AI kindness.

[semmes]

I was testing AI plans, so this is not a test, only an impression; that has been happening all the time.

An AI ridge, mostly occupied by teams, under artillery fire for something like 10 minutes. 18 units, they lost 4 leaders out of 19 and 3 gunners out of 6;  9 more casualties. In this one I wrote it down, but it is always surprising to see how AI squads were able to keep one or both of them; not at the end, but the first casualties.

I don't know if there is a different code for the AI, but, if there is, it seems to make a better use of "random"; even if three out of six, with a lot of riflemen around, looks like a huge coincidence.

 

But hey!, we got the graphics for the PzIII fixed.

[womble]

There isn't different code for the AI. What you're seeing is your very own confirmation bias off a small sample. Until you can provide data rather than anecdote, it's not even worth considering the possibility.

So set up some fire missions and do some stats on a few hundred teams, and maybe someone will buy what you're selling. Publish the scenario so other people can validate your results. It's how science works, ya know?

[Codreanu]

Not trying to be mean, and no game or company is above criticism, but I don't think I've ever seen you make a post that wasn't you complaining about the game. If you think there's something weird going on with the AI you could run some repeatable tests so there's actual data that the rest of the community can look at rather than think Battlefront is conspiring against you with how the AI works.

[Bulletpoint]

I also think it's selective memory at work here, but there MIGHT be something about leaders and gunners getting hit more often, if the game makes these troops more likely to stay up rather than to go prone.

That would mean they would be somewhat more likely to get spotted and hit or catch a splinter fragment. I don't know if the game works like that though.

[Brille]

In general I don't think that the ai gets a "buff" in survivability of certain troopers. 

However the information provided is far to little to jump to conclusions. 

What fired on the ai infantry exactly? How intense was the fire?

10 minutes of artillery fire can mean anything between 20 or 100 rounds, depending on the mission and sort of artillery. So the time provided is rather vague. 

How close were the impacts to the infantry ? 

In which positions were each particular soldiers? (prone, kneeling, standing)

 

If I reconsider my last 2 CMRT games (Head to head) one might come to the conclusion that I lost considerably more squad leaders in them than usual. Often they were the first or second in line to die. 

But I wouldn't complain about it, at least not before I did some testing myself (which I haven't, yet) . Maybe it is some fault of mine or just coincidence.

If the first don't apply and the second is far to often to be coincidence anymore then we should look at the game mechanics itself. But yeah that includes several test sessions.

Finding that something happens/seems to be wrong is the first step of the research, finding out why it happens should be the second and so on... 

 

 

[semmes]

You got the "impression", right.

 

A test, 10 or a 1.000 will give a sample, an approximation. The limited tests I have done confirm my impression, leaders and gunners take a disproportionate number of casualties; even when the code keeps the LMG behind to the right, where it cannot fire.

A total of 16 casualties, everything else is irrelevant because there were no special circumstances. For example, 18 units took casualties, is it relevant to know how many more units were there, who took no casualties?

We just had a patch, lorries are still charging forward. Someone forgot, again, to write the code to tell them to stop after Passengers Dismount. Yes, I do complaint.