Number series: find the rule before you touch the arithmetic

Number series is four to six questions in a typical bank prelims shift, and the same family of questions appears in the Mathematics section of RRB NTPC. In every case they are among the cheapest marks available, and in every case candidates lose them the same way: they start doing arithmetic on the differences before they have decided what kind of series it is.

The exam is not testing whether you can subtract. It is testing whether you can look at five or six numbers and recognise the shape of the rule that produced them. That recognition is a trained reflex, and this is how to train it.

The first look decides everything

Before you compute a single difference, read the whole sequence once and answer two questions.

How fast is it growing? If the numbers rise gently and evenly, the rule is almost certainly addition or a small constant difference. If they roughly double or triple each time, think multiplication or ratio. If they explode, think squares, cubes or a multiplication whose factor is itself changing. This single judgement eliminates most of the possibilities before you have done any work.

Does it move in one direction? A series that rises, falls and rises again is usually two interleaved series, or a pattern built on alternating operations. Do not try to force a single rule onto a sequence that visibly changes direction.

Those two observations narrow six or seven candidate rules down to one or two, and the rest is confirmation rather than search.

The rules worth knowing on sight

There are only a handful of patterns that the setters actually use, and they recur far more than any bank of “tricks” would suggest.

Constant difference. Each term differs from the last by the same amount. The easiest series and the one most likely to be a distractor for a harder one.

Changing difference. The differences themselves form a series: +2, +4, +6, +8, or +3, +5, +7. When a first look at the gaps shows them growing steadily, take the difference of the differences before anything else.

Multiplication or ratio. Each term is the previous one times a fixed number, or times a number that itself changes: ×2, ×3, ×4, or ×1.5 throughout. Rapid growth is the signal.

Squares and cubes, offset. Sequences built on 1, 4, 9, 16, 25 or 1, 8, 27, 64, often with a small constant added or subtracted. If a number is suspiciously close to a perfect square or cube, test that first. This is why the squares and cubes worth memorising earn their place: you cannot recognise a squares pattern you cannot see.

Alternating series. Two independent series woven together, one on the odd positions and one on the even. The give-away is a sequence that seems to have no single consistent rule. Split it into first, third, fifth terms and second, fourth, sixth and each half usually falls out immediately.

Mixed operation. A repeating cycle such as ×2 then +3, then ×2 then +3. Less common, but it is the standard second-hardest question in a set, so expect one.

A routine you can run under the clock

1. Read the whole sequence once and judge growth and direction.
2. If growth is gentle, write the differences under the gaps. If they are constant, you are done. If they themselves form a pattern, you are done.
3. If growth is rapid, test ratios between consecutive terms.
4. If the numbers sit near perfect squares or cubes, test that.
5. If nothing is consistent across the whole sequence, split into alternate terms and treat as two series.
6. If none of these resolves it within a minute, leave it. There is another cheap question waiting.

Run in that order, the routine solves the overwhelming majority of series questions in well under a minute, because it tests the most common rules first.

Wrong-number series: confirm before you accuse

A wrong-number set gives a complete sequence with one term that breaks the rule, and asks which term is wrong. The mistake is to spot an odd-looking number and name it without checking. The correct method reverses the order:

• Establish the rule from the terms you are confident about, usually the first two or three gaps.
• Project what each subsequent term should be under that rule.
• The first term that does not match your projection is the wrong one.

The trap is that a single wrong term makes the next gap look wrong too, so two consecutive terms will appear suspect. Only the first one is actually wrong; the second looks wrong only because it is measured from a corrupted neighbour. Always name the earlier of the two.

Where the marks leak

Forcing one rule onto an alternating series. More lost series come from this than any arithmetic slip. If a single rule will not fit, split before you struggle.

Ignoring the size of the jump. Candidates write out differences on a series that clearly doubles, wasting twenty seconds proving it is not addition. Let the growth rate choose your first test.

Chasing the hard one. In a five-question set there is usually one series built to burn time. Recognise when you are past a minute with no rule in sight, and leave it. The set is not worth failing over one term.

Arithmetic errors on an easy rule. Once the rule is found the sum is trivial, which is exactly when people get careless. Read the position asked for; a “which term comes next” and a “what replaces the question mark in the middle” are answered differently.

Practice that transfers

Recognition improves with volume and near-instant feedback, not with long single questions. Do a block of ten series in one sitting, and for each one write down the rule in words before you compute the answer, then check whether your named rule was right. You are training the first-look judgement, and the only way to train it is to make the judgement explicitly, again and again.

Work through the number series practice set, which is built from missing-term and wrong-term questions across differences, products, squares and alternating patterns, and keep the arithmetic itself fast by drilling calculation speed separately.