Blue wrote:I did not say it is working fine, I said I'm not aware of any problems of parry. And honestly parry is not block.

I'm saying parry works fine, there's absolutely no problems with it and parry basically is block. On the defender's side both are defined by the exact same parameters (dexterity and spec) with the same growth in block or parry chance, the only difference is that block has a 5% chance as soon as you equip a shield to make it possible for supporters to block. Your opponent's weaponskill or weapon type (2h halving parry, dw halving block etc) comes after that.

I don't get it how you can get parry to function properly but cause blocking to underperform with the same formula+5%. Again, obviously not an error with the formula itself.

Disdain wrote:I don't get it how you can get parry to function properly but cause blocking to underperform with the same formula+5%. Again, obviously not an error with the formula itself.

Blocking will not scale enough. This was the case for the old formula. Or do you get parry rates of 80% and up? I don't even want to talk about taking parry formula as replacement for block formula. Then we can take the old official known block formula and are in same state as before. We run in circles. Thats why I want strong test results and a CORRECT formula. Not dummy replacements where 50% are satisfied and the other 50% whine about it.

How can you not see what I'm trying to say...

Parry and block using the same formula. Same formula. How can parry be fine but block underperforms ? It makes no sense... I played a tank with 50+15 shield then, yes blocking sucked. I played a blademaster with 42+16 parry and moparry3 at the same time, he parried a lot like he should.

Obviously the old formula did not resemble the parry formula, if it did, blocking would be more reasonable like parry was(is).

Disdain wrote:Parry and block using the same formula.

You make statements without proof. Also you seem to forget that parry is checked before block so it has more chance to activate.

Do you even understand what I'm trying to say oh my god.....

I'll try to explain it one more time, please please please try to understand what I'm saying.

Parry chance: affected by dexterity and spec
Block chance: affected by dexterity and spec

If parry formula = block formula for dex, spec and mastery of parry/blocking scaling.

How can it be that parry = working fine and blockchance = terrible at the same time ? It makes NO SENSE!

Just set supporters shield spec at a fixed rate at around 5% like on live. Getting blocked 3-4 times in a fight against a mincer is totally silly. Cant see why this hasnt been changed eons ago, totally embaresing for the server with faults like this.
Set it now and we can start do testings for a correct shield spec formula and guard.

ok, this didn't go as planned.

Blue, thanks for responding to this thread. I don't know what formula Uthgard uses, but here is the formula from roclar.net:

Shield% = (5% + 0.5% * Shield)

Block: (((Dex*2)-100)/40)+(Shield/2)+(Mastery of Blocking * 3)+5. < [that's old RA master of blocking]

With this formula, a support class with 300 dex should have a 17.5% chance to block.

A defense tank with 300 dex and 42+16 shield spec should have a 46.5% chance to block. Add MoB 3 and that chance is 56.5%.

If these defense rates sound reasonable, they're not. Assuming this defense formula is live-like then it was balanced against defense reduction from an attacker's weapon skill, and weapon skill table.

My testing on live showed that a >buffed< attacker reduced his target's evade rate by 1/4 to 1/2 depending on the attacker's class. My testing also proved that defense penetration is fractional.

So let's assume that defense penetration works the same way for block, and we have a full-buffed attacker that reduces his target's defense by 1/3. We'll use the three block percentages from before:

17.5 reduced by 1/3 = 11.66%

46.5 reduced by 1/3 = 31%

56.5 reduced by 1/3 = 37.6%

Does this seem more reasonable to anyone else?

Disdain pointed out that block should use exactly the same formula as parry. Roclar's testing backs this up. What formula does Uthgard use?

Uthgard used this formula before according to blue when blockrates were too low. I was trying to point out to him that the formula is good but it's implementation was somehow wrong but he doesn't seem to understand my point. If it works for parry it should work for blocking.

Seyha wrote:

Shield% = (5% + 0.5% * Shield)

Block: (((Dex*2)-100)/40)+(Shield/2)+(Mastery of Blocking * 3)+5. < [that's old RA master of blocking]

A defense tank with 300 dex and 42+16 shield spec should have a 46.5% chance to block. Add MoB 3 and that chance is 56.5%

See what you did there :p

Seyha wrote:

Shield% = (5% + 0.5% * Shield)

Block: (((Dex*2)-100)/40)+(Shield/2)+(Mastery of Blocking * 3)+5. < [that's old RA master of blocking]

Yes that was the old formula and it was wrong or not enough. It just doesnt got enough block rate. Thats why many player complained about it. Lookup the forums. It was implemented exactly like this. And don't forget, the block rate doesnt apply to swing count but to (swing-evade-parry) count.

I'm not sure what to say then, Blue. People complained about having a 45-50% chance to block? Because that's what your typical buffed def tank will have with Roclar's formula. Add in MoB 3 or 4 and you have a 55-60% chance to block.

60% damage mitigation on top of parry and evade, and people still complained?

I'm sorry but I don't see what could possibly be wrong with that. Since defense penetration is not modeled on Uthgard a shield tank with some points in parry would stop like 75% of all attacks with the old formula. That's an insane advantage, and with the new formula the situation is even worse.

That doesn't even cover the advantage given to assassins with their capped evade chance.

Seyha wrote:I'm not sure what to say then, Blue. People complained about having a 45-50% chance to block?

Right. Obviously yes.

See the first try to show us the differences between live and Uthgard:
http://img218.imageshack.us/f/blocktestxe5.jpg

Another example:
http://img224.imageshack.us/img224/3365 ... estfs9.jpg

And more:
http://staff.uthgard-server.net/raghor/Blocktest1.JPG

Yes, the difference is obvious, but the problem is that all those tests were performed on mobs. We have no way of knowing exactly how block works in PvE. What is the base-line defense reduction for a mob on Live compared to Uthgard? How much does a mob reduce defense rate compared to a buffed player character?

We don't know any of that. We have based Uthgard PvP defense rates on PvE testing on Live. This is a major problem.

You have a system that is balanced against two sides: defense bonus and defense penetration

* defense bonus comes from spec, RAs, abilities, and stats; it is additive (spec+stats+RAs=defense rate)

* defense penetration comes from spec, stats, and weapon type; it is fractional (defense/defense penetration=actual defense rate)

On Uthgard, the only defense penetration that is modeled is the bonus from weapon type. This creates several imbalances:

* First, it imbalances high defense classes like assassins and tanks by giving them much higher defense rates against one-hand, h2h and 2hand attackers

* Second, it imbalances support classes for the same reason; much higher defense against non-DW classes

* Third, it imbalances DW/LA classes by making them the ONLY classes with defense penetration against the two most powerful defense abilities (evade and block); they effectively hit TWICE as often as other classes when attacking in the front arc, so they do at least TWICE the damage

Consequently light tanks are even more desirable than other dps classes because of their exclusive defense penetration; assassins and rangers dominate the stealth and solo aspect of the game for the same reason

All of these factors combined make playing on this server a miserable experience for a lot of classes. It really does ruin the game for a lot of people, it isn't live-like, and it isn't fair. /qq

Don't mix the results above with further tests. They were only indicative comparisons that something is propably wrong. After that every following test was done in PvP.

Thanks Blue for your work.