The Lockerbie Bomber and revenge

[The Minsky Moment]

Isn't deterrence one of the goals of the legal system?

yes but it is distinct from retribution.  you keep conflating the two.

It always annoys me when revenge is regarded to as a completely irrational action.  To me, revenge is a very rational reaction on a social level.  It lets people know that if they wrong someone badly enough, they would invite great wrath unto themselves, so great that it would be self-destructive for the person carrying out the revenge.  That terrorizing prospect tends to deter people from acting in a way that provokes revenge in the first place

See you use the word "revenge" but what you are actually talking about is deterrence.

Retibution is about making people pay for what they have done because that is a right or just result in itself, regardless of its effectiveness in deterring offenses.

[DGuller]

Isn't deterrence one of the goals of the legal system?

yes but it is distinct from retribution.  you keep conflating the two.

It always annoys me when revenge is regarded to as a completely irrational action.  To me, revenge is a very rational reaction on a social level.  It lets people know that if they wrong someone badly enough, they would invite great wrath unto themselves, so great that it would be self-destructive for the person carrying out the revenge.  That terrorizing prospect tends to deter people from acting in a way that provokes revenge in the first place

See you use the word "revenge" but what you are actually talking about is deterrence.

Retibution is about making people pay for what they have done because that is a right or just result in itself, regardless of its effectiveness in deterring offenses.

I never even talked about retribution as you define it, so I don't know how I can be conflating the two concepts.  I merely made a utilitarian argument about the benefits of two related concepts that are largely looked down upon in the West.  Acts of revenge on an individual level lead to some level of deterrence on the social level.

[ulmont]

I think Western democracies should learn some game theory.  Tit for tat, with an occasional but rare act of forgiveness, is the most effective method of discouraging actions detrimental to society.

Tit for tat with misunderstandings has a rather obvious failure mode.

The random act of forgiveness would take care of it.

Leaving you screwed from all time between misunderstanding and forgiveness.  Unless the forgiveness happens a lot more often than misunderstanding, this is not great.

[DGuller]

Leaving you screwed from all time between misunderstanding and forgiveness.  Unless the forgiveness happens a lot more often than misunderstanding, this is not great.

On average it would work out.

[ulmont]

Leaving you screwed from all time between misunderstanding and forgiveness.  Unless the forgiveness happens a lot more often than misunderstanding, this is not great.

On average it would work out.

To make this obvious, I'm going to assume that a message is misunderstood 1 in 4 rounds (alternating between player 1 and 2 as to who misunderstands), and that 1 in 8 rounds a party forgives (picking a player who would otherwise defect if possible; if both would defect, alternating between 1 and 2).

Code Select Expand


T1 Coop    Coop
T2 Coop    Coop
T3 Coop    Coop
T4 Coop    Defect (misunderstanding by Player 1)
T5 Defect  Coop
T6 Coop    Defect
T7 Defect  Coop
T8 Defect  Coop (forgiveness by Player 2, misunderstanding by Player 2)
T9 Coop    Defect
T10 Defect  Coop
T11 Coop    Defect
T12 Defect  Defect (misunderstanding by Player 1)
T13 Defect  Defect
T14 Defect  Defect
T15 Defect  Defect
T16 Coop    Defect (forgiveness by Player 1)
T17 Defect  Coop
T18 Coop    Defect
T19 Defect  Coop
T20 Defect  Defect  (misunderstanding by Player 2)This is not working out on average.

[DGuller]

Leaving you screwed from all time between misunderstanding and forgiveness.  Unless the forgiveness happens a lot more often than misunderstanding, this is not great.

On average it would work out.

To make this obvious, I'm going to assume that a message is misunderstood 1 in 4 rounds (alternating between player 1 and 2 as to who misunderstands), and that 1 in 8 rounds a party forgives (picking a player who would otherwise defect if possible; if both would defect, alternating between 1 and 2).

Code Select Expand


T1 Coop    Coop
T2 Coop    Coop
T3 Coop    Coop
T4 Coop    Defect (misunderstanding by Player 1)
T5 Defect  Coop
T6 Coop    Defect
T7 Defect  Coop
T8 Defect  Coop (forgiveness by Player 2, misunderstanding by Player 2)
T9 Coop    Defect
T10 Defect  Coop
T11 Coop    Defect
T12 Defect  Defect (misunderstanding by Player 1)
T13 Defect  Defect
T14 Defect  Defect
T15 Defect  Defect
T16 Coop    Defect (forgiveness by Player 1)
T17 Defect  Coop
T18 Coop    Defect
T19 Defect  Coop
T20 Defect  Defect  (misunderstanding by Player 2)This is not working out on average.

The flaw in your analysis is that forgiveness must be random, not with a predictable cycle.  In practice, misunderstanding would also be random, but I guess that part is up to you to define.

[ulmont]

The flaw in your analysis is that forgiveness must be random, not with a predictable cycle.  In practice, misunderstanding would also be random, but I guess that part is up to you to define.

It's not a flaw in my analysis.  Random or not, misunderstandings are going to introduce unwarranted defections until forgiveness sorts them out.  If the frequency of misunderstandings is high enough relative to forgiveness, forgiveness will not sort them out.  The predictable cycle was to avoid claims that I was cherrypicking a worst-case random scenario.

[DGuller]

The flaw in your analysis is that forgiveness must be random, not with a predictable cycle.  In practice, misunderstanding would also be random, but I guess that part is up to you to define.

It's not a flaw in my analysis.  Random or not, misunderstandings are going to introduce unwarranted defections until forgiveness sorts them out.  If the frequency of misunderstandings is high enough relative to forgiveness, forgiveness will not sort them out.  The predictable cycle was to avoid claims that I was cherrypicking a worst-case random scenario.

Obviously few systems are going to work with people who are unable to reliably perceive reality most of the time.

[ulmont]

The flaw in your analysis is that forgiveness must be random, not with a predictable cycle.  In practice, misunderstanding would also be random, but I guess that part is up to you to define.

It's not a flaw in my analysis.  Random or not, misunderstandings are going to introduce unwarranted defections until forgiveness sorts them out.  If the frequency of misunderstandings is high enough relative to forgiveness, forgiveness will not sort them out.  The predictable cycle was to avoid claims that I was cherrypicking a worst-case random scenario.

Obviously few systems are going to work with people who are unable to reliably perceive reality most of the time.

Right.  How rare are you expecting to find misunderstandings, and how often are you expecting to forgive?

[Zanza]

In my mind, reeducation is morally acceptable only for petty criminals.
A murderer should be punished to the maximum any given country's laws allow to, and not waste our time thinking of the reeducation of this murderer. That's what I mean by revenge. "You are a convicted killer, and so we will kill you or put you in jail for the rest of your life without possibility of parole (or whatever you call it)".

The recidivism rate among released murderers is much lower than that of petty criminals. So why do you think that it we have a better chance to reeducate petty criminals than murderers? Maybe we should execute all shoplifters (very likely to reoffend) but try to reeducate the murderers? 

Numbers: http://www.cor.state.pa.us/stats/lib/stats/BJS%20Recidivism%20Study.pdf

[DGuller]

Right.  How rare are you expecting to find misunderstandings, and how often are you expecting to forgive?

I expect misunderstandingsto be quite rare.  Most of the time people in any culture get along just fine, meaning that they choose to cooperate without a hitch.

[DontSayBanana]

The recidivism rate among released murderers is much lower than that of petty criminals. So why do you think that it we have a better chance to reeducate petty criminals than murderers? Maybe we should execute all shoplifters (very likely to reoffend) but try to reeducate the murderers? 

Numbers: http://www.cor.state.pa.us/stats/lib/stats/BJS%20Recidivism%20Study.pdf

Petty crimes tend to be influenced more by class and standard of living and are less likely to be positively affected by penal rehabilitation methods?

[Valmy]

Petty crimes tend to be influenced more by class and standard of living and are less likely to be positively affected by penal rehabilitation methods?

Or influenced by things like drug addiction that are hard to rehabilitate.

[ulmont]

The recidivism rate among released murderers is much lower than that of petty criminals. So why do you think that it we have a better chance to reeducate petty criminals than murderers? Maybe we should execute all shoplifters (very likely to reoffend) but try to reeducate the murderers? 

Numbers: http://www.cor.state.pa.us/stats/lib/stats/BJS%20Recidivism%20Study.pdf

Yeah, I've often thought everyone should get one free manslaughter offense for that reason.

[ulmont]

I expect misunderstandingsto be quite rare.

According to recent research published in the Journal of Personality and Social Psychology, I've only a 50-50 chance of ascertaining the tone of any e-mail message.

http://www.wired.com/science/discoveries/news/2006/02/70179