or
Why We Keep Guessing at the Wrapped Present
The future is the next moment. We live in a permanent transition between the past and what comes next, and the thin strip we call the present is mostly a polite fiction. Yet we spend enormous energy trying to predict what lies on the other side of it — and we are, with remarkable consistency, wrong.
The easy prediction is: predicting that our predictions will fail.
The harder questions are the ones we rarely stop to ask. Why do we predict in
the first place? What exactly are we predicting? And would our behaviour be any
different because of our predictions?
I cannot answer these questions definitively. But I think they are
worth unpacking, because how we answer them determines whether political
prediction is a useful discipline or an elaborate form of self-reassurance, or
self-destruction.
Not all predictions are alike. At the quietest end of the spectrum
sit predictions that neither affect our behaviour nor are affected by it.
Scientists tell us that the Andromeda galaxy will collide with our own in
roughly two billion years. This is an educated estimate, grounded in precise
mathematics and physics — and still, technically, a prediction, since a great
deal could happen between now and then. But we will not reorganise our foreign
policy around it. No government will call an emergency session. No pundit will
appear on television to analyse the implications for the next election cycle.
It is the one category of prediction that is genuinely free of us, and perhaps
for that reason, the one we find easiest to accept.
A step closer to home are predictions that will not be changed by
our behaviour, even if they change us. The weather is the classic example. We
do not alter tomorrow's rain by predicting it; we only alter how we dress. The
forecast does not make itself true or false — it simply describes what is
coming, and we adjust accordingly. This category has limited but real
application in politics: there are situations where the actions of one part of
the world are simply too distant, too disconnected, to bend the outcome in
another.
· · ·
The most interesting category, and the one that matters most in
political life, is where prediction and behaviour fold back on each other.
These are predictions that change the very thing they are predicting.
A teacher who tells a student they are likely to fail may be
issuing a warning that becomes a wake-up call — the student works harder,
performs better, and the prediction is proved wrong by the act of having been
made. Or the same warning lands as a verdict, and the student stops trying, and
the prediction fulfils itself with terrible efficiency. The prediction is not a
passive observation; it is an intervention.
The mechanism is familiar from markets. When speculation about a
potential war affecting oil supply begins to circulate seriously, traders do
not wait for the shortage to materialise. They act on the prediction. They buy
in advance. Prices rise. And so the prediction — higher oil prices — comes true
not because of any actual disruption, but because enough people believed it
would. The shortage was conjured into existence by the act of forecasting it.
British electoral politics offers a particularly clear version of
this loop. For decades, polls consistently predicted that one of the two main
parties will win any given constituency, which leads many voters to conclude
that a vote for the Greens or the Liberal Democrats is a wasted vote, which
leads them either to vote tactically or not to vote at all, which then confirms
the poll's prediction. The forecast creates the conditions for its own
accuracy. But that loop is not unbreakable. In 2024 it cracked — the Lib Dems
surged, the Greens returned MPs, and Reform demonstrated that new predictions
can generate new realities just as powerfully as old ones can suppress them.
What is shifting now is not just the voting patterns but the underlying belief:
once enough people stop treating a prediction as a fact, the prediction loses
its grip. Elections may be the most honest laboratory we have for watching that
process happen in real time.
So why do we predict at all, given that we know this is how it
works?
Part of the answer is simple: impatience. We are creatures who
cannot wait. Hand someone a wrapped present and watch as they before have
touched the ribbon, they are already guessing. We watch a political situation
unfold in real time and immediately reach for interpretation, projection,
resolution. We spend hours listening to commentators analyse an evolving
crisis, speculating about what will happen — and when the expected thing
happens, we treat it as confirmation of our analytical powers, without pausing
to ask whether the expectation itself shaped the outcome.
But we also predict because we are afraid of being surprised. The
guess at the wrapped present is not only eagerness — it is armour. If I can
name what is inside before I open it, the opening cannot shock me. And this
fear-driven prediction is just as capable of making itself real as the
optimistic kind: the candidate who believes they will lose and stops fighting,
the movement that convinces itself change is impossible and stops pushing. The
anxiety becomes the outcome.
Which brings us to the most honest thing one can say about why our
predictions so often fail. We do not predict the future neutrally. We predict
through a lens ground by the past — our experiences, our fears, our desires —
and that lens bends everything. The optimist sees the conditions that favour
the outcome they want; the pessimist sees only the obstacles. We say we are
being analytical. We are mostly being autobiographical. And because that bias
shapes which predictions we act on, it shapes which self-fulfilling loops we
set in motion. The bias and the behaviour-change are not separate causes of
error. One feeds the other.
When I flip a coin, I know I have a fifty-fifty chance of heads or
tails. I can predict with confidence because the system is closed and the
possibilities are finite. But most political coins have an infinite number of
sides. Some of those sides only come into existence depending on what I do
before the flip. Others emerge only from what I do after it. In a system like
that — which is to say, in any system involving human beings in large numbers —
prediction is not really prediction at all. It is a declaration of intent, or a
confession of fear, dressed up as analysis.
We are also seduced by pattern. History does not repeat itself —
but it rhymes, and we love the rhyme. The danger is not that we notice it. The
danger is that we approach every new situation already looking for where it
rhymes, and in doing so we miss what is genuinely new about it. The notes we
cannot place are usually the ones that matter.
Ahmad baker
30.03.2026