Hit Recovery and Block Speed Break Points

[lemekim]

It's great to see someone actually put the info together on the FHR and FBR info =) When the older info came out, we looked at it, but it was quickly evident that the tables didn't actually match the data, but people apparently didn't care =\ So we tried to put together some actual test data with break values for FBR and FHR, and I guess we were at least on a right track, for it seems your FBR tables match whatever info we managed to collect back then.

Still, I was wondering if you got a chance to test FHR numbers yet; and if you did, what method did you use? FHR is somewhat hard to measure, I wanted to see what "pros" were using :lol: From what we have seen though in use, and in actual test that we performed, the FHR breaks seem to come less often then your tables indicate; but the breakpoints do match it seems. For example, after some testing we were able to approximate that necromancers have FHR breakpoint at 26, and 86 FHR. But it seems between them there is no breaks; or perhaps the change is so subtle, that there is almost no difference between, say, 35 and 75 FHR.

I am not sure why such discrepancy... For testing we used an assassin with Mindblast + Wake of Fire to provide constant stun + hit recovery animation for the subject, and simply counted the number of stuns versus a significant amount of time. While this might not be a good way to find the actual FHR frames, it is a good way to establish that a break actually exists in that area. We repeated tests to ensure that the numbers we were getting are correct. Perhaps something in the test routine? Or is something else happening?

Thanks,
lem.

[LavCat]

I am confused as usual. I just got around to reading this thread. Last winter I wrote down a table of FHR breakpoints, based on a thread that was current at that time. Not sure if this was the older information of which you speak. I've been searching and the only thread I could find was:

http://www.lurkerlounge.com/forums/index.p...very,and,frames

This was not what I was thinking of, though, although I had added the act II mercenery numbers from this thread. (I sure wish the forum search would accept "FHR" as a search term.)

The thing is, the data in Tommi's tables is far different, not even close. The number of frames is much higher in the newer data. For example the sorceress had 8 frames at 0%. I balanced the HR of several characters based on the older numbers, and now I'm wondering if the older data is just wrong or if it applied to an older patch? I wish I could be more exact, but I can't find the post that gave the source of the old data that I've been using.

Any clarification would be appreciated.

[Tommi]

Let me clarify:

The incorrect tables were derived by using the weapon attack rate formula as follows.
- Base = Base frames for hit recovery (known)
- Speed Increase = EFHR = [(120 * FHR)/(120 + FHR)]
THIS IS THE INCORRECT FORMULA

(Several people - lemekim at the LL, Tao and colds at the AB, among others - noted that real data did not match these. Well, nobody knew anything better.)

Some time ago Karon/Rivo (the same guy, different aliases) conducted extensive tests (I do not know details), through which he derived empirical data for FHR and FBR breakpoints. With the help of Hammerman, we were able to find out some theoretical implications (from the game code), for example for amazons blocking with swinging weapons. I compiled the page from the posts for easier reference (as I usually do, good info should be stored). I did not do any tests, just referred to Karon's/Rivo's results and Hammerman's info. So:

Karon/Rivo --> The guy who made the tests
Hammerman --> The guy who read the game code
Tommi --> The guy who wrote this all down from the forum posts and asked the above two guys for all missing information

These tests and work lead to formulas close to previous ones, BUT with some important differences. First, the basis before Speed Increase was 50 normally, 100 only with Holy Shield. Second, there was no +1 before Base. Animation speed was not 256 in all cases (not visible in the weapon speed formula). See formulas in my page.

Lemekim, I suggest that you correspond with Rivo/Karon (Kar0n at The LL, Rivo at the AB, Karon at dii.net) about the tests. Some breakpoints based on Hammerman's data are not tested, especially those where Animation Speed differs, I believe. Perhaps you are able to find explanations to your observations, or further elaborate the FHR/FBR info. I'll update the page whenever new reliable data is acquired.

LavCat, if you look at the thread you posted and read the quote from me (in ShadowHM's post), it is apparent that the tables are purely theoretical and untested. They were derived using the same incorrect method as I described at the beginning of this post.

Cheers,

Tommi

[LavCat]

Let me clarify:

And the clarification is most appreciated! As well as all the work that went into it.

[lemekim]

Thanks for clarification Tommi =) I will definately talk to Rivo/Karon, perhaps make some heads or tails out of it. This new info really clarifies a lot of things.

Btw colds on AB is me ;)