i feel like this video is a prime example of why ravenfield needs a bot rework/overhaul bots in general just feel super random and i still see the majority of fights being just whole ass groups passing each other by without trying to kill each other the bot ai is just dumb. would honestly be interesting to see you do a video on the bot related mods maybe we can finally fix the bots by tweaking this mod and maybe using others like Better Bot Targeting, Replace Gunner for All Vehicles, and maybe even the "Vanilla Styled Classes again..." mod would be cool to have like a group effort to see what mods would be best or not good to use to have RF bots finally get smart.
The "passing through eachother" seems to be the AI throttling to avoid your computer from devolving into a 1 frame per 5 seconds powerpoint slideshow. Someone did a test by disabling bot sight and it increased performance, obviously... after all, making 300 bots see eachother without killing framerate is one of the biggest challenges you can give yourself. But yeah, plenty of these "better bots" mods should be added to vanilla, at least to give new players quick access to them.
@@migueeeelet see the problem is their is years apon years of video games that just have better ai than what is currently in ravenfield and its not like we are just at the start of the internet so theirs tons of resources out their for making good ai i really see no excuse besides the fact mans is a solo dev which that bit i can understand but not the ai being just serviceable.
@@tumble8323 you shouldnt NEED to use lower bot counts is the issue like remind you this is a battlefield-ish game people didnt come to this game to do 10 v 10s people want big battles like 100 v 100 heck even the defaults of like what is it 60 v 60 or whatever still has bots being dumb.
this is very interesting! i believe the reason more skilled bots destroy more vehicles is that all bots "evaluate shot difficulty" when shooting. weapons and vehicle weapons have different difficulty levels (auto, easy, challenging, hard, and impossible) depending on the target type (infantry, infantry groups, ground vehicles, helicopters, and planes). more skilled bots are likely taking the more difficult shots more often, though i'm not entirely sure how it works. the fact that more skilled bots take down more vehicles but not other infantry may be because modders usually set the difficulty against infantry and infantry groups to "easy" (or use the "auto" difficulty, which defaults to easy against infantry, among other targets). amazing work, and thank you for your dedication! i also think i can make more skilled bots target and take down infantry more effectively by tweaking my better bot targeting mutator :0
I think the reason nolberg got similar results in Kills is because Harder bots may be harder to kill. Maybe they position themselves better. What if the eagle bots were veteran level and the raven were beginner for example? Metrics we can use also dependent on environment. I agree with the reaction time statistics however the matchmaking data should be reissued, if there is something I dont hear from video of course, maybe he did but I didnt give attention at the moment. Edit: In skirmish mode, if you die, you see that some bots have equipped with just bad weapons, maybe the skill_level config also changes that spec.
The bit about having one team being entirely one skill level is shown in the video, near the end. Nolberg ran results on Island where the Eagles all have the Veteran skill level while the Ravens were all Beginners. The result was that the Eagles won significantly more than the Ravens, about nearly 97% of the time. When the skill levels were flipped (Eagles were Beginners, Ravens were Veterans), the Eagles still won but not by much over the Ravens. They were about nearly 50-50 with the score. Keep in mind that this all takes place on Island, which is always in favor of Eagle than Raven.
I think there are a lot of factors that change these results. If you were testing the beginner bots vs veteran bots, thing like: weapon, map, if they were using a vehicle or not, what team was winning have to be taken into account. You can't have real results if those factors are unbalanced. Also, bot count matters a lot. Because if veteran bots usually focus on using vehicles and there are 120 enemy bots on a relatively small map they will allways get shot down without significant kills. Every ravenfield game past 100 bots are a stalemate because bots don't take their lives into account, they only move to predefined positions and that can affect a lot their capacity at making progress
Since the new skill tweaker can improve accuracy of bots does that mean I can just unsubscribe to the stormtrooper aim mod that I have been using? Also what does FOV check mean from the mod?
I believe so, though they could be slightly different in how they tweak spread specifically. That said, the accuracy spread multiplier definitely has an effect. As for fov check, it essentially makes it so a bot can see in all directions and is no longer limited to just front vision
I would say Wiki needs more editors that can fact-check most of details inside the game, but consider RavenScript secrecy - it's not surprising that Wiki is inaccurate at all. Only sheer tests with sweat and blood like this video can be used as factual information, but we still can miss something because of very same RavenScript secrecy on how it works.
@@Bean_guy2 Duh Its What Centuries Of Of Intermingling With Eachother Does To People And Thier Culture TlDr: I Was Trying To Check If His Like Those Nasty Ultra-Nationalist You Know The Kind That Revisons But Deep Down Are
I'm not sure if this has been brought up before, but what if bot skills are only noticeable against the player? Like If a beginner and veteran bot takes 7 shots to hit an enemy bot, but a beginner would take 7 shots to hit the player while the veteran would take 4?
I'm assuming that the player is also counted as a bot in their field of view. The same reaction time and TTK it would take for the player is about the same it would take for any other bot.
For 6 USD, I would say Ravenfield is definitely worth it! Id recommend looking at the price chart from steamDB for your country. The current sale for me looks to be pretty solid
I love when nolberg fills me with information and entertainment
🤤🤤🤤🤤
🤨📸
i feel like this video is a prime example of why ravenfield needs a bot rework/overhaul bots in general just feel super random and i still see the majority of fights being just whole ass groups passing each other by without trying to kill each other the bot ai is just dumb. would honestly be interesting to see you do a video on the bot related mods maybe we can finally fix the bots by tweaking this mod and maybe using others like Better Bot Targeting, Replace Gunner for All Vehicles, and maybe even the "Vanilla Styled Classes again..." mod would be cool to have like a group effort to see what mods would be best or not good to use to have RF bots finally get smart.
The "passing through eachother" seems to be the AI throttling to avoid your computer from devolving into a 1 frame per 5 seconds powerpoint slideshow. Someone did a test by disabling bot sight and it increased performance, obviously... after all, making 300 bots see eachother without killing framerate is one of the biggest challenges you can give yourself.
But yeah, plenty of these "better bots" mods should be added to vanilla, at least to give new players quick access to them.
@@migueeeelet see the problem is their is years apon years of video games that just have better ai than what is currently in ravenfield and its not like we are just at the start of the internet so theirs tons of resources out their for making good ai i really see no excuse besides the fact mans is a solo dev which that bit i can understand but not the ai being just serviceable.
How many bots are you using? They won’t pass by each-other much if you have lower bot counts.
@@tumble8323 you shouldnt NEED to use lower bot counts is the issue like remind you this is a battlefield-ish game people didnt come to this game to do 10 v 10s people want big battles like 100 v 100 heck even the defaults of like what is it 60 v 60 or whatever still has bots being dumb.
At this point, Rising Front has better AI than Ravenfield even though there isn't any vehicle options.
this is very interesting! i believe the reason more skilled bots destroy more vehicles is that all bots "evaluate shot difficulty" when shooting. weapons and vehicle weapons have different difficulty levels (auto, easy, challenging, hard, and impossible) depending on the target type (infantry, infantry groups, ground vehicles, helicopters, and planes). more skilled bots are likely taking the more difficult shots more often, though i'm not entirely sure how it works. the fact that more skilled bots take down more vehicles but not other infantry may be because modders usually set the difficulty against infantry and infantry groups to "easy" (or use the "auto" difficulty, which defaults to easy against infantry, among other targets). amazing work, and thank you for your dedication!
i also think i can make more skilled bots target and take down infantry more effectively by tweaking my better bot targeting mutator :0
Thanks for this I’ve always wanted a comprehensive explanation for what Bot Skill Tweaker does
I think the reason nolberg got similar results in Kills is because Harder bots may be harder to kill. Maybe they position themselves better. What if the eagle bots were veteran level and the raven were beginner for example? Metrics we can use also dependent on environment. I agree with the reaction time statistics however the matchmaking data should be reissued, if there is something I dont hear from video of course, maybe he did but I didnt give attention at the moment.
Edit: In skirmish mode, if you die, you see that some bots have equipped with just bad weapons, maybe the skill_level config also changes that spec.
The bit about having one team being entirely one skill level is shown in the video, near the end. Nolberg ran results on Island where the Eagles all have the Veteran skill level while the Ravens were all Beginners. The result was that the Eagles won significantly more than the Ravens, about nearly 97% of the time.
When the skill levels were flipped (Eagles were Beginners, Ravens were Veterans), the Eagles still won but not by much over the Ravens. They were about nearly 50-50 with the score.
Keep in mind that this all takes place on Island, which is always in favor of Eagle than Raven.
Here's to the good times😂😂😅, the bad times😢, and the don't tell dad times!🤣🤣😁😳
I always noticed only certain bots used their tools like grenades, anti armor and secondary weapons. Now i know
that was so incredibly fascinating
absolutely wild to think that did guy can probably work in silicon valley if he wanted but chooses to grace us with ravenfield content instead
he finally cracked the code
As close as I will get😂
I think there are a lot of factors that change these results. If you were testing the beginner bots vs veteran bots, thing like: weapon, map, if they were using a vehicle or not, what team was winning have to be taken into account. You can't have real results if those factors are unbalanced. Also, bot count matters a lot. Because if veteran bots usually focus on using vehicles and there are 120 enemy bots on a relatively small map they will allways get shot down without significant kills. Every ravenfield game past 100 bots are a stalemate because bots don't take their lives into account, they only move to predefined positions and that can affect a lot their capacity at making progress
Respect to the modders fr
9:23 bot went flying
another great video from Nolberg
Even elite bots still failed Drivers Ed
Since the new skill tweaker can improve accuracy of bots does that mean I can just unsubscribe to the stormtrooper aim mod that I have been using? Also what does FOV check mean from the mod?
I believe so, though they could be slightly different in how they tweak spread specifically. That said, the accuracy spread multiplier definitely has an effect.
As for fov check, it essentially makes it so a bot can see in all directions and is no longer limited to just front vision
@@nicolala7132 isn't Stormtrooper aim mutator makes enemy bots completely useless?
I would say Wiki needs more editors that can fact-check most of details inside the game, but consider RavenScript secrecy - it's not surprising that Wiki is inaccurate at all. Only sheer tests with sweat and blood like this video can be used as factual information, but we still can miss something because of very same RavenScript secrecy on how it works.
I Dont Want To Be Condensending But Shouldnt Your Name Be Volod Instead Of Uhhh Vlad
@Alfonse-dm6ht I'm Vladislav, not Volodimir, so Vlad.
@@Alfonse-dm6ht You do realise ukranians can be called vlad as well right
@@Bean_guy2 Duh Its What Centuries Of Of Intermingling With Eachother Does To People And Thier Culture
TlDr: I Was Trying To Check If His Like Those Nasty Ultra-Nationalist You Know The Kind That Revisons But Deep Down Are
@@Bean_guy2 Yeah Yeah
How about you make a video on best ww2 mods
And what mods did you use
what is that sick desert map at 1:55?!
Sorry for the late response, but its Helmand:
steamcommunity.com/sharedfiles/filedetails/?id=3368466198
I'm not sure if this has been brought up before, but what if bot skills are only noticeable against the player? Like If a beginner and veteran bot takes 7 shots to hit an enemy bot, but a beginner would take 7 shots to hit the player while the veteran would take 4?
I'm assuming that the player is also counted as a bot in their field of view. The same reaction time and TTK it would take for the player is about the same it would take for any other bot.
Great video
Im suprised RF isn't open source. Looking at how it has modding at it's core
Its almost impossible to make an paid open source game
@@KronovMpg What about Source 2 \ GoldSrc games? Team Fortress and CS were born from those engines wich were made for Half-life games
What desert map and grenade launch do you use in the game play
Hey dude ravenfield become 6 dolar in here which is 200 if you convert to my countrys currency should i buy or wait till it gets cheaper
For 6 USD, I would say Ravenfield is definitely worth it! Id recommend looking at the price chart from steamDB for your country. The current sale for me looks to be pretty solid
Nice vid
Cool
based
oorah
12min ago is very fwuffy uwu
does it work on the crack version of ravenfield?
No, it's a workshop mod
Yeah it works, just use it but it doesn't really show any difference
@@ZyraX-21 one word: topmods
Hello