Max number of prisoners ? (no mod)

[Kunovega]

I know you can edit the save file; but whats the most prisoners you can get without doing this?

So far I seem to be stuck with around 150 as my max.

The problem is only a random number between 8 and 29 seem to show up each day, and it's usually closer to 8. With only that few prisoners coming there is now about the same number of prisoners leaving each day for completing their sentence so it ends up just balancing itself out. (I have all 3 prisoner types turned on)

I see no way to get the system to allow 200+ prisoners without basically cheating and editing the save file

I'm basically at a point where if I start a new game I can easily complete every grant, make all the prisoners happy and than just sit there and cycle in cash. My latest prison is at a 512 cell capacity and I've bought enough land around it that I could easily turn it into a 1000 inmate prison. But I can't seem to actually get the prisoners needed to fill the prison. So I just have a big empty prison.

[xander]

At this point, the equilibrium capacity (which I am not going to compute right now)* is of the order of magnitude that you describe. By closing your prison to everyone except maximum security prisoners, you can stretch things a bit more and there are ways of gaming incomplete game logic to get more prisoners into your prison, but you are around the limit without doing anything cheesy.

xander

* I started to write about continuous time Markov chains, but then decided that (a) I don't have enough date handy to work the model and (b) no one but me really cares. :(

[Geordie1892]

Hi both

Apologies if I am way off topic here but I don't believe I am. I don't run any mods and am currently at 309 prisoners accepting only minimum and medium security prisoners. I don't seem to have the issue you guys are discussing. I am using a medium sized map and have not bought any additional land. The capacity is 450 currently. I am receiving between 8 and 29 prisoners each day (as you suggest) but only lose at most 5 prisoners per day due to release.

Aware that this doesn't answer the question and perhaps raises a few more. Would be interesting to see if anyone sees a limit at around 150 prisoners or not.

[Kunovega]

Hi both

Apologies if I am way off topic here but I don't believe I am. I don't run any mods and am currently at 309 prisoners accepting only minimum and medium security prisoners. I don't seem to have the issue you guys are discussing. I am using a medium sized map and have not bought any additional land. The capacity is 450 currently. I am receiving between 8 and 29 prisoners each day (as you suggest) but only lose at most 5 prisoners per day due to release.

Aware that this doesn't answer the question and perhaps raises a few more. Would be interesting to see if anyone sees a limit at around 150 prisoners or not.

To some degree it may be random luck of how many prisoners come in each day. If you are getting 29s fairly frequently you could get to 300 in 11 days

On the other hand if I'm getting 8's I would be at 88 prisoners in 11 days; I just had 6 days in a row with 8 prisoners a day (and 5 to 10 leave a day)

I don't know what algorithm they are using for the random number generation; but if they are using something that doesn't change frequently enough for a seed number it can lead to strings of the same number coming up more frequently than would be expected

[xander]

I don't know what algorithm they are using for the random number generation; but if they are using something that doesn't change frequently enough for a seed number it can lead to strings of the same number coming up more frequently than would be expected

In all probability (haha) they are using the the rand() function provided by C++ (I can't see any reason why they would want to use anything else). For the purposes of a video game, rand() is perfectly acceptable and will lead to behaviour that appears random to all but the most find-grained statistical analysis. Long streaks of the same number of prisoners appearing day after day is actually a sign that the random number generator is working well (streaks are *expected* in truly random data). The problem is that people really don't understand random behaviour very well.

Again, this could all be modeled with Markov chains to get a better idea of what the equilibrium population should look like it (and, since I have been thinking about this a bit more, one could use discrete time Markov chains by assuming that prisoners arrive and are released at the same time---this assumption makes things much simpler, and does a pretty good job of modeling the situation). Unfortunately, it has been four or five years since I have computed equilibria for Markov chains, most of my books are in boxes at the moment (we are moving to California in two weeks), and I don't have the energy to deal with it right now.

If I get motivated, I'll have another look at it in four or five weeks when I have settled down and unboxed some books.

xander