1,254 Seats and Nearly 14,000 Students: How Many Library Seats Does Monash Malaysia Need?

At two in the afternoon during semester, I often find myself carrying a laptop around the MUM library, looking for a seat. Before I have written a line of my assignment, I have toured all three floors.

On Level 1, conversations fill the room. Upstairs on Level 2, it is quieter, but there is still nowhere to sit. Level 3 offers another round of lecture notes, closed MacBooks, leftover drinks and charging cables plugged into power strips. The belongings are there. Their owners are not, and there is no telling when they will return.

Someone finally starts packing up. Before I can get across the room, another student has taken the seat. After twenty minutes, I leave with my laptop. Either I go back to my room or head to Sunway Square and pay RM18 for a coffee that comes with a power outlet.

After enough afternoons like this, it gets irritating. Seat hogging is annoying. Leaving for lunch or class while your backpack continues occupying a desk is annoying. But the thing I actually want to complain about is simpler: if every unattended bag vanished, would the university have provided enough chairs in the first place?

So I took the university’s own published numbers and did the arithmetic. Every assumption is written down before the conclusion. A banner telling students to “study considerately” does not get to replace a capacity calculation.

Start with the University’s Own Yardstick

In a 2017 anti-hogging campaign, the library reported 1,174 seats, or one seat for every 6.6 students. The announcement compared this with roughly one seat per seven students at Clayton’s Matheson Library and treated the two as broadly comparable [1][1] M. U. M. Library and L. Commons, “LLC Campaign: Park Yourself, Not Your Bags,” 2017. Historical seating count and comparison ratio; campaign began May 29, 2017. Accessed September 16, 2026. https://www.monash.edu.my/library/about/glance/director-spotlight/2017/park-yourself-not-your-bags. Read backwards, that ratio implies a student population of roughly 7,750 in 2017.

Fine. Let us use the same yardstick today.

After the Level 3 refurbishment in 2018, the published total was 1,254 seats [2][2] M. U. M. Library and L. Commons, “Ready, Set, Go! Level 3 New Refurbished Area Now Open,” 2018. Reports a total seating capacity of 1254 after the Level 3 refurbishment. Accessed September 16, 2026. https://www.monash.edu.my/library/about/glance/director-spotlight/2018/level-3-new-refurbished-area-open. That is the last complete official seat count I can find. Level 2 was refurbished in 2025, but the notice did not publish a new library-wide total [3][3] M. U. M. Library and L. Commons, “Level 2 Renovation,” 2025. Posted July 3, 2025. Announces closure from June 30 to the end of October, without a before-and-after seat count. Accessed September 16, 2026. https://www.monash.edu.my/library/about/news/2025/articles/level-2-renovation. Until the university publishes a current number, 1,254 remains the only official baseline available in its public material.

As of 16 September 2026, the university’s Welcome page reports approximately 13,800 students [4][4] M. U. Malaysia, “Welcome to Monash Malaysia,” . Page reports approximately 13,800 students. Accessed September 16, 2026; not a count of simultaneous Sunway campus attendance. https://www.monash.edu.my/about/welcome. Divide one by the other:

$$\frac{13{,}800}{1254}\approx 11.0.$$

A seat that corresponded to 6.6 students in the university’s old comparison now corresponds to about 11. That is roughly 1.67 times as many students per published seat, and the implied population has grown by close to 80 percent since 2017.

Turn the comparison around. To maintain the old 1:6.6 level for a population $N$, the reference capacity is:

$$C_{\mathrm{ref}}(N)=\left\lceil\frac{N}{6.6}\right\rceil.$$

Not all 13,800 are on the Sunway campus in any given semester. Some are on exchange, some on industry placement, some intermitting, some taking fully online units. Discount for all of that aggressively, by about one sixth, and work with 11,500 students. The result is still:

Population scenario	Seats at 1:6.6	Above the 1,254 reported in 2018
13,800	2,091	837
11,500	1,743	489

By the university’s own old comparison, today’s population corresponds to a provision level roughly 490 to 840 seats above the 2018 published count.

If the current library has already added hundreds of seats, excellent. Publish the number. That is exactly the problem here: enrollment has publicly grown to nearly fourteen thousand, while students are left guessing how much study-space capacity actually grew with it.

Give the University Every Seat

In real life, a chair is not automatically useful. Level 1 is built for interaction and collaboration. Level 2 is the Quiet Zone. Level 3 mixes individual study, quiet areas and group discussion spaces [5][5] M. U. M. Library and L. Commons, “Library Spaces,” . Official floor functions and space descriptions. Accessed September 16, 2026. https://www.monash.edu.my/library/services-facilities/library-spaces. A free chair in a discussion area is not equivalent to a quiet desk when you need to code. An empty table in the Quiet Zone is not a group meeting room. A laptop with five percent battery also cares whether the nearest outlet works.

For the calculations below, I ignore all of that.

All 1,254 seats count as perfectly usable. No broken outlets, no task mismatch, no awkward communal tables, and no unattended bags. This is already the most generous possible version of the library.

That makes the rest of the argument cleaner. If capacity is inadequate even when every physical seat is assumed to be available, blaming students for using those seats badly cannot solve the underlying problem.

How Large Is Peak Demand?

Take one moment between two and four on a weekday afternoon during semester. Let $X$ be the number of students who want a library seat at that moment, including those already seated and those still searching.

For any student, define:

$$\alpha=\mathbb P(\text{on campus now}), \qquad \beta=\mathbb P(\text{needs a library seat now}\mid\text{on campus now}).$$

Then:

$$p=\alpha\beta.$$

Use two peak-demand scenarios:

$$\begin{aligned} p_{\text{low}}&=0.55\times0.20=11\%,\\ p_{\text{high}}&=0.60\times0.25=15\%. \end{aligned}$$

Under the usual equal-probability independence approximation:

$$X\sim\operatorname{Binomial}(N,p), \qquad \mathbb E[X]=Np, \qquad \operatorname{Var}(X)=Np(1-p).$$

Start with the expected demand:

Population scenario	Demand rate	Expected concurrent demand	Above 1,254
13,800	11%	1,518	264
13,800	15%	2,070	816
11,500	11%	1,265	11
11,500	15%	1,725	471

Even the softest combination, 11,500 students with an 11 percent demand rate, already puts expected demand above 1,254. More importantly, this is not merely a mean that sits eleven seats over the line. The exact binomial probability is:

$$\mathbb P(X>1254)\approx62.2\% \quad(N=11{,}500,\ p=0.11).$$

So even under that very forgiving scenario, the historical capacity is insufficient more often than not.

Now take 13,800 students and a 15 percent demand rate:

$$\mu=2070, \qquad \sigma^2=1759.5, \qquad \mu-1254=816.$$

Cantelli’s inequality gives:

$$\begin{aligned} \mathbb P(X\le1254) &\le\frac{1759.5}{1759.5+816^2}\\ &\approx0.0026355. \end{aligned}$$

The upper bound is only about 0.264 percent. In that peak scenario, 1,254 seats covering simultaneous demand is essentially a miracle ^{1Cantelli’s left-tail inequality is $\mathbb{P}(X-\mu \le -k)\le {\sigma }^2/({\sigma }^2+k^2)$ for $k>0$. The main calculation uses $k=\mu -1254=816$ and requires no normal approximation.} .

Real students are not independent Bernoulli trials either. Lectures end together, project groups work together, and deadlines arrive together. Those common causes add positive covariance terms to the variance ^{2More generally, if $X={\sum }_i{I}_i$, then $\mathrm{Var}(X)={\sum }_i{p}_i(1-p_i)+2{\sum }_{i<j}\mathrm{Cov}(I_i,I_j)$. Shared schedules and common deadlines make those covariance terms nonzero.} . To be fair, a larger variance loosens the Cantelli bound above rather than tightening it. But it does not move the mean: expected demand still sits 816 seats above capacity. What positive correlation can change is the shape of the peak. Every calculation below still assumes independence. If demand clusters around shared timetables, the busiest moments can be busier than the independent model predicts, and the independent model is already short.

The Question That Matters Is Whether a Student Can Actually Sit Down

A library is not an expectation-value exercise. From a student’s perspective, the useful question is much simpler: if I walk in now, what is the chance I actually find a seat?

Set an extremely modest service target:

During normal teaching-week peak hours, a student who needs the library should have at least an 80 percent chance of immediately finding a seat.

Eighty percent is hardly luxurious. It still permits failure one time in five.

For a focal student who is looking for a seat, let $Y$ be the number of other students simultaneously seeking one. Under the same binomial approximation:

$$Y\sim\operatorname{Binomial}(N-1,p).$$

With $C$ usable seats, the focal student succeeds exactly when $Y\le C-1$. Therefore the 80 percent service requirement is:

$$\mathbb P(Y\le C-1)\ge0.8.$$

The minimum required capacity is:

$$C_{80}=1+F^{-1}_{\operatorname{Binomial}(N-1,p)}(0.8).$$

Evaluating the four scenarios gives:

Population scenario	Demand rate	Chance of finding a seat with 1,254 seats	Seats needed for 80%	Gap above 1,254
11,500	11%	36.8%	1,294	40
13,800	11%	<0.001%	1,550	296
11,500	15%	<0.001%	1,758	504
13,800	15%	<0.001%	2,106	852

This is the part I find hardest to excuse.

Even with the most forgiving scenario, 11,500 students and 11 percent demand, 1,254 seats give a student only about a 36.8 percent chance of finding a place. To raise that to a still-mediocre 80 percent requires 1,294 seats.

Use the university’s own current figure of about 13,800 students while keeping demand at only 11 percent, and the 80 percent service target needs 1,550 seats, which is 296 more than the 2018 published total. At the higher demand rate, the gap becomes 504 to 852 seats.

That is why the real problem is not a handful of people leaving bags on tables. The moment the target changes from “do not exceed capacity on average” to “students should usually be able to sit down when they arrive”, required capacity jumps substantially.

Clearing Bags Reduces Waste. It Does Not Manufacture Chairs.

The July 2025 policy says belongings left unattended for more than 30 minutes may be moved beside the table so another student can use the seat [6][6] M. U. M. Library and L. Commons, “Say No to Seat-Hogging,” 2025. Posted July 29, 2025; effective July 28. Items unattended for more than 30 minutes are moved next to the table. Accessed September 16, 2026. https://www.monash.edu.my/library/about/news/2025/articles/say-no-to-seat-hogging. A 2017 campaign already used red reminder slips for the same 30-minute threshold [1][1] M. U. M. Library and L. Commons, “LLC Campaign: Park Yourself, Not Your Bags,” 2017. Historical seating count and comparison ratio; campaign began May 29, 2017. Accessed September 16, 2026. https://www.monash.edu.my/library/about/glance/director-spotlight/2017/park-yourself-not-your-bags.

Fine. Clear the bags. Queueing theory still tells us exactly what kind of problem this can solve. It changes how long an existing seat remains claimed. It does not change how many physical seats exist.

Little’s Law says that in a stable system the average number of occupied positions equals the arrival rate times the average time each arrival stays. With arrival rate $\lambda$ and sojourn time $T$:

$$\bar L=\lambda\,\mathbb E[T].$$

Apply the same identity to one kind of occupation only: a seat held by belongings while its owner is away. Let such absences begin at rate $\nu$ and last $B$ minutes. Without clearing, the average number of seats held this way is:

$$\bar L_{\text{away}}=\nu\,\mathbb E[B].$$

With perfect release exactly 30 minutes after departure, unattended holding time becomes $\min(B,30)$, so the recovered time per absence is:

$$(B-30)_+.$$

Now include a manual patrol. Suppose staff complete one sweep every 30 minutes and departures are uniform relative to that cycle. Time until first detection is:

$$U\sim\operatorname{Uniform}(0,30).$$

If the seat is cleared on the next sweep after the 30-minute threshold, potential clearance time is:

$$D=U+30 \sim\operatorname{Uniform}(30,60), \qquad \mathbb E[D]=45\text{ minutes}.$$

Recovered unattended time becomes $(B-D)_+$, and necessarily:

$$\mathbb E[(B-D)_+] \le \mathbb E[(B-30)_+].$$

Manual patrols can therefore recover at most what an ideal, zero-latency 30-minute rule would recover, and in practice less.

But the most important line is simpler than any of these equations: after every unattended bag has been removed, there are still only 1,254 physical seats in the historical baseline.

The 80 percent service table above already assumes exactly that perfect world. Every one of the 1,254 seats is always available to a person. No hogging at all. Yet the softest scenario is still 40 seats short, the 13,800-student low-demand scenario is 296 short, and the higher-demand scenarios are short by 504 to 852 seats.

Anti-hogging enforcement can recover wasted use within 1,254. It cannot push physical capacity past 1,254, and it certainly cannot conjure 1,550, 1,758 or 2,106 seats into existence.

Turning a capacity problem into a lecture about student behaviour may be administratively convenient. The chairs do not reproduce because someone left a red reminder slip on a desk.

The University Gets the Growth. Students Should Not Get All the Congestion.

This is the part that actually makes me angry.

Enrollment has climbed to about 13,800 students. More students mean more tuition revenue and a larger institution. But the people being admitted are not numbers on a webpage. They need somewhere to attend class, study, charge laptops, meet project teammates and finish assignments before a deadline.

There is nothing sacred about refusing to expand. The university can admit more students. But every additional student carries a physical-space cost. That cost does not disappear because old furniture was refurbished, a room was relabelled as a Quiet Zone, or another anti-hogging notice went up.

If the institution keeps the revenue and scale that come with enrollment growth while students absorb the congestion, the search time, the lack of meeting space, the shortage of outlets and the café bill when they give up, then the cost of expansion has simply been pushed onto students.

What makes this especially frustrating is how generous the arithmetic above already is. I did not discount noisy seats. I did not remove seats with bad power access. I did not subtract a single chair for hogging. I assumed every one of the 1,254 seats was perfectly usable. Even then, an 80 percent chance of finding a seat at peak time still requires substantially more capacity.

At that point, focusing the discussion on whether students returned to their bags quickly enough feels like building a pool too small for the crowd and then studying whether everyone can shower five minutes faster.

The university does not need a mysterious solution. Publish the current seat count. Publish the net seats added by each refurbishment. Publish the capacities of quiet, collaborative and temporary study areas. Plan around a peak service level rather than a comfortable average. A normal target should be that most students who walk into the library during a teaching week can sit down without touring every floor first.

The university’s 2025 announcement puts the future TRX campus on a timeline beginning in 2032 [7][7] M. University, “Monash University's Future Campus in Malaysia,” 2025. Published October 28, 2025; describes a future Kuala Lumpur campus from 2032. Accessed September 16, 2026. https://www.monash.edu/news/articles/monash-universitys-future-campus-in-malaysia4. A beautiful campus in 2032 does not create a desk six years early for a student carrying a laptop around Sunway today.

If enrollment is going to keep growing, study space has to be part of the cost of that growth. Keep admitting students if you want, but chairs, outlets and floor space have to grow with them. Do not collect the upside of expansion, leave the congestion to students, and call a red “no seat hogging” slip the solution.

References

1. [1] M. U. M. Library and L. Commons, “LLC Campaign: Park Yourself, Not Your Bags,” 2017. Historical seating count and comparison ratio; campaign began May 29, 2017. Accessed September 16, 2026. https://www.monash.edu.my/library/about/glance/director-spotlight/2017/park-yourself-not-your-bags a b
2. [2] M. U. M. Library and L. Commons, “Ready, Set, Go! Level 3 New Refurbished Area Now Open,” 2018. Reports a total seating capacity of 1254 after the Level 3 refurbishment. Accessed September 16, 2026. https://www.monash.edu.my/library/about/glance/director-spotlight/2018/level-3-new-refurbished-area-open ↩
3. [3] M. U. M. Library and L. Commons, “Level 2 Renovation,” 2025. Posted July 3, 2025. Announces closure from June 30 to the end of October, without a before-and-after seat count. Accessed September 16, 2026. https://www.monash.edu.my/library/about/news/2025/articles/level-2-renovation ↩
4. [4] M. U. Malaysia, “Welcome to Monash Malaysia,” . Page reports approximately 13,800 students. Accessed September 16, 2026; not a count of simultaneous Sunway campus attendance. https://www.monash.edu.my/about/welcome ↩
5. [5] M. U. M. Library and L. Commons, “Library Spaces,” . Official floor functions and space descriptions. Accessed September 16, 2026. https://www.monash.edu.my/library/services-facilities/library-spaces ↩
6. [6] M. U. M. Library and L. Commons, “Say No to Seat-Hogging,” 2025. Posted July 29, 2025; effective July 28. Items unattended for more than 30 minutes are moved next to the table. Accessed September 16, 2026. https://www.monash.edu.my/library/about/news/2025/articles/say-no-to-seat-hogging ↩
7. [7] M. University, “Monash University's Future Campus in Malaysia,” 2025. Published October 28, 2025; describes a future Kuala Lumpur campus from 2032. Accessed September 16, 2026. https://www.monash.edu/news/articles/monash-universitys-future-campus-in-malaysia4 ↩

Footnotes

1. Cantelli’s left-tail inequality is $\mathbb P(X-\mu\le-k)\le \sigma^2/(\sigma^2+k^2)$ for $k>0$. The main calculation uses $k=\mu-1254=816$ and requires no normal approximation. ↩

2. More generally, if $X=\sum_i I_i$, then $\operatorname{Var}(X)=\sum_i p_i(1-p_i)+2\sum_{i<j}\operatorname{Cov}(I_i,I_j)$. Shared schedules and common deadlines make those covariance terms nonzero. ↩