You'll Take the Win and Hand Me the Loss — Even With the Receipts on the Table
July 6, 2026 · Dana Cole, Predictably Irrational~6 min read
You know the feeling where a piece of work goes well and some quiet, unbothered part of you thinks: yes, that was mostly me? And then the same piece of work goes sideways a fortnight later, and the same quiet part thinks: well, honestly, if they'd only listened to me on the Tuesday. Same you. Same fortnight. Two very different accountants doing the books. There's a new study out about exactly that little bit of creative accounting, and the uncomfortable part isn't that we do it — we've long suspected we do — it's when we still do it.
A silly little game about berries and predators
A team at Caltech — Dean Mobbs, who runs a cognitive-neuroscience lab there, with Ketika Garg and the lead author, Wenning Deng — sat people down in front of a foraging game. You wander a virtual patch looking for food, choosing among nine possible spots, whilst trying not to get eaten by a predator. First you play 120 rounds on your own, learning where you like to forage. Then they pair you with another player for 60 rounds, and now the two of you have to land on the same spot together. (I know. It sounds like the least glamorous afternoon ever handed out for science. Bear with me, because the boring wrapper is hiding something quite sharp.)
Most of us, it turns out, meet in the middle
Here's the first finding, and it's almost sweet: most people compromise. You liked spot three, your partner liked spot seven, and the pair of you drift towards five — some quiet middle ground between the two of you. And it holds, because it's reciprocal: I bend towards you this round, you bend towards me the next, and the little arrangement keeps itself alive. Which is, frankly, a kinder picture of human beings than we usually get handed. A lot of the time, we are willing to walk halfway, and to keep walking it as long as the other person does too. Hold on to that, because the story is about to complicate it — the same two people who found their middle ground so graciously are about to disagree, quite sharply, about who deserves the credit for having found it.
And then we do the books
Now the sharp bit. When the pair did well, and when it did badly, the researchers looked at how each person carved up the responsibility — and people took noticeably more credit for the wins than blame for the losses. Not a hard, cackling this was all me and none of it was you. Just a gentle, persistent tilt. The good outcome? A touch more mine than yours. The bad one? A touch more yours than mine. (I do this constantly, by the way, and I had years of training that was meant to make me catch myself at it. It didn't take.) Psychologists have a stiff name for it — egocentric bias in responsibility — which is a long way of saying: I round in my own favour.
The twist: it doesn't need the fog
Here's the part that made the researchers themselves sit up. You'd assume this sort of self-flattering maths needs a bit of murk to work — that we only claim the win because it genuinely wasn't clear who did what. Twenty years ago Dan Ariely wrote a book, Predictably Irrational, and one of its quietest, most damning ideas is what he called the fudge factor: give us a little wiggle room, a little ambiguity, and we'll shave a sliver in our own favour whilst still, sincerely, feeling like honest people. The wiggle room was doing the work. Take it away, the theory ran, and the shaving should stop. Except this study took it away. They made each player's contribution completely, unambiguously clear — you could see exactly who chose what — and the bias just carried on regardless. As Deng put it, the egocentric bias is usually seen under ambiguity, and yet even when the feedback made each person's contribution clear, people still tilted the books their own way. So the update on Ariely is a slightly bleaker one. It isn't that the fog lets us cheat. It's that we'll do it in broad daylight, with the receipts sitting on the table between us.
You don't need the fog to cheat
Ariely said a little ambiguity lets us round in our own favour; this study says we'll take the win and hand over the loss even when everyone can see exactly who did what.
Whether the team survives depends on how hard you round
And this is the part that actually lands on a Monday, because the bias isn't just a private vanity — it predicts whether the two of you can work together at all. Partners who wanted different things but rounded gently, who weren't too biased, drifted towards that shared middle and stayed there. Partners who both rounded hard — each quietly certain the wins were theirs and the losses the other's — spiralled. Each one's self-flattery cancelled out the other's, over and over, a small engine of grievance feeding itself. Same disagreement, wildly different endings, and the thing that sorted them wasn't who was right. It was how generously each person did their own accounting.
What to do with this on Monday
I'm not going to hand you a five-step plan for becoming a scrupulously fair auditor of your own contributions — partly because I don't have one, and partly because I'd be the last person qualified to sell it. But there's a smaller, more honest move. The next time a shared thing goes well and you feel that warm little that was mostly me, you don't have to fight it. Just notice it's there, and assume your partner is feeling the exact same warm thing about themselves, with exactly as much conviction. And when it goes badly and the blame drifts helpfully away from you, notice that drift too. Nobody's asking you to flagellate yourself. You're only being asked to remember that the receipts might not say what your gut insists they say — and that the person across from you is reading them every bit as creatively as you are.
Framing drawn from Dan Ariely, Predictably Irrational (the "fudge factor": a little ambiguity lets us behave dishonestly whilst still feeling honest). The findings come from a 2026 Caltech study in PNAS on how partners share responsibility for a joint decision. This is a popular-science reading, not professional advice — and it rests on a single lab-based game, so treat it as a sharp early result rather than settled law; correlation isn't causation.
In a 2026 Caltech study (PNAS) using a foraging game — 120 solo rounds, then 60 paired across nine locations — partners took more credit for wins than blame for losses, and the bias held even when each player's contribution was made completely clear. Framework: Dan Ariely, Predictably Irrational (the "fudge factor"). Popular-science reading, not professional advice; one lab study, correlation is not causation.
下面这段,是让研究者自己都坐直了的地方。你大概会觉得,这种给自己脸上贴金的算法,得有点浑水才好摸鱼——我们敢揽功,是因为本来就说不清到底谁干的。二十年前,丹·艾瑞里写过一本《怪诞行为学》,里头最不动声色、也最扎心的一个说法,他管它叫「作弊余地」(fudge factor):只要给我们留一点回旋、一点模糊,我们就会悄悄往自己这边刮一点,同时还真心觉得自己是个老实人。是那点模糊在替我们干活。按理说,把模糊抽掉,这点小作弊就该停了。可这个研究偏偏把模糊抽掉了。他们让每个人的贡献都清清楚楚、一点不含糊——谁选了哪个点,你看得明明白白——那点偏差却照旧不误。用领衔作者 Deng 的话说:这种偏差通常是在模糊的情况下才见到的,可即便反馈把两个人各自的贡献都摆清楚了,人们还是照样往自己这边算。所以对艾瑞里那句话,这个研究给的更新,是更冷一点的:不是那层雾放我们作弊,而是账本摊开摆在桌上、大白天里,我们照样这么算。
研究者自身が、ここで背筋を伸ばした。こういう自分に甘い計算って、少し霧がかかってないと成り立たない気がしませんか。誰がやったのか本当に分からないから、勝ちを自分のものにできる、って。二十年前、ダン・アリエリーが『予想どおりに不合理』という本を書きました。その中でいちばん静かに刺さる考えが、彼の言う「ごまかしの余地(fudge factor)」。ちょっとの逃げ道、ちょっとの曖昧さがあれば、わたしたちは自分に有利なほうへそっと削る。しかも、自分はちゃんと正直だと本気で思ったまま。その曖昧さが、仕事をしてくれてたわけです。じゃあ曖昧さを消せば、削るのも止まるはず。ところがこの研究、その曖昧さを消してみた。誰がどこを選んだか、一人ひとりの貢献を、完全に、はっきり見えるようにした。それでも、バイアスはそのまま続いた。筆頭著者の Deng はこう言っています。自己中心バイアスはふつう曖昧なときに見られるものだけど、二人それぞれの貢献がはっきり分かるフィードバックがあっても、人はやっぱり自分に都合よく傾いた、と。だからアリエリーへの更新は、少しだけ、さみしい。霧がわたしたちに削らせるんじゃない。スコアが机の上に全部出ていても、真っ昼間に、わたしたちは削るんです。