less beds should be available under that option on campus.

Look, all this choice in dining is expensive. MIT cannot offer every option to everybody everywhere at a price people are willing to pay. The vast majority of people I talk to in Baker say that they don’t want all that choice at that ($3,800) price. It is also ironic that that vast majority of people defending the “choice” of the dining hall forms live in dorms without dining halls. The 305 freshmen who listed Baker as their top choice in my year chose Baker because they either wanted the dinning under the current arrangements or the $600/year was an acceptable tradeoff. What my plan does is let them rerun their preference function and make a new choice based on the new plan. Their choice according to their preference function. EACH STUDENT SHOULD HAVE THEIR OWN PERSONAL PREFERENCE FUNCTION. Whose job is it to say that dining service level or cost should not go in anybody’s dorm preference function!!!!! This is the EC argument. Sure we would prefer that cost or dining service not be an issue. But the constraint is that we cannot have dining service choice (inside each dorm) and no associated cost. Look some people want to pay $3800 for all the hot breakfast and dinner they can eat. Some say no thank you. Give students a choice. Each student should be allowed to weight the dorms with the factors of their choosing and if a factor changes they should be allowed (not forced, however) to move out.


My plan leads to more choice. More choice than now. Far more choice than HDAG’s plan.

Thank you very much -Michael Plasmeier

(David’s plan goes a step further and signs you up for your desired dining plan, which may be at another dorm. In exchange, it allows current students to stay in their current dorm at the dining level they desire, which could be less than what the dorm offers.) Under my plan if you choose to stay at your current dorm you must accept the dining level that the minimizes the number of people who would have to move on campus – likely the majority view of people in your dorm. See previous email about how costs can only be met through a pre-paid plan.

What do you think will happen

I’m guessing 1 dorm would be hot breakfast, 1-2 would be grab and go, and 1-2 would be dinner only. I’d bet all the no dining hall dorms would stay no dining hall dorms. McCormick would be difficult - they would likely just have a simple vote of what they would prefer - with prices and the vote would be binding.

Effects on Financial Aid

I would say that it would not affect financial aid, except that the cost of the B,L,D plan should equal the fin aid allowance. Currently any money you do not spend on food is money in your pocket. I think it should continue to stay this way. The alternate would be to give people more financial aid because they don’t want to cook their own breakfast. I think this is a bad idea and one I think most people, including MIT, I believe, reject. (The flip argument is that you don’t think students should trade off food for money – in which case you might want to vary fin aid)

Also, remember many of the students have said that they don’t want the HDAG plan even if their financial aid was increased to cover the cost of it because they feel that the scarce fin aid resources could be used in more efficient means. The UA bill had a line on this.

So in my opinion trading off food for money is a good thing, and students should continue to be able to keep the difference between the fin aid allowance and their actual total expenditures.

Having a B,L,D plan above the financial aid allowance is not a good thing in my mind. The fin aid allowance should match the 21 meal/week plan, otherwise it would look very weird.

Want to move into Baker, but can not afford it

Hi (removed, EC resident),

Under the HDAG plan, currently scheduled by the administration for implementation next year: you could continue to live in EC without a dining hall just like today. However, if you chose to move into dorm with a dining hall (Baker, Next, Simmons) as a junior next year you would have to purchase at least 5 breakfasts and 5 dinners a week at a project price of $2,900/year. Freshmen in those 3 dorms (plus McCormick) would be required to purchase 7 breakfasts and 7 dinners under the projected plan at a cost of $3,800/year. Under the HDAG plan you get a slight break. Currently there is a readjustment lottery after each year. You can switch to Baker if someone from baker wanted to move to EC.

Under my proposed plan you could select which level of the dining plan you want (prices are projected):

   * Hot Breakfast, Dinner $3,800/year
   * Grab and Go Breakfast, Dinner $3,600/year 
   * Dinner $1,800/year
   * No dining hall (what you currently have)

At that point, (assuming enough people want to move out of a dorm to avoid the new plan; this was Nils objection), there would be a good chance there would be a spot for you in a dorm with the dining level you chose, since MIT would choose how many dorms to offer a program in based on actual demand). Which dorm that is I do not know yet. Which dorms offered what would be done to minimize the number of people who would think about moving out – basically what the majority of people in a certain dorm choose. At that point you could choose a) to stay in EC and forget your dining b) from among the dorms that offer that dining level, if there was uneven demand for dorms at the same dining level there would be a lottery. Also you could also choose, another dorm with a different dining level than what you originally chose, and not your current dorm, but at a lower priority to keep room for the people who preferred that dorm’s dining level in the first place.

If ~300 students in dorms that currently have dining choose no dining, then one of the dining halls would be closed (again which one that is, would minimize campus wide moves) , thus creating more capacity in no dining dorms. This is much better than HDAG’s plan which just assigned 1,200 beds at hot breakfast and dinner.

Under any plan with the opening of Massey Hall (W1) crowding will disappear, I’m pretty sure.

You question about not fitting in with the others would have the same likelihood of happening today and it would be handled the same way as today.

Remember the algorithm is desired to minimize the number of people who feel that they would want to move to get the dining level of their choice. So it is unlikely you will end up with all new people in your hall, and it will be very unlikely that EC will be flooded with new people seeking to avoid dining. Plus I believe that my plan will actually reduce the number of people trying to flood into EC because it is cheap vs HDAG – since under my plan if enough people don’t want a dining plan and thus dining hall, one of the dining hall dorms could be shut down, thus creating additional non dining capacity.

Maximize Happiness Instead

A problem with Plaz's algorithm:

Suppose you have 3 dorms, A, B & C. Each dorm has 300 students. In each dorm, 100 students want a dining plan, and 200 students do not want a dining plan. Every student who wants a dining plan wouldn't want to move out of his/her dorm if his/her dorm didn't get a dining plan. Every student who doesn't want a dining plan would want to move out of his/her dorm if his/her dorm did get a dining plan.

Under Plaz's plan, an arbitrary dorm (let's say dorm A) would get a dining plan. However, the 200 students of dorm A who didn't want the plan would be unable to move out (even though they want to), making them very unhappy. The 100 students of dorm B and the 100 students of dorm C would all be semi-unhappy because they didn't get the dining plan they got, but they don't want to move out. In summary, you get 200 very unhappy students and 100 semi-unhappy students.

However, if instead you let all 3 dorms not have dining plans, then you'd get 300 semi-unhappy students, which would be a far better scenario than having 200 very unhappy students and 100 semi-unhappy students.

My point is that Plaz's plan can fail massively for maximizing happiness if chunks of people decide not to move out of their dorms. A better algorithm would ask people under what circumstances they would want to move out of their dorm, and then maximize student happiness in some manner. (All algorithms proposed are better than HDAG's solution.)

Well I think it all depends on the level of unhappiness. You break it down into 3 levels: -Happiness - I got the dorm and dining I wanted -Semi-Unhappiness – I have the dorm I want, but not the dining plan -Very Unhappiness – I am forced to pay for a dining plan I do not want

But how will you be determining the levels of happiness relative to each other so that you can trade them off? Also with my assertion that everyone’s dorm preference function is distinct and independent of everyone else’s how will you determine how to maximize happiness. You could attempt to come up with an algorithm to at least measure the average relative unhappiness between the different options. My plan focuses on only one type of unhappiness: not getting your desired dining level. If we were able to agree on the relative weights of happiness, your plan would be able to better maximize happiness in some edge cases.

You also simplify the dining plan to 2 levels (dining/no dining) and only assume 3 houses are in play.

My plan also operates under the assumption that at least one dorm needs to offer hot breakfast. It depends what you weight: people not getting less of a dining plan than they wanted or people not paying for something they do not want. You prefer that students get less dining rather than more. Also you assume that the administration would not run breakfast even if only 50 people campus wide want it. I have a feeling that they will insist on offering breakfast at least somewhere, even if the demand is ~50 people. They may decide this due to “obligations” or admissions reasons.

You could also run the modifications some have suggested through your scenario. Take for example the plan proposed by David where people get the dining level they select, no matter where they live (a move that protects revenue) AND students can continue to live in their current dorm at their selected dining plan. “David Balanced Plan”. Under the scenario you proposed, a random dorm (since preference was the same in each), say Dorm A would offer the dining plan. Again 400 people in Dorms B and C would be happy at no dining plan. The 100 students who wanted a dining plan in A would be happy to have one in A. The 200 students from Dorm B and C who wanted a dining plan would be signed up for one. These people would be a Semi-Happy (a new happiness level) in that they get the dining they want, but may need to walk next door. (They could likely not move into A, since no one would want to move out of A) The 200 students who did not want a dining plan in A would just ignore the dining hall in A. Final total: 700 happy people; 200 semi happy people; Dining hall would have 300 customers so it would be happy and maximize its revenue in line with the HDAG plan.

If you wanted to be a bit more student friendly you could take only the student friendly half of David’s plan. “David’s Plan Unbalanced” Students would not have to sign up for the dining plans they chose if it meant walking next door. However, students who did not want the dorm’s plan could still stay in their dorm. Under this plan everyone would be happy, except the dining hall would be money losing with only 100 customers. MIT could choose to eat that loss until the next generation of students come, who would be subject to that dorm’s dining plan.

Although I would agree with you that algorithms proposed would maximize student happiness more than HDAG’s plan, even if no one decides to move anywhere. If students decide to move because of their own personal preference function, happiness would increase under each plan (even David’s Unbalanced Plan would become more financially sustainable).


Dining at MIT

Various things I have said about dining at MIT, in various contexts: