The Psychology of Doing the Hard Thing First

There is a reasonable case for doing difficult work early in the day, and almost none of it comes from the theory people cite when they make it. The advice to "eat the frog" is normally justified by a model of willpower as a depleting resource. That model failed a preregistered replication across 23 laboratories. What survives is a smaller and more individual claim: do the hard thing at your peak, which for a substantial minority of people is not first.

Where the advice comes from

The frog line is nearly always attributed to Mark Twain. There is no record of him writing it. The saying circulated in the nineteenth century and was popularized in its current shape by Brian Tracy's 2001 book of that title. A small point, but a useful indicator: advice whose most-repeated credential is a fabricated attribution has usually not been examined closely.

The justification that collapsed

The standard mechanism offered for front-loading is ego depletion, the claim that self-control draws on a limited resource that empties with use. If that were true, the ordering argument follows directly: spend the full tank on the hardest thing and let easy work absorb the depleted hours.

In 2016, Martin Hagger, Nikos Chatzisarantis and a large collaboration published a Registered Replication Report in Perspectives on Psychological Science (volume 11, issue 4, pages 546 to 573). Twenty-three laboratories ran a standardized, preregistered depletion protocol on a combined 2,141 participants. The pooled effect was d = 0.04, with a 95 percent confidence interval of negative 0.07 to 0.15. The interval contains zero. The effect the protocol was built to detect had been reported at d = 0.69.

That is not a small effect. It is an absent one, and every subsequent use of ego depletion as a justification for anything should be stated alongside it.

Decision fatigue and the judges

The related popular claim is decision fatigue, and its evidence is almost always a single study. Shai Danziger, Jonathan Levav and Liora Avnaim-Pesso analyzed 1,112 parole rulings by eight Israeli judges over roughly ten months, published in PNAS in 2011. The proportion of favorable rulings began each session near 65 percent and fell to close to zero by the end of it, then reset to 65 percent after a food break. The authors speculated cautiously about mental depletion. The internet did not speculate cautiously about anything.

Two problems have since been documented. Keren Weinshall-Margel and John Shapard, writing in the same journal, reported from interviews with panel personnel that case ordering was not random: prisoners with attorneys were typically heard first, and boards worked through all cases from one prison before breaking. That is a confound sitting exactly where the effect is.

Andreas Gloeckner then showed in Judgment and Decision Making in 2016 (volume 11, issue 6, pages 601 to 610) that a comparable within-session decline can be generated as a statistical artifact, because favorable rulings take longer to process than unfavorable ones. His conclusion was not that the effect is fictional but that its magnitude has been substantially overestimated.

So the two pillars under "hard thing first" are a replication failure and a confounded observational study. That does not make the advice wrong. It makes the usual argument for it worthless.

What does support front-loading

Three things do, and none of them are about willpower.

First, cognition genuinely varies across the day. Christina Schmidt, Fabienne Collette, Christian Cajochen and Philippe Peigneux reviewed this in Cognitive Neuropsychology in 2007 (volume 24, issue 7, pages 755 to 789). Time-of-day modulations affect attention, executive function and memory, driven by the interaction of circadian phase and accumulated sleep pressure. This is chronobiology, and it is well established.

Second, there is population-scale evidence that later is worse for demanding cognitive work. Hans Henrik Sievertsen, Francesca Gino and Marco Piovesan analyzed every standardized test taken in Danish public schools across four school years, published in PNAS in 2016 (volume 113, issue 10, pages 2621 to 2624). Each hour later in the day cost 0.9 percent of a standard deviation in test scores. Worth equal attention: a break of 20 to 30 minutes recovered 1.7 percent of a standard deviation, with a confidence interval of 1.2 to 2.2 percent. That is administrative data on millions of tests, which is a different evidentiary class from a lab paradigm.

Notice what the break result implies. If a short pause buys back roughly two hours' worth of decline, whatever is accumulating is cheap to reverse, which is not how a depleting resource is supposed to behave.

Third, and least psychological, reactive demand accumulates. By mid-afternoon other people's requests have arrived. That is a fact about calendars rather than brains, and it may be the strongest argument for front-loading anything.

The part that breaks the rule

Elise Facer-Childs and Roland Brandstaetter tested when peak performance actually occurs, published in Current Biology in 2015 (volume 25, issue 4, pages 518 to 522). They found that the best predictor of an individual's peak was not clock time but time since entrained awakening. Early chronotypes peaked roughly six hours after waking. Late chronotypes peaked roughly eleven hours after waking. Individual performance varied by as much as 26 percent across a single day.

Eleven hours after waking, for someone who gets up at eight, is seven in the evening. The instruction to do hard work first thing is a lark's schedule generalized into a universal rule, and for a late chronotype it prescribes exactly the trough.

Chronotype is also not a two-category thing. Till Roenneberg's Munich Chronotype Questionnaire measures it continuously, from the midpoint of sleep on free days corrected for accumulated sleep debt, and the distribution is a spread with most people between the extremes. His related term social jetlag, the gap between your biological schedule and your imposed one, names what happens when a population runs on lark time regardless.

What to do instead, ranked

1. Find your own peak before optimizing order. Two weeks of logging when demanding work went well is more informative than any general rule, and it is the only step here that is specific to you.
2. Protect the peak, whenever it is. Ordering matters far less than whether the best hours are available at all. Most people spend theirs in meetings and then wonder about sequencing.
3. Front-load for the calendar reason. Early hours are defensible because nobody has claimed them yet. That argument survives every replication problem above, because it is not a psychological claim.
4. Take the break. A 20 to 30 minute pause bought back 1.7 percent of a standard deviation in Danish test scores. Almost nothing else in this literature is that cheap.
5. Stop justifying anything with ego depletion. After Hagger and colleagues, a plan that requires the depletion model to be true is a plan without support.
6. Do not confuse hard with important. "Eat the frog" selects for aversiveness. Aversiveness and importance are different variables, and optimizing the first at the expense of the second is a common and expensive error.

The honest conclusion

The defensible version of this advice is: do the hard thing at your peak, protect that window structurally, and front-load only because mornings are less contested. That is duller than the frog, and it has the advantage of surviving contact with the evidence.

It also relocates the question. Once the willpower story is gone, what remains is whether the difficult work is scheduled at all, which is a matter of what actually sustains effort over time rather than of daily ordering. Morning routines get their attention not because mornings are cognitively special for everyone but because they are the most defensible block on most calendars, which is the real case for building a sequence you do not have to decide about. Waiting to feel ready remains the least supported option here, since the evidence on whether action produces motivation or the reverse runs in the direction people find least convenient.