The Transformers Forged to Fight community has officially moved to Discord. For all the latest news and updates, join us here!
Options
Raid AI difficulty scaling with medals?
Qliver
Posts: 54
Does the raid difficulty scale with medals? Recently I've noticed that the AI is insane, I am often losing one bot per fight. And that's the regular overpowered Grindor with HA, megatron with LG, motormaster with EF we're talking about. By that I mean these are bots that are on every base so I definitely notice the difference. They even do impossible moves like blocking after I dodge their charge, since when is that possible? Also it's suddenly very hard to bait their specials. But in general they seem super evasive and twitchy.
Tagged:
0
Comments
Get better and you will die less. Seriously.
Enough of these posts crying about not being able to win 100% of everything all the time.
If you're losing against strong bases, you'll end up dropping to where you belong. That's the whole point of the raid system.
Dude I am at 2100 medals now and my best rank has been in the top 20, so I am not new to this game and I am also not crying. I didn't ask about the bot rating getting higher, I asked about the AI. I am only using 3* bots and I am definitely gaining more medals than I am losing (never used energon) so your argument about medals is invalid. Also I can't say that I have noticed any difference in the bases as far as the ratings go, they are usually around 5000-10000 regardless if I drop some medals or stay at 2000+ medals.
You're contradicting your original post, so I'm gonna have to break it down.
"I'm often losing 1 bot per fight"
This is against strong bots with strong mods vs, as you say, your 3* bots. Small mistakes against stronger bot/mod combinations will make you pay dearly, so yes you will lose bots when facing those 10k bases.
"They regularly do impossible things like block after they charge"
Although I somewhat agree, this is actually not impossible, and you can often do the same thing vs the AI if you perfectly time your block after your dash animation frame. It's not easy to do but it is possible. Just don't expect 100% success.
"In general they seem super evasive and twitchy."
Yup, higher power bot/mod combinations have faster reaction times, no mystery there and it's also expected.
So you've validated almost all my major points, but let's break it down further vs a semi-typical 10k ish base, with easier mods than what you're describing:
So, let's do the math here. Subtract any two side node bots in each section
Total avg power: ~7000pi
If you are usually taking 3x 3* bots with you, even if they are 4/40 unforged, you are only bringing a squad with around 55-60% of the total power of the bots you are facing.
Yeah, it's gonna seem lopsided if you lose, especially if you lose the bot you are the best with.
There is a huge difference in hp and attack rating between 3* and 4* bots, even if your 3* bot is fully forged. The bot pi doesn't tell the whole story. Compare baseline stats for bots in your roster and you'll see what I mean. While 3* bots have better abilities at rank 4 giving them the illusion of higher pi, they still have inferior base stats.
Know what situations they will thrive in, and die in.
Most 10k bases are using 3* mods or full Sig 2* mods. These will either (3* mods) add a huge chunk of HP and attack rating or (2*mods)massive ability bonuses to the bots you are fighting.
Against HA, attack rating is everything when it comes to DoT effects. You don't have much time to take it down, and any mistake will compound the problem. You need to fight that node with a very strong bot to reduce the damage taken by HA. An r3 (even an r2) 4* bot will take far less damage than any 4/40 3* bot from HA because they will take less base melee damage from the bot on the node. Damage is relative there.
The inverse goes for high Sig LGM nodes. One mistake will make it feel like the AI is Teflon when it really isn't. You take big damage from small mistakes. Sorry but, you're just fighting impatiently for the situation.
Exo filter is insane, you need to use a bot that self buffs instead of debuffs or you will trigger purify/unstoppable a thousand times, again making it feel like the AI is way better than it actually is. Use the wrong bot and you have to play defense too much... you'll often get cornered and violated.
All these things together make it feel like "something is wrong" when it's just how players are perceiving it.
I'm honestly trying to shed some light on the situation for you. Do what you want with the information.
With that in mind, as I mentioned before, there are certain situations where skill can overcome the difficulty, and others where you're more likely to lose due to manadatory damage taken.
Playing cautiously and recognizing AI patterns will go a long way to minimize the damage, but occasionally the game will simply decide that you're going to take damage.
I'm not being flippant with this comment. Timing your attacks so the AI is on the back foot is how you create openings. And timing your attacks so you aren't left off balance is how you prevent the AI walking all over you
It seems reasonable that there could be one of the ai thresholds around 2k where Qliver is and because Manthro's medal count is so low in comparison he wouldn't know about it.
But then I also feel like the ai is different when the bot is max level. Before all the updates, there was a time when I remember max 2 star bots having much harder ai than a 3 star rank 3 level 30. Maybe I was extra lazy against the 2* bases because there was no chance to lose, but it always seemed like I ended up taking more damage and the ai being far more aggressive/defensive.
I haven't really paid much attention as of late, all of the ai was buffed in a recent patch anyway so now I just roll with whatever the game throws at me.
Did the ai feel different? Try dropping to 1k, for science
Off-topic, is it worth saving 2* mods and max their sig ability? I usually forge them into my 3* mods, and only keep 3* or 4* mods.
i know that before they shortened the time on motormaster's unstoppable, I was able to do some pretty crazy things with it. So I won't say it's impossible for a human to do this also. I will say that if a human can use unstoppable to escape a combo or start blocking mid special attack, the window of opportunity must be very small.
"An r3 (even an r2) 4* bot will take far less damage than any 4/40 3* bot from HA because they will take less base melee damage from the bot on the node. Damage is relative there." -- does that mean a r2 4* bot, being at about the same PI with a maxed 3*, has better armor and resistance? Again, my question about how PI is calculated in this post.
For example, my MV OP against IH in the Arena. I throw out a punch right after the round starts. He didn't get hit, didn't dodge or backstep. Instead, he immediately punches me back in the face like he's got unstoppable (whereas we're in Arena), ouch.
Yes. When looking at the same raw PI 3* vs 4*, the 4* will have higher armor, resistance, greater attack rating and more hp.
In short, 4* bots have higher base stats when comparing near identical PI.
Bleed (or any damage over time effect) is based upon how much damage your bot would take from base melee damage. It says so in the description.
Therefore, the 4* bot will take less overall passive damage from HA than a 3* bot.
Having said that, a forge lvl 100 3* bot will take less damage than a unforged 4* r2 bot, however... the same forge 100 3* bot will take more damage than an r3 unforged 4*.
Prowl has also been forged a bit and barely has the same amount of HP:
Armor is one of those mystery things yeah? We don't know exactly how it works from bot to bot and from star to star and rank to rank, hidden stats yeah?