TOLD, the Politecnico di Milano Design test: structure, weighted score and how to prepare

The TOLD (Test del Design) is the test the Politecnico di Milano uses to admit students to the degrees of its School of Design. It differs from the CISIA tests in one respect: the score is not the sum of your answers but a weighted sum, with sections that count four and a half times more than others and with minimum numbers of correct answers to reach section by section. Anyone who does not know the formula prepares the wrong test: hours go into English, which weighs little, while geometry, which weighs a lot and is the first tie-breaker, is neglected.

Everything below comes from the 2026/27 call for the Design degree of the Politecnico (official PDF), read on 29 October 2026. The 2027/28 call is not published yet: where a figure may change I say so, and where I have no source I write it.

Structure: six areas, two versions

The test is taken in a computer room, on a computer, with five-option questions and one correct answer. The 2026/27 call has two versions:

The English version, reserved for the English-taught degree (Interaction Design), has 60 questions in 70 minutes: the same five areas, without the English-language section.

Look at questions against time: history of design and art has 12 questions in 10 minutes (under a minute each), general knowledge 24 in 20, geometry 12 in 20. There is no uniform time per question: a knowledge section is answered in seconds if you know, and the time saved there is the time you need for geometry.

What is inside each area

The call describes the content quite precisely. In short:

• Geometry and representation. Foundations of analytic and Euclidean geometry, abstract problems solved by logical reasoning, reading graphs, drawings and iconic representations, elementary notions of visual representation.
• Logic. Abstract verbal and symbolic reasoning: sets, hypothesis and conclusion, necessary and sufficient conditions, reasoning errors.
• Verbal comprehension. Essay, fiction, newspaper and magazine texts: command of the language and consistency with the premises of the text.
• History of design and of art. Events and protagonists by period, chronological orientation, the evolution of the arts from antiquity to today, the history of industrial design (people, products, events).
• General knowledge. Common upper-secondary knowledge and historical, technological, scientific, political and social contemporary affairs.
• English. Comprehension of everyday-language texts, such as descriptions of events, feelings and wishes.

What I have not verified: I have read no public TOLD paper (neither an official one nor an archive), so I cannot tell you how hard a typical question is. From the threshold and the minimums it looks like the entry filter is low and the ranking is competitive, but that is my reading, not a figure from the call.

Scoring: +1, 0, −0.25, then the weights

Each question is worth +1 if correct, 0 if left blank, −0.25 if wrong. With five options a blind guess is worth zero on average: $\frac{1}{5}\cdot 1-\frac{4}{5}\cdot 0.25=0$. The expected-value calculation when you can discard options is in the article on the English-PSY: the penalty is the same here.

Then comes the surprising part. The score of each section is computed, and the total weighted score is

where $P_g$, $P_l$, $P_v$, $P_s$, $P_c$ and $P_i$ are the scores (with penalties) of geometry, logic, verbal comprehension, history, general knowledge and English. The five non-language sections weigh 1.5, English one third.

Check on the maximum: the five areas have $12+6+6+12+24=60$ questions, so $1.5\cdot 60=90$; English has 30, so $30/3=10$. Total 100, the maximum of the Italian version. In the English version, which has no English section, the maximum is 90.

How much one answer is worth

A correct answer in a non-language section is worth 1.5 points, in English $1/3$: the ratio is

So one correct geometry answer is worth four and a half correct English answers. The error weighs more too: a wrong answer outside English costs $0.25\cdot 1.5=0.375$ points.

A worked calculation, step by step

Take an imaginary candidate (not a real paper) who answers as follows:

The five non-language sections add up to $6.25+3.75+4.75+5+12.5=32.25$. Apply the formula:

Now the conditions in the call:

• Minimum correct answers per section (Geometry 2, Logic 1, Verbal comprehension 1, History 2, General knowledge 5, no minimum in English): 7, 4, 5, 6 and 14 correct, all cleared.
• Threshold on the weighted total: 30/100 for EU applicants. $56.04\ge 30$: cleared.
• English requirement: at least 24 correct answers are needed, and here there are exactly 24: satisfied, by a single question.

The example shows two things. First, the 30 threshold is low: the candidate clears it comfortably while getting 19 questions out of 90 wrong and leaving 11 blank. Second, the ranking is decided elsewhere: turning just three geometry answers from wrong to correct lifts the total to $56.04+3\cdot (1.5+0.375)\approx 61.7$, while three more English questions are worth only $3\cdot (1+0.25)/3=1.25$ points.

Minimums and tie-breakers

Two rules in the call change your strategy:

1. Section minimums. The total is not enough: the call sets a minimum number of correct answers per section, to be respected on top of the threshold on the total (the call states that missing a minimum in even one section excludes the candidate from the ranking). The minimums are low (2 in geometry, 1 in logic, 1 in comprehension, 2 in history, 5 in general knowledge), so the risk is not the section you cannot do but the one you leave blank for fear of the penalty: in general knowledge, with 24 questions and a minimum of 5 correct, leaving almost everything blank out of caution is the easiest way not to get there.
2. Order of tie-breakers. On equal scores the following count in order: Geometry, Verbal comprehension, History of design and art, Logic, General knowledge, English. Geometry opens the list and has a high weight: it is the section worth investing in most.

Two side rules: the Italian TOLD with at least 30/100 counts as the proof of Italian for those applying to the English-taught degree, and thresholds change for the reserved quota (20/100 in Italian; 18/90 in English).

Windows: what I know and what I do not

The 2026/27 call had test windows from 1 March to 31 May 2026 for the Italian version (two blocks: 1 March-11 April and 12 April-31 May) and from 14 March to 14 May 2026 for the English one. They are past windows: for 2027/28 no dates are published. When the call comes out we will report it in the tests calendar and at the top of this page. Seats available and the score of the last admitted candidate I have not verified: I do not quote them.

How to prepare, area by area

The 1.5 weight and the tie-breaker order suggest this hierarchy of hours: geometry, then general knowledge (24 questions, 27 % of the test), then history, logic and comprehension, and English last, only if you do not already have a level that gives you 24 out of 30.

Geometry and representation. You do not need a full science-lyceum programme, you need the ability to see. An original example: a right triangle has legs of 6 and 8 cm; what is the radius of the circle through its three vertices? Remember that a right angle is inscribed in a semicircle, so the hypotenuse (10 cm) is a diameter and the radius is 5 cm. The question asks for no formulas: it asks you to connect a property to a figure. Train with a sheet of paper, always drawing before choosing. For coordinates and the equations of lines and circles, the maths page covers the basics.

Logic. Few rules, applied rigorously. An original example: «All the awarded prototypes were tested. Some tested objects are uncomfortable.» What follows? Nothing about the awarded prototypes: the uncomfortable tested objects might not be awarded. The question measures whether you tell «some» from «all» and notice that the link between the two sentences is not guaranteed. For drills in series, the article on logical reasoning covers the same mechanisms.

Verbal comprehension. The right answer is the one the text supports, not the one that is true in the world. Six questions in 10 minutes: a careful reading of a short passage and a check against each option.

History of design and of art. It is the only area that cannot be rebuilt by reasoning: you study it. A chronology in large blocks (from the early twentieth-century avant-gardes to the Ulm school, from postwar Italian design to contemporary design) with the people, products and events the call names as the subject of questions. With 12 questions in 10 minutes you have no time to recall: you must recognise. Short cards and review spaced over days work better than a single reading.

General knowledge. 24 questions are the numerical pillar of the test. The content is broad (history, science, technology, contemporary society): the strategy is reading a newspaper every day and getting used to tying a fact to a date. I have verified no official detailed programme beyond what the call says, so I do not promise you a topic list.

English. Everyday level, 30 questions in 15 minutes. If you are comfortable at B2, a few vocabulary and grammar drills are enough to keep 24 out of 30.

Typical mistakes

1. Spending hours on English. It is worth $1/3$ of the score and has no minimum. It serves the language requirement (24 correct), not the ranking.
2. Leaving a whole section blank for fear of −0.25. With five options the expected value of a blind attempt is zero, and a missed minimum is worth more than any penalty.
3. Spreading time evenly. History has under a minute a question; geometry almost two. Similar times on both is wrong for one of them.
4. Trusting figures from an old call. Weights, minimums and thresholds are those of 2026/27. When the next call is out, compare everything again.

A plan in four steps

1. Week 1: read the whole call and compute by hand your score on a hypothetical paper, as in the example above. Understanding the formula changes your priorities.
2. Weeks 2-6: geometry and design history daily; logic and comprehension twice a week; general knowledge every day in small doses.
3. Weeks 7-8: full papers over 85 minutes, with the six sections in sequence and the call’s timings.
4. Last days: review of the chronologies, no new topics.

To try all this under test conditions there is the TOLD practice test: six timed sections, a weighted score out of 100 and per-section minimums as in the call (structure reconstructed from the 2026/27 call, original questions, no real paper read; the questions are in Italian). If you want a tailored plan for design, the free introductory call is for that: in half an hour we see where you start and how to split hours between geometry, logic and general knowledge. The tests I prepare are listed on the admission tests page.