Paradox had 18 QA peeps playing this for weeks?

[Sensuki]

BG1 combat is fun, even if it's not super active.

 

Just like a point in dexterity increases your chance do deal damage (which is, in turn, modified by might).

No. Low Accuracy High Interrupt is a gimped character, whereas due to the Attack Resolution system, hits are normalized so varying Might and Dexterity is viable.

 

Interrupt is an absolute effect or no effect, so the reliance is much higher.

Edited September 6, 2014 by Sensuki

[Valorian]

BG1 combat is fun, even if it's not super active.

 

Just like a point in dexterity increases your chance do deal damage (which is, in turn, modified by might).

No. Low Accuracy High Interrupt is a gimped character.

 

 

You're wrong, because you can't lower your dexterity to the point of not hitting anything. There's a base accuracy that advances as you level up, not to mention abilities/spells/gear.

 

That would be like saying; low accuracy (you rarely hit anything!) + high damage (from might) is a gimped character; might is useless! It's just an unlikely scenario, like the above. 

 

You'll graze/hit things regardless.

[Valorian]

Let's make it simple:

 

Character A with 3 PER and 18 DEX. His chance to hit against a specific opponent is 50%. His chance to interrupt on a hit is 20%, on a graze 10%.

 

Character B with 18 PER and 3 DEX. His chance to hit against a specific opponent is 35%. His chance to interrupt on a hit is 60%, on a graze 30%.

 

Character B is not "gimped".

[Sensuki]

I'm not wrong. Might is extremely useful as it adds something every time you deal damage. Accuracy is less useful (and thus, gives less overall DPS except in extreme edge cases).

 

Interrupt is primarily item based. Interrupt gives a percentage increase to the item's base interrupt you are using. It gives you a benefit less often than Dexterity does, and it is reliant on Dexterity. If you have a low Dexterity, you will end up getting mostly grazes and hits, with misses. This means that your chances of causing interrupts will be pretty darn ****ty, and only average if you use a Morningstar.

 

Making such a build is effectively gimping yourself, as you would have been much, much better off putting those points into Dexterity instead of Perception because it gives you a DPS increase every point.

 

I know exactly how the system works.

 

FTR, here are the Interrupt mechanics

 

These are the RAW, though I don't currently have the constants handy:

 

***

 

Here are the basic mechanics for Interrupt (modified from the attack's Base Interrupt) and Concentration (derived from a constant Base Concentration), which are two opposed values used for determining if someone plays a hit reaction when they take damage. It comes down to a comparison of an Interrupt value (from the attack and attacker's stats) vs. the defenders' Concentration values (from the defender's stats) with a random percentile roll.

 

A hit reaction always stops movement, cooldown timer, reloading, and all other actions. Movement, cooldowns and reloads do not reset but are simply paused and resumed. All other actions (attacks, spells, abilities, etc.) are stopped and must be restarted from the beginning.

 

In general, these values do not scale tremendously over the course of the game. They also are generally applied based on the speed and area of effect of the attack. A fast AoE will have a miniscule Base Interrupt. A slow single-target attack will have a high Base Interrupt.

 

The general rule is this: the slower the attack, the higher the Base Interrupt. The more targets it affects, the lower the Base Interrupt.

 

Interrupt Scale:

[None]

Weakest

Weaker

Weak

Average

Strong

Stronger

Strongest

 

These names are associated with constants. If the constant is adjusted for "Stronger", it will automatically affect everything set to have a Stronger interrupt.

 

Every Attack can have an Interrupt value, though it should be set to [None] unless damage is part of the attack. Note that even if damage is not, in the final tally, applied to the target, it can still be interrupted. However, the attacker must at least score a Graze. The formula for calculating interrupt follows:

 

Interrupt = Base Interrupt * (1 + [{Perception * 3}/100])

 

The attack roll can further modify this value. The final result increases by 50% if the attack was a Crit and is cut to 50% if the attack was a Graze. Additionally, a Disengagement Attack automatically increases the final result by 50%. A Disengagement Attack that Crits would increase the result by 50%, then that result by 50%.

 

For example, let's say someone attacks with a Sword. The Sword has an Average Base Interrupt. We've defined Average to be 40. The attacker has a Perception of 15. Interrupt = 40 * (1 + [{15*3}/100]) or Interrupt = 40 * 1.45, so Interrupt = 58.

 

Concentration = Base Concentration * (1 + [{Resolve * 3}/100])

 

Base Concentration is a constant. Let's say it's 50, assuming we want a lower number of Average attacks to call a hit reaction (we may not, which is why this is a constant we can tune). The defender's Resolve is 18. Concentration = 50 * (1 + [{18*3}/100]) or Concentration = 50 * 1.54, so Concentration = 77.

 

To call a hit reaction, the percentile roll needs to be 51 or higher. The roll is 1-100 + Interrupt - Concentration. In this case, it's 1-100 + 58 - 77. It's kinda hard to call a reaction (only on 70 or higher), but it wouldn't be rare.

 

Now let's say with the same characters an attacker is using Stilettos with Weak Base Interrupt. We define Weak as being 30, so the Interrupt is 44 (43.5). With a compared 77 Concentration, the roll now needs to be an 84 or higher to call an Interrupt!

 

If the attacker really wants to call an reaction, he or she may use a Morning Star, which has a Strong Base Interrupt (50). This results in a 77 Interrupt, which exactly matches the 77 Concentration. A roll of 51 or higher (half of all Hits) will call a reaction. On a Crit, the Interrupt is increased to 116 (115.5), meaning even a roll of 12 would call a reaction! A Crit Disengagement Attack would do it every single time, though obvious those attacks have to be provoked.

 

Magical weapons can have a special property (called Superior Interruption) that bumps their set/listed Base Interrupt rating up by one category. E.g., a Stiletto would be set to Weak but be bumped to Average.

Magical items (of any sort, but usually armor) and spells/abilities can have a special property, called Interrupting, that adds to the modifier normally calculated from Perception. This adds directly into the percentage increase of the Base Interrupt. Conversely, they can have a property called Concentrating that specifically adds to the modifier normally calculated from Resolve.

[Valorian]

Clearly it depends on the numbers: does 1 point in perception increase my chance to interrupt (considerably) more than 1 point in dexterity. That's hardly a shocking revelation.

[Sensuki]

I haven't done those maths, but I would assume it does give you more of a chance to cause an interrupt than a point in Dex does. However, a point in Perception is worth less per point the lower your Accuracy is, and there's a reason why the majority of players are dumping Res and Con.

 

One of the aims of the system was to prevent yourself from making a terrible character, I'd say the system has kinda done a good job with that, but one of the worst characters you could make IMO is High Perception, Low Dex.

[Ineth]

Classes: [...] conceptually are already leagues ahead of IE games. [...] all the classes feel so much different from each other in PoE.

 

Not when it comes to melee fitness, for example.

 

In Baldur's Gate, a fighter wielding warrior-only weapons with grandmastery, was probably 20 times as deadly as an un-buffed mage who joined melee combat with one of the few weapons that mages were allowed to wield and only a single proficiency point in it (and crappy Mage base thac0 and attacks-per-round).

 

In PoE, that distinction is all but blurred.

 

And class differences aside, what PoE lack the most is real combat-relevant differentiation between different characters of the same class. Right now, two characters with different attribute distributions will still be able to viably do all the same things (at best, one of them may do some things a little better than the other), and it looks like every character can viably wield any weapon regardless of chargen choices.

 

If Obsidian rejects...

• classes with significant impact on melee dps
• attributes with significant impact on anything
• weapon specialization
• weapon-style specialization
• thieving skills specialization
• etc.

...then character build customization will have to come from elsewhere. And I'm not convinced that the talent system will be able to fill that gap by itself.

[Valorian]

I haven't done those maths, but I would assume it does give you more of a chance to cause an interrupt than a point in Dex does. However, a point in Perception is worth less per point the lower your Accuracy is, and there's a reason why the majority of players are dumping Res and Con.

 

 

 

Not to derail this thread any further, but...

 

Just like 18 dex + 18 might will deal more damage than 3 dex + 18 might, 3 dex + 18 PER will not interrupt as often as 18 + 18. 

Yes, attributes are interlinked.

[Sensuki]

You've conveniently ignored the second part of my post, but suit yourself.

[Valorian]

Ok.

[Falkon Swiftblade]

To be fair to Obsidian, I purchased the Enhanced edition to BG 1 and 2 which came out I think 2 yrs ago, and even though they were using a much improved engine, they still had hundreds of bugs they managed to fix, and they've gone through several huge patched since and I think last count they were well over 700 bugs fixed from the old game and the new version and counting even with the help of the old modders and newer tech being used. Similarly with this game, there's a ton of systems in the game with a lot of choose your own path to play, so that compounds things. Finding the bugs is a good thing. However, with all due respect to Obsidian, they have a rotten reputation for putting out some really buggy games. This time though, there are a LOT more people helping them test and balance things, so this might be well worth the headaches now. I'm sure we've got a solid 3 months before launch which is a long time to iterate and polish this off