Nosy: forum

Advanced Jail Bust Guide

I remember a few years ago when I was looking to understand how busting works and couldn’t find any guide about it. I have seen pieces of advice here and there but they were mostly hearsay or very approximative (maybe I didn’t know where to look). I forgot about it until a little over a year ago when I started busting people more actively and started wondering how I could maximize my daily bust count. In the meantime I joined a law firm which has a perk that let’s you see your success chance before attempting a bust. This perk let me experiment a few ideas and get a better understanding of the mechanics at play. Most of this was done earlier this year and now I've decided it's time to share it with all of you who might be interested.


This guide might not yet be helpful to everyone since everything could only be tested on 10* law firm employees. Actually its place is somewhere between guide/discussion. Nevertheless, everyone can get a better understanding of the trends. It is important to note that some things are still speculations, I simply propose a model and theory that I could test with the help of other law firm employees. I’ll gladly accept any feedback as it will probably be positive for our future and better understanding of how the jail works.


If you are already very familiar with busting and are only interested in some graphs and slightly more in-depth explanation of the findings, you can skip ahead.

Note: Appart from busts there are other ways to get out of jail. The lawyer business card (instantly get out of jail when using it, no questions asked), a perk from the law city job (get anyone out of jail for 15 job points), bails or self busts. None of these options count towards the total bust stat (thanks to Politics [2250481] for testing the federal judge perk).

The basics

What is the point of spending all this nerve on busting? Some say it is a waste. Some say it will gives you some crime experience, but the gains are probably very little. To me, the best thing about these is there are bunch of merits to collect! (4 awards and 7 medals. awards at 250, 1000, 2500 and 10000 busts and medals at 250, 500, 1000, 2000, 4000, 6000 and 8000). A lot of player overlook them because they require a long term commitment. It is up to you to decide if you want to add busting to your daily torn routine and if you do, this guide is here to help you. With the recent changes related to crime experience loss for failing crimes, the jail will likely get packed and it would be the perfect time to get started!

The outcomes

Everytime you perform a bust, there are three possible outcomes. You will either succeed (yay!), fail without penalty or fail, get caught and sent to jail yourself. Please note that getting jailed for a failed bust will not damage your crime experience or nerve bar. There is probably no loss, and if there is one it is so little that it’s negligible (I tested that unintentionally by being reckless and getting jailed a lot doing this research, however not only did I remain part of a PA team but my crime experience kept increasing).


After you click the bust button, the outcome is dependant on a few things:

• The bust's difficulty (target level and time they're in for)
• Your skill (your level, your perks)
• The penalty from previous busts (those you did in the past 72 hours)

The perks

Education

• LAW2920: Gain a 5% bonus to your skill in busting
• LAW2970: Gain a 10% bonus to your skill in busting
• LAW3102: Gain two bonuses: Busting is 50% easier and buying is 50% cheaper

Company (Law Firm)

• Closing argument(5*): Easier to bust more people at once.
• Educated decision(10*): View success chance of potential busts.

Faction (criminality branch)

• Bust skill I-X: Increases bust success chance by 5% (so up to 50% with all ten upgrades)
• Bust nerve I-III: Decreases bust nerve cost by 1 (so busts will only cost 2 nerve with all three upgrades)

The Book Shawshank Sure Ain’t For me!

• Large jail bust & escape boost for 31 days

Of these perks,“Bust nerve” is the only one that will not influence your success chance, but it will significantly reduce the nerve you have to spend, especially if you want all the merits.


The wording for the education and faction perks are similar but different, it is unclear if they act the same way. If they’re the same, that would be a bonus of up to 147.5% in busting skill (to be tested). How the 5* law firm perk works is still unknown and a mystery. I suspect it isn’t like the other ones due to the very different phrasing. However,both the faction and education perk allow you to "bust more people at once" so maybe they’re the same in the end. The problem is it cannot be tested yet as we would still need the 10* perk to study its effect…
The effect of the book hasn’t been thoroughly tested either. As a reference, anecdotal evidence showed daily count about tripled for an active buster. The bonus is huge and it is highly recommended that you join a law firm and a faction with the perks while reading this book.

Level and sentence duration -> difficulty:

The higher your level, the easier it will be to bust someone out of jail. Inversely, the higher their level, the harder it will become. Also, the longer their sentence, the harder it will be. Without considering the perks and recent busts, the difficulty of a bust is proportional to your level, inversely proportional to your target’s level and proportional to their sentence duration arbitrarily increased by 3hours. This means it gets twice as easy if you double your level, and it also means that for short durations, the time is less important than the targets level.

As an example, imagine a level 20 jailed for 1h and a level 10 jailed for 3h. 20*(1+3) =80, 10*(3+3)=60. In this case it would actually be easier to bust the level 10 even though he’s in jail for much longer. But a level 15 that is in for 1 hour is exactly as difficult to bust as the level 10 we just mentioned. 15*(1+3)=60

Obviously you won’t do this every time but it can help you recognize more favourable potential busts. These optimal targets never stay in jail for long so you’ll have to be quick.

If you want to know more about the formula, it will be explained in more details later.
It won't tell you your success chance but it can help you decide which target is best.


Note: The reason why a person is in jail has no influence on the difficulty of the bust. In bold because I have read that it is supposed to be easier to bust someone if "They were sentenced for a failed bust or self-bust" on the wiki. During our research we noticed no significant difference between failed busts and any other reason. In fact, if you failed to a bust with a lowere difficulty than yours (formula above), don't attempt to self-bust because you will inevitably fail. Maybe they meant they’re easier to bust because the sentence is shorter, in which case it is true but misleading.
It is also stated that "There needs to be a ten-minute break between each bust chain.".
Again, nothing in our results supports a “bust chain” theory. Actually it was something we were trying to find but didn’t. My hypothesis is that they only experienced the recovery that is initially very quick just after each bust. And if you just did a bunch of them, it adds up and you can do one more soon after.

Penalty

All the busts you did recently have a negative impact on your ability to successfully bust more people out of jail. The penalty will be the same no matter who you busted, their level, the reason they got caught or the time they're in for. These penalties can add up to the point where you need to take a break and wait a while before attempting any more busts. That means it will get harder as you bust more people but counter-intuitively, you will maximize your daily count by flirting with the limit (the greater your penalty, the greater the recovery). “Recently” is relative and might not be the best word as it takes 72 hours (3 days) to completely recover from a single bust. however the recovery is highly non-linear in time and you will have recovered from half the penalty in only 10 hours. Taking a break after a large amount of busts is actually a good piece advice.
It is worth mentioning that each bust incurs an individual penalty, and your total penalty is simply the sum of the contribution from each bust.

One more thing, the higher your level, the smaller the penalty for each bust. Think about it if you are level holding but want to increase your bust potential.

How to know your limit?

If you want to reach your maximum daily bust count, bust aggressively for a few days, focus on low level/time targets and you’ll get a rough idea of how many you can do. If you fail too many, take a break and try again later. How long you have to wait is what you’re trying to find out (note that it depends largely on your level and perks). Once you know your limit approximately, you will be able to achieve it daily without wasting too much nerve on failed attempts. A good indication of whether you should take a break is the outcome of a failed busts.The lower your success chance is, the more likely you will get jail time instead of simply failing without consequence. Also, if you fail but don’t get jailed, you know with certainty that your success chance was greater than 0%. (This was actually stated in a bug report sadly I can’t find the post anymore).
The difference between taking a break and waiting for a while before next attempt or spending all your nerve on busts might only change your daily count by a few more busts and it is probably not worth it. What matters most is picking the right target, but in the end, you will also be limited by what the jail has to offer.


I’ll try to keep thepost updated as our understanding improves. Because of the 3 daysbust penalty it can get tedious to run some test and validate an hypothesis.


That’s it for the basics. I hope this will help you towards your 10k busts merit and possibly even the top 250 HoF.


Self busts:
A quick test of the self bust option revealed that it doesn't follow the same rules as "regular" busts. The nerve cost is much higher, a portion of your total nerve instead of 5 (or less with the right faction perks). It costs half your maximum nerve to attempt a self bust, this ammount can be reduced to a third of max nerve after completion of the education class LAW2990: International Human Rights. I am unsure about the cost without reduction, maybe someone that hasn't completed the education can confirm.
Successfully busting yourself out of jail doesn't affect your success chance for regular busts, and it doesn't seem like regular bust affect self busting either. So far we don't know how success or failure is decided for this specific case, maybe a fixed rate of success. Considering the extreme nerve cost, difficulty to test this feature and the fact it doesn't count as a bust in the stats, it won't be investigated in further detail for now.

Graphs - formulas - remarks - validation

Bust difficulty and skill

A bust’s difficulty is proportional to the inmate's level and to their sentence duration arbitrarily increased by 3 hours.

[image: i.imgur.com]

Your skill is proportional to your own level and increased at least by the faction perk.


How the perks are applied is still uncertain but early testing on the faction perk suggests they simply amplify the buster’s level. It will be hard to find testers without the education perks and simply impossible to test the law firm perk for now.


Remark: it has been anounced (by Ched) that the total amount of busts has aninfluence but we couldn’t find any meaningful difference at all from our testing (busters ranging from 1k to 4k busts)… This post is 6 years old so it is very likely outdated as it is the case for this statement: "For failing a bust and being sent to jail, like before you'll receive double the amount of jail time your target had." .(The total amount of busts you've made will allow you to bust people more often)
Correction: it seems the total amount of busts does increase your success chance. By how much is still unknown, most probably more noticeable when you get started and have done very few busts.. There still is a cap on your success chance (determined by level and perks) and that cap won't increase with more busts.

Here is a representation of the relation between success chance, level (skill) and bust difficulty. All the testers had all the perks (education,faction and law firm). Those that do not have the "no penalty" mention had performed one or more busts within the previous 72 hours.


[image: i.imgur.com]

First thing to notice is that all the points gathered without penalty describe the same line. In that "no penalty" state, buster level, in-mate level and jail time are enough to predict success chance accurately (this will be helpful later). The second thing is that all the lines are parallel, previous busts shift the line to the left, effectively reducing success chance for subsequent busts.

Upon request here is a comparison between DocTorn's time*level and the new metric for difficulty applied to a single set of in-mates. The data

[image: i.imgur.com]

Penalty

Now, penalty was a pain… It really was a pain. Initially we had no idea that there was a 3 days recovery period, some things are still unclear. For instance we could observe that the first bust after a long break (how long?) always has a reduced penalty but the recovery rate for that bust is the same as for any other. It seems that there is a cap on your maximum success chance and that you will regenerate your ability beyond it, giving the impression of a “reduced penalty”. We can see it as some sort of discount if you haven’t done any in a while. It could explain the phenomenon but it’s still too recent to be categoric. I wouldn’t be surprised if we find a simpler explanation later…
Another thing is, the recovery is gradual and there is a noticeable and sudden “jump” at the end of the 72 hours recovery. That is in fact what led us to notice this specific duration for the penalty.


Enough writing, now here’s what the evolution of a bust’s penalty looks like for a level 61 with all the perks. The initial penalty P0 is inversely proportional to the buster’s level.


[image: i.imgur.com]
[image: i.imgur.com] [image: i.imgur.com]

c = 0.1 [-/h]

This formula tells us half the penalty for a bust is gone after exactly 10 hours.

Success chance

With the previous definitions, success chance for any bust can be evaluated with this formula:

[image: i.imgur.com]

The two constants a and b are supposedly shared by all players, a was the same for all three testers (level 28-61) and is not influenced by the faction perk. The faction perk factors both the skill and penalty terms. I will rewrite the formula once known we know more about the other perks.

a = 266.6 [-] , b = 0.427 [- / minute]

Application of the model for testing

The model has been tested on two players so far, one short test (level 53, 8 busts over a 30h period) and a long test (level 61, 81 busts over about 100h).
The aim was to verify that, knowing when previous busts occurred, the model could predict success chance of subsequent busts accurately. In this case we will compare two values, the penalty measured in jail at some given points in time and the prediction of the model.

Penalty(measured)
Our reference will be the penalty (measured). It was obtained by gathering measures from the jail (time of measure, level of in-mate and duration of their stay) and substracting it to the success chance without penalty (That we can get from the equation of the line on the first graph). The measure is actually an average over as many players as possible at a given time. The idea of taking the average was to reduce the error but in some cases (heavy penalty or almost empty jail) the penalty was computed based on a single measure.
Each blue dot represents a measure in jail.

Penalty(estimated)
Keeping track of all the busts, the estimation is computed according to the success chance formula presented above.
Basically, it's as simple as keeping track of when you busted someone out of jail, applying the penalty formula and add their individual contributions.
The red line is purely analytical.

[image: i.imgur.com][image: i.imgur.com]
Considering there is an intrinsic error in the measure of the success chance, the results are very promising. 1% off from the actual success chance is surprisingly close and I am therefore confident in the model even though its applications are still limited.

Possible improvements

At this point there is a lot that can be improved, but mainly two points.

• The perks require more attention so that the formula can be generalized for every player
• It would be good to get better estimates of the various constants (a, b, P0)

I’d say with time, a busting tool (similar to the law firm perk) could possibly be developed from this, or we could at least get an easy formula that lets you know what maximum daily bust count to aim for based on your level/perks.

On this matter, can an API Master tell me whether the log can be accessed via API (to get busts timestamps)?


Thank you for reading, I hope this was helpful and/or interesting. If some of you are interested in doing some testing on the subject just send me a mail. I can also set up a google sheet with some examples if that would help make all this more clear, let me know!


Thanks to those who contributed to this project.

TRex, ThtAstronautGuy and panieri for gathering data.
CloudJumper for providing usefull links and information.
Special thanks to sacham for testing and providing a crucial data gathering tool.
and all of them again for sharing their insight and discussing ideas.

ShadowRazers and RazerLaw for the perks.

Extra credit to Nobody, a pioneer in busting -> His guide


tl;dr: To maximize your daily bust count, focus on low level targets with short sentences. after a long break you can start with players around your level and gradually aim lower.
You will then be limited by your perks and your own level (a higher level makes it easier).

Come now and get a free elbow to the nose!!

Hello,

So my problem is that I lost classes that were previously validated. The only proof I have is that I need to retake some introduction classes even though I validated classes that require it. I guess I can't do much about the other ones but it would already be nice to recover the introduction classes.

Here's an example: swimming requires the introduction class and it's validated, however the introduction class is not validated anymore.

[image: i.imgur.com]

Can you do something about it? Thank you in advance.

N.

Did a test on plane hijackings after reading some of the comments here. This week: 8 PH teams with about 85k arsons total each. 5/8 were successful. I'll report next PHs here when they're done.

Hi, here's the problem

Attacker wins, attacker is defeated and then "someone fought him/her off" event is displayed. I don't think this is intended.
From what I've seen, it happens when the initial fight is already over but attacker hasn't clicked left/mugged/hospitalized button. What makes me think that is the timestamps are the same for both attacker and defender losing. This probably is a remnant from the time there wasn't auto-finish at the end of fights.

Here are the events and the fight log

outspoking attacked you but Colonel fought her off [view] 17:20:51 09/07/22
outspoking (and 0 others) attacked you [view] 17:20:51 09/07/22

The attack log: https://www.torn.com/loader.php?sid=attackLog&ID=0b25f56fbe8336d07fb2282b84417dbd

Looking for someone still

Hiring again

Ideal for bust merit or PA

Applications open -> RazerLaw

Here's some data I've been gathering (667 PA outcomes so far) plus my hypothesis for success rate formula.
[image: i.imgur.com]

Some samples are very small for some intervals but it gives a rough idea

Still gathering data and I'll make a post about it when I have the time, before OC 2.0 hopefully

Congrats

Update: now at 15 success 2 fails for 120-140k total arsons teams. Still think we are way above the actual success rate for that range since our main teams are around 200-220k with a 70% success rate

Still it seems to me that 120k arsons would be a good starting point for factions that are getting started with PAs

It is slow to gather results with the 8 days wait. 5 success 1 fail in the 120k-140k range. I was very surprised that many teams succeeded, we've been super lucky. Can't say much more for now. Bellow 120k, success chance seems to be somewhere in the 10-25%.

Considering this and what Fatbears reported I wouldn't go bellow 120k arsons total. But If you do please report anyway

Hey, sorry I tried giving credit, didn't know doctorn was using your formula at the time, my bad.

Been very busy lately that's why I haven't corrected it since your recent chat. Didn't mean to hurt your feelings...
I didn't steal your formula though, just proposed one that's more fitting.

Edit: added a link to your guide

Sadly it didn't happen...

It's something I can't test myself and I'm not actively looking for testers anymore as I am more busy now. Though If someone volunteers we could do it.

If we do, it would be great to have the assistance from someone who's decent with API too, it would make things easier (get bust success/fail from log for example so nothing too complicated).

Very iteresting. Currently experimenting with weak PA teams as well as recording stronger ones. Started recording in july so I'd rather share later when there's more data. So far I have data points in the 70-120k and 180-240k ranges, nothing in between.

What I can share is we've had a success with a team just over 75k total arsons although as expected success rate is bad.

My observations fit with maledict's experiment in the corresponding range.

Also I am still not convinced there is a soft cap for success

Been tracking our members CE so I could test hypothetical teams using your method. I was using a similar one too.

Pretty good balancing for how simple it is (within 5-10% of average in our case).

Then all that is left to do is figuring out how many teams to plan.

Ok. It's not really a problem anyway, just thought it didn't make much sense.

Thank you

Hi, here's the problem

Attacker wins, attacker is defeated and then "someone fought him/her off" event is displayed. I don't think this is intended.
From what I've seen, it happens when the initial fight is already over but attacker hasn't clicked left/mugged/hospitalized button. What makes me think that is the timestamps are the same for both attacker and defender losing. This probably is a remnant from the time there wasn't auto-finish at the end of fights.

Here are the events and the fight log

outspoking attacked you but Colonel fought her off [view] 17:20:51 09/07/22
outspoking (and 0 others) attacked you [view] 17:20:51 09/07/22

The attack log: https://www.torn.com/loader.php?sid=attackLog&ID=0b25f56fbe8336d07fb2282b84417dbd

I didn't cover the CE rewards for busts as I have no idea about that besides what I read in a newspaper article. I think that article is what you're refering to.

Busting might help you get a little CE but it won't replace crimes. I'd go for the easy ones to increase bust count faster, it takes a very long time to reach 10k. If you're more interested in CE, do very few hard ones and focus on crimes instead.

I'm not sure what your question is about the law perk.

Good call, my bad and thank you for reading thoroughly... I figured it might have changed without me knowing as it was a 5 year old post and I wasn't following the changes at the time and for a long time. I must have missed the "expect change to jail times for failed busts" part which which would have influenced my interpretation.

I just tried a similar test with roughly 6k more busts than when I started testing (2k at the time of the earlier testing and about 8k now). There is a noticeable difference since then, which could be (very probably is) due to the additionnal busts. The difference is very subtle though... Average eror was 0.15, it is now -0.67 (error = penalty measured - penalty estimated). Similar distribution
So we're looking at +0.5 to +1 bonus to success chance for 6k more busts. To be clear, it's not a relative increase, it's a flat increase to the success chance displayed. a 10% bust would become 10.5-11% or a 90% bust would become 90.5-91%. The difference is probably more important at earlier stages as Stente stated... I hope we'll be able to test it soon with the help of someone that has done very few busts so far...

The cap to success chance without penalty didn't move though, so I suspect the bonus provided by number of busts is this extra success chance that would go above cap, if that makes sense. That's only a supposition and isn't worth much for now, just a hunch... If that's the case, the bonus would still be worth less than the penalty from a single bust at 8k total. It's something but not as good as any of the perks (edu, law, faction).

I'll edit the post, thanks for calling me out on this.

Also I'm a little too busy to start some large scale tests at this time... I'll take some more measures for later but the thinking will probably have to wait for a while.

Yeah I'm sure it didn't get more difficult at 10k. makes sense to slow down after the merits.

Was your level the same when you could do 2-3 per day and 8-10 per day? Were you attempting busts of equal difficulty?

I agree there are a lot of variables, that's why it's difficult to draw that kind of conclusions just from looking at the graphs. As you said, it might work like stat gains at the gym. It isn't out of the question though I'd still bet on the level changes.

Luckily I just found someone willing to help me test this with only 134 busts at the moment. We first need to find him a 10* law firm and then we can start. Considering the limited amount of daily busts it will take a while though.

With that graph and that logic we could also assume that it gets much harder after 10K busts?

Are you level holding at 88 since july 2017? Otherwise I'd say it's most likely the increasing level that's responsible for the easier busts and increasing slope.

Same question for JussiJernkuuk?

The point at which the slope becomes constant doesn't appear to be the same in both curves. And red graph looks more like two lines that intersect around 2.4k busts. (level holding then sudden increase early 2018?)

Still doesn't exclude that it has an influence but I remain highly sceptical that it does.

PS: impressive regularity, my graph is a mess in comparison

Personally I don't trust Yata's estimates at all. It Works fine for people that didn't get jailed much, stayed away from drug/hacking crime etc, not so much for the rest. It tends to overestimate CE, sometimes by a lot.

I use some of our members that I have a fair estimate of their CE as references, for the 60NNB crowd that is. Then for the rest, I use the NNB milestones and assume they're evenly distributed If that makes sense. Not 100% accurate but still better than Yata's estimates most of the time