www.lesswrong.com/posts/3rxMBRCYEmHCNDLhu/the-pavlov-strategy
1 correction found
they’ll be stuck in a mutual defection equilibrium.
That is not how Tit-for-Tat behaves after a single mistaken defection. Standard Tit-for-Tat gets trapped in an alternating C/D, D/C retaliation loop, not a mutual-defection (D/D) equilibrium.
Full reasoning
Under the standard definition of Tit-for-Tat, each player cooperates on the first round and then copies the opponent’s previous move. If two TFT players are cooperating and one defects by mistake, the next rounds go:
- Round 1 mistake: D / C
- Round 2: the mistaken player copies the opponent’s previous C, while the other copies the observed D → C / D
- Round 3: they swap again → D / C
- Round 4: C / D
So the error produces an alternating retaliation cycle, not a stable mutual defection state (D / D).
This matters because the post’s comparison is specifically about how quickly different strategies recover from noise. Saying TFT falls into mutual defection misdescribes the actual failure mode: it is vulnerable to a persistent echo of alternating retaliation.
1 source
- Prisoner’s Dilemma (Stanford Encyclopedia of Philosophy)
Tit for Tat (henceforth TFT) simply cooperates on the first round and imitates its opponent’s previous move thereafter.