So let's apply some deductive reasoning here.
- They ran around 20 simulation games and did not notice AI Germany being beaten by Russia in most games by 41, definitely by 42.
- They ran around 20 simulation games, did notice AI Germany being beaten by Russia in most games by 41, definitely by 42 and published 1.3.3 anyway.
- They ran significantly less than 20 games, so few that they took AI Germany losing in 41/42 as a fluke instead of a consistent outcome.
- They did not run any simulation games.
Now I for one believe PDS as a software game development company has enough experience developing software to have good enough QA processes to catch something as big as AI Germany losing consistently to Russia in 41. So the 1st line of deductive reasoning just doesn't make sense. Which leaves us with the last three options as being possible, which one do you think it is? You could add another option if you think there is one.
I'm pretty sure we had already discussed this in the feedback thread on the AI already, and keeping on topic is nice, but sure:
As I said Germany has statistically always lost to soviet union in all our patches, sometimes more sometimes less. We ran our games and it shows germany failing earlier than before.
Looking at reasons it was because soviet union no longer:
1) built really trash motorized divisions
2) no longer wasted lots of troops on neutral borders
3) nations overall now value their own safety over helping allies (goal: italy should better protect itself and be less soft underbelly)
Basically soviet AI had improved a lot. Germany's also improved, but soviets more. The big issue for germany was front shuffling halting their push and letting soviets overwhelm and push them back. Japan and china also had front shuffling issues limiting them, but about in equal measures so that helped to delay the conflict.
Because we changed production quite a bit we looked at that, but it did not have a very large impact on AI performance overall. It may have slightly impact things, but if so the smallest of several options.
To clarify, the reason a germany losing earlier is bad is that for minors or non UK-allies they may not have time to involve themselves in the conflict which diminishes their experience if they are trying for the historical war. We then looked at consequences and options:
1) Nerf Soviet union - Since a vast majority play germany, or end up fighting the soviets this would make their experience worse. It also feels wierd to nerf a nation for AI reasons rather than player reasons when fighting them. removing some AI improvements were also on the table, but is clearly a bad idea.
2) Buff Germany - Most people already consider germany easy to play and there are also buff sliders available so we didnt consider this a good option
3) Fix front shuffling for good and rewrite logic for how AI nations should manage exp forces - Downside is that this is not quick nor easy. both a long term problems since the start of HOI4 (and hoi3 too to a certain extent). We estimated 3-4 weeks of work if all went well.
We felt 1 and 2 would negatively impact the majority of players and 3 would take so long we couldn't wait a month to release the patch. We also looked at the feedback from the 2k or so players who had been testing the open beta of 1.3.3 for a week (I think it was?) and while there was some comments on germany soviet balance it was a minority while majority seemed to like things. So we released the beta as is with some minor tweaks and in the next dev diary we talked about how we were making front shuffling our top priority.
Was it the right choice? Up to you to decide. I think so.
) - if, instead of features, the diary had focussed on a roadmap along these lines that would definitely have been useful and a good approach. ...... If Podcat had the time, I think it'd even be worth having a special 'bonus' roadmap DD, or to bump next week's content to put this in.
