Testing method: Start the game with “x86” and “x64” respectively, and only Cadence and Blitz were used to continuously attack Craining Dummy for a long time (lasting more than 1 minute), then stop attack. It was observed that the trauma damage of “x64” games was around 250000, while the trauma damage of “x86” games could reach over 500000.
Speculation: When starting the game with “x86”, the trauma damage of the deadline moment can be repeatedly stacked, resulting in an abnormally high final trauma damage.
i think it might have something todo with deadly momentum expiration and new timed attacks
atleast i could get same dmg on both x86 and x64, but i noticed i got more frequent dmg overlaps or “highest trauma” ticks onx86, which seemed to correspond somewhat to how/when blitz got applied in relation to deadly momentum triggering
TLDR:
basically i think if DM expires it’s not counted as “same” source as previous or next DM, and each actual new instance of DM is “new”/separate WD counted
and for whatever reason seems like this is easier to occur on x86(more lenient hit timings/doesn’t have to truly expire) but not impossible to force on x64
i went into DB and tried to mod the duration of deadly momentum,. and lowering it from 5 sec to 1sec, i “instantly” got the same higher dmg (and higher peak even), which seemed to be related to “new” deadly momentums triggering instead of just maintaining the same old/constant buff
similar peak dmg ticks, just much easier to reach compared to x64 5sec DM manual timings
maybe something on x86 is handling the timer different/in a way where it’s easier for “fresh” DeadlyMomentums to happen and be counted as new similar to having DM being 1 sec duration on x64 ?
As a part-time programmer, im gonna guess it has to do with the data structure. Could it be the difference between the value of say, an integer type of value, whose limit would be twice as large in a 64-bit computer than a 32-bit computer.
Just stabbing in the dark…
There is a high probability that you are right, maybe in x86, DM is not counted as same source as previous or next DM, even if DM is not expires. Therefore, the behavior of DM in x86 is very similar to that of extremely short DM duration.
obviously Crate knew about these all along, and probably has like a list of "top 500 bugs we left in the game for players to find", so there would be continued engagement how else would they have fixes for them so fast, everytime ?!
*note to self; running low on tinfoil, order more, in bulk
In addition, regarding Cadence, after testing, it was found that “reducing ICD from 0.5 to 0.35” in 1.2.1 seemed backfired, resulting in a significantly lower proportion of Cadence in continuous attacks.
The Cadence timer changes in v1.2.1 happened at the request of the playtesters, and we saw no harm in it. As Cadence is in a good place right now, no further tinkering is necessary.
