20 yard range as in "Ignore trash further away than"?I'm doing GR80 using Fisher's "Power" setup and getting pretty good results:
Code:[Adventurer][Rift] Runs count: 12 [Adventurer][Rift] This run time: 0:11:43 [Adventurer][Rift] Average run time: 0:11:40 [Adventurer][Rift] Best run time: 0:09:49 [Adventurer][Rift] Worse run time: 0:14:39 [Adventurer][Rift] This run XP Gained: 31,115,920,873 [Adventurer][Rift] This run / Hour: 159,304,041,272 [Adventurer][Rift] Total XP Gained: 398,637,721,515 [Adventurer][Rift] Total XP / Hour: 170,695,770,882 [Adventurer][Rift] Best XP / Hour (single run): 211,422,505,590 [Adventurer][Rift] Worse XP / Hour (single run): 134,258,854,500 [Adventurer][Rift] Best XP / Single run: 35,670,106,742 [Adventurer][Rift] Worse XP / Single run: 28,939,592,185
Trash pack size 8, 20 yard range.
Turbohud stats:
- 183 billion xp/hr
- 47.8 legs/hr
- 5.7 deaths/hr
Actually my gear is mediocre and I'm only around paragon 1100. My stats are 19500 str, 30% elite, 100% AD, 63% CDR, 55% Phys, 30% Bomb, 300m toughness sitting in town, 240k thorns, all ancients, all res around 1300
In case you don't believe me how good my setup is, here are three back-to-back Grift 75 runs in a row:
![]()
![]()
![]()
The Crusader is already amazingly efficient (awesome job, by the way!), but I imagine this is something of increased importance to other classes. If I had to give you an ideal picture, I would want to emulate how a human would decide to stop:
- Combat trigger logic stays the same with the ability to choose either # or % within a chosen radius (I agree with Jubisman that % should have more priority than #)
- Once triggered, the bot would keep the combat cycle active on that radius until it reduced its #/% to a user-designated minimum setting like 3/2%, then move on
- Additionally (to prevent the stucks you were referring to), include a user-designated trash combat timer to prevent wasting too much time on a trash pack -- so if the trash pack isn't reduced to the designated minimums by X time, then ignore all trash for 3-7 seconds, which will force the bot to path away and find a new cluster to focus on (similar to what humans do -- if the trash pack doesn't die efficiently enough, they move on)
If anyone else has any suggestions to enhance this logic, please chip in!
That's pretty good, well done.![]()
to make stricken more effective than powerful you need mostly manual playing...for bots powerful is better i guessWhy do you recommend Powerful over Stricken? From what I remember, powerful is additive (with skill dmg or such?), while the stricken passive is multiplicative. As damage on trash is irrelevant, Stricken should be much better,
First of all I wanted to say thank you, all the guides you've posted have been extremely professional and helpful. I did have a couple questions though; originally I had my "manually set" trinity settings based upon your guide and was clearing 60 Grifts easily in about 5 minutes. When I decided to use your SVN imported settings for some reason the efficiency has greatly diminished. I found when watching the bot it will occasionally "skip" elites, it will engage in combat yet simply not finish the kill, usually leaving said elite at ~33% health. This abnormality doesn't appear to occur when engaging blue elites, only gold - I confirmed that "always kill elites under 100% is indeed checked. Any idea what might be causing this? It appears to happen about 25% of the time. It also appears to not efficiently collect progress globes, which has been identified as an issue prior but I thought was somewhat fixed? Thank you once again sir.
Why do you recommend Powerful over Stricken? From what I remember, powerful is additive (with skill dmg or such?), while the stricken passive is multiplicative. As damage on trash is irrelevant, Stricken should be much better,
Check the TPS setting. Latest settings I downloaded here limited the TPS to 15, no idea why. That might fix itFirst of all I wanted to say thank you, all the guides you've posted have been extremely professional and helpful. I did have a couple questions though; originally I had my "manually set" trinity settings based upon your guide and was clearing 60 Grifts easily in about 5 minutes. When I decided to use your SVN imported settings for some reason the efficiency has greatly diminished. I found when watching the bot it will occasionally "skip" elites, it will engage in combat yet simply not finish the kill, usually leaving said elite at ~33% health. This abnormality doesn't appear to occur when engaging blue elites, only gold - I confirmed that "always kill elites under 100% is indeed checked. Any idea what might be causing this? It appears to happen about 25% of the time. It also appears to not efficiently collect progress globes, which has been identified as an issue prior but I thought was somewhat fixed? Thank you once again sir.
[Adventurer][Rift] Worse XP / Single run: 851,933,846Stats running GR75 + Torment X :
[Adventurer][Rift] Runs count: 15
[Adventurer][Rift] This run time: 0:14:46
[Adventurer][Rift] Average run time: 0:09:04
[Adventurer][Rift] Best run time: 0:04:39
[Adventurer][Rift] Worse run time: 0:14:46
[Adventurer][Rift] This run XP Gained: 23,007,229,123
[Adventurer][Rift] This run / Hour: 93,428,047,986
[Adventurer][Rift] Total XP Gained: 203,630,902,125
[Adventurer][Rift] Total XP / Hour: 89,761,058,585
[Adventurer][Rift] Best XP / Hour (single run): 144,532,057,444
[Adventurer][Rift] Worse XP / Hour (single run): 8,310,920,298
[Adventurer][Rift] Best XP / Single run: 27,630,936,208
[Adventurer][Rift] Worse XP / Single run: 851,933,846
Paragon : 1000
Unless the bot gets stuck in corners/objects or doesn't use portal/doors. Then yes.Great guide by the way! Horribly unoptimized gear crusader and I'm still running GR65 in an 8min average. I despise the manual playstyle of this build but it is definitely well handled with this bot.![]()
where do i put these SVN folders this is where im lost
Alright guys, special thanks to Jubisman -- I think the elite skipping bug is mostly solved. I've updated my trinity settings file in the OP. After about half an hour of testing, it only skipped 2 elites. I'm gonna go ahead and call that "good enough."
I'd like some additional testing from you all though, so please let me know how it goes. You will notice the bot attacks trash a bit more than before, but it's actually enhanced the clearspeed in my testing, so I'll be leaving that in there for now.
Also, make sure you are using the latest version of SVN or this won't work.
Thanks for the update! Could you share your finding and what you modified? Might be helpful for testing purposes so we can know what variables to play with.