Is there a sweet spot for Critical Chance?

[lanceuppercut]

Since crit chance is the "in thing" since IAS bit the dust, is there a point where it's better to stack stuff into crit damage instead of piling an extra 1% here and there into chance?

Am sitting at 34.5% crit chance (without follower buffs) and the only thing I can see myself changing at this point are my gloves which have 5.5% crit chance, is it worth dropping 5.5% in favour of some more damage which at the moment, am at 170%. Or maybe my bracers which have 3% crit chance.

[Karth]

I haven't made a comparison including item budgets and the like, but a 10:1 crit damage : crit chance ratio is what you want if you don't get any other benefit from chance to crit. Like anything else multiplicative, the farther from 10:1 you are, the more valuable the lower stat becomes. You could probably gain substantial dps by swapping your crit chance gloves for gloves with decent crit damage or better yet, crit chance and crit damage (with preference given to crit damage). Or you could switch your amulet to one with crit damage; they're pretty cheap and can get crit damage (65% instead of 50%) than gloves. You won't find crit damage on bracers, though; crit chance is the best offensive mod rare bracers can get.

Of course, I'm saying this at 4 in the morning and I don't know what unhealthy hour it was when I decided 10:1 was about right, so a second opinion couldn't hurt you.

Edit: and again, like magicrectangle says, if you get some other benefit from crit chance, it is often better to increase your crit chance in spite of a dps disadvantage. My WW/Sprint barb for example could drop 6% ctc for ~50% crit damage and gain about 12% more dps... but he'd lose fury generation, crits with which to refresh Overpower, WotB against tough mobs, etc., so I'm content to be patient about increasing my crit damage.

[magicrectangle]

It is impossible to answer this question without knowing your class, build, etc. The short answer is that there are many damage calculators out there now for the various classes. Simply plug in the stats of the item you're looking at buying, and see if it will increase your damage or not.

The longer answer is that many builds out there now rely on crit chance not just for damage, but for some additional desirable effect. In these builds often it can be a good idea to stack crit chance even when it lowers your damage. For example on a critical mass wizard I like to maintain at least 40% crit chance. I could give up some crit chance and gain a lot of damage, but it would harm my ability to tank.

[lanceuppercut]

Not really sure if I can pigeon hole my build into a type but here goes, am a HC DH using Natalya's 4 piece set with a xbow, running minimal LoH (200 or so) but around 1.70 or so attack speed. Am highly mobile with +25 MS, caltrops to stun/slow and a double vault with tactical advantage let's me get away, if needed. The usual SS to break CC and prep for the 60% heal. Evasive Fire/CF and impale/overpenetration round out the offensive parts of my build.

I tend to kite using traps the best I can to slow enemies. I don't think there is anything that I can see on my build that *relies* on crit chance (I guess you mean things like the Barb's WotB rune where a crit hit increases the timer?), other than seeing the big yellow numbers appear

Am at 35k DPS and HP, around 3.3k armour and lowest all resist is about 190, it's so low as fitting in Nat's gear is the cause of that. Biggest ballpark crit number I've seen is 100,000.

[nintes]

With a crit chance as high as that, you may like to consider using the nightsalker passive, if you are not already. With nat's gear, your crit chance and 1.7 attack speed and some +max disc gear, it becomes reasonably viable to smoke screen spam. Especially if you use the caltrops rune that gives you a 10% to crit chance while standing in the circle. Something to consider

[lanceuppercut]

I use the standard-y passives of tacticsl advantage, steady aim and archery. only one I could see myself dropping is steady aim but tactical advantage, evasive fire and steady aim working so well in unison, it's hard to break up for nightstalker.

getting a +disc nat's is hard, there is one on the HC AH at the moment but 15m is just not worth it which leaves the quiver, paid a pretty penny for a +19 IAS, 7.5 crit chance, dex/vit and hatred dead mans legacy.

[Vanadin]

Like anything else multiplicative, the farther from 10:1 you are, the more valuable the lower stat becomes.

The 10:1 rule is only for resistances, and that's because 50% reduction = 3,000 armor but 300 resist.

In general, the actual issue is the ratio of changes. For example: if you have 20% crit chance and 100% crit damage, then on average you do an extra 20% damage/hit. If you want to increase that number to 30% damage/hit, you can increase the crit chance to 30% (up it 50% from its current value) or increase crit damage to 150% (ditto).

So the key is the % change in each stat. The larger the % change, the more DPS you'll have.

[Karth]

The 10:1 rule is only for resistances, and that's because 50% reduction = 3,000 armor but 300 resist.

In general, the actual issue is the ratio of changes. For example: if you have 20% crit chance and 100% crit damage, then on average you do an extra 20% damage/hit. If you want to increase that number to 30% damage/hit, you can increase the crit chance to 30% (up it 50% from its current value) or increase crit damage to 150% (ditto).

So the key is the % change in each stat. The larger the % change, the more DPS you'll have.

...with 10:1 crit damage to crit chance being most optimal. Nowhere am I talking about resistances; please do not assume I am a moron. An increase from 5% ctc to 15% ctc is the same increase in dps as an increase from 50% crit damage to 150% crit damage. However, if you instead split the increase between chance to crit and crit damage, for 10% ctc and 100% crit damage, you're doing more dps than by focusing entirely on either one.

[Vanadin]

...with 10:1 crit damage to crit chance being most optimal. Nowhere am I talking about resistances; please do not assume I am a moron. An increase from 5% ctc to 15% ctc is the same increase in dps as an increase from 50% crit damage to 150% crit damage. However, if you instead split the increase between chance to crit and crit damage, for 10% ctc and 100% crit damage, you're doing more dps than by focusing entirely on either one.

Your logic is actually true independent of the ratio between the two:

An increase from X% ctc to 3X% ctc is the same increase in dps as an increase from Z% crit damage to 3Z% crit damage. However, if you instead split the increase between chance to crit and crit damage, for 2X% ctc and 2Z% crit damage, you're doing more dps than by focusing entirely on either one.

To take things further, let's say I have 10% chance and 100% damage, per your example above. Also, suppose that I have 4.5% chance and 15% damage on my current pair of gloves. I find a new pair of gloves with 4% chance and 25% damage—that would make the new numbers 9.5% chance and 110% damage. Just using the "10:1 rule" would suggest that this would be worse (110:9.5 = 11.58:1), but it isn't: 10% x 100% = 10% additional DPS, while 9.5% x 110% = 10.45% additional DPS.

There is simply no such thing as an "optimal ratio" here; you simply want the product of the two stats to be as large as possible.

[Karth]

Your logic is actually true independent of the ratio between the two:



To take things further, let's say I have 10% chance and 100% damage, per your example above. Also, suppose that I have 4.5% chance and 15% damage on my current pair of gloves. I find a new pair of gloves with 4% chance and 25% damage—that would make the new numbers 9.5% chance and 110% damage. Just using the "10:1 rule" would suggest that this would be worse (110:9.5 = 11.58:1), but it isn't: 10% x 100% = 10% additional DPS, while 9.5% x 110% = 10.45% additional DPS.

There is simply no such thing as an "optimal ratio" here; you simply want the product of the two stats to be as large as possible.

And 10:1 is about the most resource-efficient ratio for maximizing the product of the two stats, much like the information the OP seemed to desire. If you diverge far from this, you're spending too much of your resources (the affix budget for items and your gold) on one of them when increasing the other would give you more dps.

Of course an example cherry picked to cause this rule to fail will succeed. When I say 10:1 is most efficient it doesn't mean I think water runs uphill and 2+2 can equal five. It means that when you have 20% chance to crit and 100% crit damage you're wasting a whole lot of your gold and thus potential dps; you could have 15% ctc and 150% crit damage, do more damage, and probably save some gold along the way.