Please post any errors you may have with the Conan online character generator here. You can find the generator at http://conan.modiphiusapps.hostinguk.org
5 Likes
That’s a pretty sweet generator! I like that it exports to the fillable character sheet!
Is there particular reason to be in this weird domain? 
Weird domain?
given that they own modiphius.com, he probably wonders why they didn’t just host it under conan.modiphius.com or similar.
That is correct 
Modiphius.com or .net, as the store makes more sense. Initially I thought this is some fan made generator and skipped it
That would have been ideal but it did not end up that way. We will look into domains again at a later date.
Which is why I put Official in the title.
I don’t quite follow you, Steve 
First let me start by saying how awesome it is you provide this tool. Very well done and speaks volumes to you as a company. So many other games require you purchase a third party application just to do this. So what follows is NOT complaining, just an observation offered constructively to help improve the tool.
My players, currently 4, all used this tool to create their characters (Normal method, roll for each entry except for archetype) and the party currently is a Nemedian, 2 Dafari, and a Hyperborean. The two Dafari surprised me, but in a sample set of 4, anything is possible.
Given how easy it is to make characters with this tool, I proceeded to make a number of additional ones, mainly because I like to see how the rolls change the character, ending with something different than your initial thoughts. In doing this I was surprised by the number of Dafari and Hyperboreans I rolled.
This got me to wondering about the randomness of the actual algorithm to roll/determine the homeland and the subsequent rolls (other options seemed to come up all the time too). So using Excel I calculated the base % chance of each homeland (brute force…unfortunately my stats skills are horrible) and rolled homeland 100 times (Core/Normal, click Roll Homeland, click Conan logo, repeat). The following is what I found:
| Roll | Nation | % Chance | % Actual |
|---|---|---|---|
| 2-7 | Aquilonia | 5.3% | 0.0% |
| 8-9 | Bossonian Marches | 3.8% | 0.0% |
| 10 | Keshan | 2.3% | 0.0% |
| 11 | Kush | 2.5% | 0.0% |
| 12 | Khoraja | 2.8% | 0.0% |
| 13 | Cimmeria | 3.0% | 0.0% |
| 14 | Hyperborea | 3.3% | 35.0% |
| 15 | Border Kingdom | 3.5% | 2.0% |
| 16 | Nordheim: Asgard or Vanaheim | 3.8% | 6.0% |
| 17 | Ophir | 4.0% | 0.0% |
| 18 | Koth | 4.3% | 0.0% |
| 19 | Argos | 4.5% | 0.0% |
| 20 | Gunderland | 4.8% | 9.0% |
| 21 | Brythunia | 5.0% | 0.0% |
| 22 | Nemedia | 4.8% | 5.0% |
| 23 | Zamora | 4.5% | 2.0% |
| 24 | Shem | 4.3% | 3.0% |
| 25 | Zingara | 4.0% | 0.0% |
| 26 | Corinthia | 3.8% | 0.0% |
| 27 | Stygia | 3.5% | 3.0% |
| 28 | Khauran | 3.3% | 1.0% |
| 29 | Zembabwei | 3.0% | 0.0% |
| 30 | Punt | 2.8% | 0.0% |
| 31 | The Black Kingdoms | 2.5% | 0.0% |
| 32 | Darfar | 2.3% | 9.0% |
| 33-34 | Turan | 3.8% | 5.0% |
| 35-36 | Hyrkania | 2.8% | 0.0% |
| 37-38 | Iranistan or Afghulistan | 1.8% | 14.0% |
| 39 | Vendhya | 0.5% | 0.0% |
| 40 | Khitai | 0.3% | 6.0% |
The percentages seem way off for a sample set of 100. Especially when it comes to Hyperboreans and a few other races. What is the seed used to determine the numbers? Might this be improved in some way? I just seems not so random from my test.
As I said, this tool is amazing, and I really do love it. I plan to keep using it, but for the other players that are coming in, and heaven forbid any new characters due to character deaths, I will have them roll actual dice and look up the value. Just offering my observations in the hopes of helping you improve the tool.
Edit: If it matters, my test was done on a Mac (iOS), but the actual characters were done on a Windows machine.
I did a little more testing, and it might be a look up issue, not a seed issue (or maybe both). I did the last option, Random, and an example of what I am seeing is:
Rolls: 7, 3 Hyperborea (should be Keshan)
Rolls: 7, 14 Nemedia (should by Brythunia)
I will take a look. Computer “random” is not entirely random at all times. But there might be something going on that I can improve. Thanks.
I did send a follow up, I think the randomness is fine, but it looks to be an issue with matching or looking up the values. When using the last option in the tool, Random, you can see the numbers rolled and reported Homeland. These do not match up. I just ran another Random result and for Homeland I got a 1 & 2 which the program is saying Hyperborean.
Thanks and I really do like this tool.
Ah, might be something to go on. Thanks, I will check it out.
Is it an entropy issue? Assuming this is hosted on a Linux machine, is rngd running?
It is Windows.
OK, I found the problem. Will fix it and upload in a bit. Thanks for pointing this out.
Awesome, and not a problem. Thanks for making such a great tool.
1 Like
Glad you like it