Pseudo random distribution to critical strikes

  • Added pseudo random distribution to critical strikes and procs.

What this means is that everytime you attack and don’t trigger a critical strike, the chance of you triggering a critical strike increases slightly. This should help flatten the bell curve slightly when it comes to large periods of either no critical strikes happening, or all critical strikes triggering in a row.

Is there a hidden plot going on, which wants to promote Swift Slaying even more? What part of crits do Careers that dont have Talents, or passives affected by those have from more crit rate, apart from enhanced left click spamming and better SS uptime?

If they wanna hand us more crit, do it via timed blocks, heavys to the head, or something else that lets us play for them. Dont just promote left click spam and put SS into an even better spot. A Trait that hands us four properties worth of Attack Speed for a random event.

6 Likes

Can someone explain to me if this change is a buff or a nerf?

From what i understand this is a buff, and it will just distribute the crits more consistently? So it will be less ‘rng’ but not as fixed as 'every 5th strike is a critical hit?

If that’s the case, why not add fixed crit changes to Every weapon seperatly, like ‘every 10th strike crit’ on DD for example and ‘every 5th strike crit’ on 2h hammer. Wouldn’t that accomplish the same result, while balancing crits in the meantime?

To me it sounds like a massive buff especially if you have extremely low crit chance while mid/high crit will certainly benefit but not as massively.

3 Likes

I think it’s not a buff nor a nerf… it (if I have understood) just makes “more true” your %… In live you can have 4-5 crits in a row… like you can fight for more than a minute without seein a single crit… and this is frustrating (anyway, to me, every “RNG mechanics” is bad).

For example some hours ago I played as Huntsman (high crit chance career)… and I could not active Swift Slaying: each time the horde died before.

3 Likes

Strictly speaking, it’s neither a buff nor a nerf. More a… I hesitate to call it a quality of life feature, but I suppose that’ll do.

Statistically, over the course of even just a single game, PRNG shouldn’t (in theory) make a difference at all in terms of how many crits you get. In practical terms, it basically just means that crits will be less prone to “streakiness” - You’re less likely to hit a random dry spell without any crits, but you’re also less likely to get a bunch of crits all in a row.

This is all, of course, assuming that it’s been implemented correctly.

2 Likes

You would need to know the exact crit distribution to know how exactly it affects anything, but it should make you less likely to chain crits (chain stagger lock and blazing echo oof) and have better SS uptime on lower crit in that you would be less likely to have long dry streaks.

Here’s an example of what a 25% crit chance distribution could look like: ( Well it could look like anything but you normally choose the constants very carefully)
1 8.5%
2 17%
3 25.5%
4 34%
5 42.5%
etc…

2 Likes

Thank you.
Edit: actually why does it not start at 25% crit chance and ramp up? Or is that because they don’t want you to chain crits anymore, or be less likely to?

uhh. how could it start at 25% and ramp up unless you want more crits than 25%? the uniform random looks like
1 25%
2 25%
3 25%
4 25%
5 25%

2 Likes

Ah wait… nvm ignore my question. I should think before typing, your first comment was enough, thanks.

1 Like

This is a buff to Swift Slaying. Plain and simple.
It means that your average up-time on Swift Slaying is higher since the crits are forced to be more spread out.

1 Like

How are they forces to be spread out? If I don’t misunderstand the dev comment the following applies:

If you don’t crit on an attack you receive a small bonus to crit chance on your next attack. The crit chance increases with every non-crit attack you take.

When you crit on an attack your extra crit chance is reset. Your next non-crit attack after that receives a small bonus and on it goes.

For a high crit build the change wouldn’t be noticeable at all, except for when you’re on the odd non-crit streak. Result: slightly less rng, slight buff to all weapons where non-crit builds and low-crit weapons are the winners in regard to extra damage done.

Edit: The added consistency does obviously apply to swift slaying as well. Same goes, should be more noticeable on low crit weapons and low crit builds.

2 Likes

This is all true, it seems like a straight up buff to crit chance… If there isn’t also a mechanic that reduces your crit chance after scoring a crit… And the way the explanation was worded, that possibility can’t be excluded…

@Brytarn , any chance of a more in-depth explanation of the mechanism?

It’s been interesting watching this play out and there are still some things that could use answers.

The distribution itself is simple enough and it may be assumed that having crit chance on your gear and talents contributes toward the distribution, so you will get crits more often.
What I have yet to determine through play is how higher crit attacks on certain weapons, such as upward swing attacks, factor into the distribution.

Are they stand-alone? Does the large percentage get appended to the current PRD percentage? Does the PRD percentage continue from where the high crit attack left off or does it go back down to where it would be if it was a normal attack?

2 Likes

Because the way it was worded was “flattening the bell curve”.
The flatter the bell curve the more spread out the occurrences.

This isn’t exactly a buff to crits in general, since on a large scale they should still appear at the same frequency on average, but it is a buff to Swift Slaying since the standard deviation seems to be reduced.

Clarification: this post assumes they also reduce your crit chance the more crits you get in a small period of time, since in the patch they wrote that this change should also take care of multiple crits happening in a row.

Wondering very similar things. While we’re at it, do your melee and ranged attacks have independent crit distributions and do they interact, eg does not getting a crit for a number of melee strikes then changing to ranged mean a higher crit chance?

1 Like

Good shout. I didn’t even consider the relation between melee and ranged PRD.

1 Like

It is not only restricted to range/melee swaps. Imagine you use Executioner light attack 1&2, making a few swings without critting and as such stacking the crit mutliplier/correcter and then do a heavy attack with innate 20 % crit chance. Does the crit modifier transer in this case? If so, it could be abused to a certain degree.

The interaction between different actions and the crit modifier is something which would need some clarification.

Also, as it said flattening the bell curve, it seems to work in both directions. Meaning that consecutive critting will reduced the chance for further crits:

  • Scenario 1 (base crit 10 %, consecutive no critting): 10 - 11 - 12 - 13 - ….
  • Scenario 2 (base crit 10 %, consecutive critting ): 10 - 9 - 8 - 7 - ……

This would also need some clarifications. It mainly depends on the increments between each step.

1 Like

It’s most likely an implementation of the Pseudo-random distributed procs mod on the workshop by Zaphio. The workshop page has some answers to the principles behind PRD.

From the workshop mod:
" What is a pseudo-random distribution?

The pseudo-random distribution (often shortened to PRD) is a mechanic in which the event’s chance increases every time it does not occur, but is lower in the first place as compensation.
This results in the effects occurring more consistently while mantaining the same nominal probability .

This has the following implications:

  • There’s a limit on the maximum number of attempts without a proc. All procs benefit.

  • Getting two procs in a row is less likely. This has a negative effect on duping effects.

  • Procs are more evenly spread out. Traits such as Swift Slaying benefit from this greatly."

Math for nerds

The PRD is a markov chain with N states labelled from 1 to N.
Proc chance at state n is n*c, where c is a carefully chosen constant (see above).
Whenever there is a proc, the state resets back to 1. Otherwise it goes to n+1."

6 Likes

I have read this and also looked at two of the links. However, it doesnt give any information how the game handles attacks with different crit chances. Does the adjustment work over all attacks middled or is it adjusted per attack?

Also doesn’t give the increments which determine how different the PRD will be from the TRD.

1 Like

Yes, this is my understanding as well. I’ve programmed some trials, and the naive approach that naturally occurs from reading the patch notes of course skews the crit chance in positive direction. For example, here is a model of 1 million strikes with critical chance of 5% and the skew parameter is what gets added every consecutive non-critical:

As can be seen, the gap between critical strikes indeed gets a little more consistent with this naive approach, however choosingthe right starting chance multiplier and or power proves to be a chore and asks for an analytic approach instead.

I can dig a bit into the Markov chain analysis if it is indeed what has been implemented, but the caveat is that the “carefully chosen constant” must be carefully chosen for every possible crit chance we need.

1 Like
Why not join the Fatshark Discord https://discord.gg/K6gyMpu