I wrote down (well, typed in) the Attack, as given by Manuel, each time.
I manually added that to the running total - but it agreed precisely with what it said when I used the flyer.
flyeredML - potential rounding problem
- Thread starter Magus_Prime
- Start date
I manually added that to the running total - but it agreed precisely with what it said when I used the flyer.
The run I just started gives my Hot. That means that I missed Cold 15 times - but I will have every other Tower Contest factoid when I get The Lavalier for the third time in a few days.
If each element is equally likely - a flat 1 in 5 chance - there is a 3.5% chance of not getting an element in 15 attempts.
Bale
Minion
Happy dance! Happy dance!
It took me 15 ascensions, but my telescope tells me that I'm finally going to be fighting Stinky adventurers when I reach the tower of that naughty sorceress!
Hopefully you'll find Cold on your 16 try.
Magus_Prime
Well-known member
This time everything worked as expected. After every combat I performed a "get flyeredML" in the gCLI and it matched exactly with the reported percentage in the relay browser window.
The major difference this run is that the character is a Sauceror with a Saucepanic equipped.
Last run, when something went wrong, the character was a Pastamancer with a Hand that rocks the Ladle equipped.
The ML mismatch isn't something that happened on every run.
The major difference this run is that the character is a Sauceror with a Saucepanic equipped.
Last run, when something went wrong, the character was a Pastamancer with a Hand that rocks the Ladle equipped.
The ML mismatch isn't something that happened on every run.
As Mafia adds the information to the relay browser, that is no surprise.
Bale
Minion
I cannot image how that could be even slightly relevant, but if you tested and it did make a difference you'd have a real discovery.
Magus_Prime
Well-known member
It happened again. The value of flyeredML is 10090 but the I'm unable to complete the quest. This was as an Accordion Thief in a soft-core standard ascension. Since the Monster Manual factoids are complete for all monsters in a normal run that should mean that the value of flyeredML is accurate.
One more combat brought flyeredML to 10209 and I was able to complete the quest.
I'm open to ideas on how to track this down.
One more combat brought flyeredML to 10209 and I was able to complete the quest.
I'm open to ideas on how to track this down.
Last edited:
To be completely clear - you are saying that THIS CHARACTER has a Monster Manual and has gotten at least one factoid for every monster you visited?
How to track it down? Your session log has every single "flyer" event you performed. They look like this:
Code:
Round 1: Veracity uses the rock band flyers!Magus_Prime
Well-known member
Magus_Prime
Well-known member
No. Nor any other effects or items that are known to mess with parsing combat results.
OK. I am trying to figure out if there is any possible way that you used flyers twice in the same battle and we incremented flyeredML twice. For example, I frequently double throw flyers + a molybdenum magnet against gremlins. The first time, it says "You slap a flyer up on your opponent. It enrages it." and we increment the setting. The second time it says "You already advertised here" and we don't.
Now, under the effects of anapests, I don't know what it reports in the second case, but it might very well get double counted.
I'll see about putting a message into the session log whenever we increment the setting. It may not look nice, since we don't increment the setting in the same place that we log combat actions; we do that AFTER we do the session logging. But if we, somehow, do it twice in the same fight, we should be able to see that happening.
Now, under the effects of anapests, I don't know what it reports in the second case, but it might very well get double counted.
I'll see about putting a message into the session log whenever we increment the setting. It may not look nice, since we don't increment the setting in the same place that we log combat actions; we do that AFTER we do the session logging. But if we, somehow, do it twice in the same fight, we should be able to see that happening.
OK, I think I understand how this could happen - but only if you have Manuel. 
If you DON'T have Manuel, the calculation is this:
MonsterStatusTracker.originalAttack + MonsterStatusTracker.attackModifier - MonsterStatusTracker.attackModifier.
If you DO have Manuel, the calculation is this:
MonsterStatusTracker.attackManuel - MonsterStatusTracker.attackModifier;
If you do not delevel the monster, that is the equivalent, since Manuel reports the monsters adjusted attack at the beginning of the round. But if you DO delevel it, that the attackModifier is a larger (negative) number and we give the monster too much attack.
I broke this when I "simplified" the calculation recently. I'll figure out what to do.
If you DON'T have Manuel, the calculation is this:
MonsterStatusTracker.originalAttack + MonsterStatusTracker.attackModifier - MonsterStatusTracker.attackModifier.
If you DO have Manuel, the calculation is this:
MonsterStatusTracker.attackManuel - MonsterStatusTracker.attackModifier;
If you do not delevel the monster, that is the equivalent, since Manuel reports the monsters adjusted attack at the beginning of the round. But if you DO delevel it, that the attackModifier is a larger (negative) number and we give the monster too much attack.
I broke this when I "simplified" the calculation recently. I'll figure out what to do.
Magus_Prime
Well-known member
All the flyer combats this run were in one of three places: Oasis, Desert, or Pyramid.
All combats were scripted, via Batbrain, SmartStasis, and WHAM.
I've never noticed SmartStasis using flyers more than once in a combat but I suppose it's possible.
Edit: You posted a new theory while I wrote the above. What you propose could easily happen. SmartStasis and WHAM frequently delevel after the start of combat.
All combats were scripted, via Batbrain, SmartStasis, and WHAM.
I've never noticed SmartStasis using flyers more than once in a combat but I suppose it's possible.
Edit: You posted a new theory while I wrote the above. What you propose could easily happen. SmartStasis and WHAM frequently delevel after the start of combat.
Last edited:
Magus_Prime
Well-known member
It's going to be a while before I can test this. I just ascended into a hard-core "Ed" run.
Now I'm wondering how KoL handles flyer progress as Ed if you delevel, then die, then come back and flyer again. Is it the starting value from the current encounter, or from the first encounter?
Magus_Prime
Well-known member
Same here. Thank you Veracity!
xKiv
Active member
Something else must have been at play too, though, because I have manuel, always delevel in most fights, and usually go back to arena when mafia reports 100%, and never have this problem. I don't fully automate my fights, though (playing in relay browser, button-clicking on native macro).