[Megathread] DotA 2

[Sorcereria]

The table says two rings is worse than 1 stout when it's not on most ranged heroes.

You should discard your reading comprehension and/or alt + F5 the page couse we have no idea what are you talking about.

[Deleted]

The table has been updated to include a 100% block rate for hero attacks with a PMS. Whats in the PMS columns is as accurate as it can be. As for blocking over 50% of a heroes attack, well.. CM has a max damage of 41 at level 1. You'd be unlucky to block less than half with a PMS as a melee hero.

You also have to remember that this table is taking into account the cost of the items too. The important columns are the ones that state how much gold you pay for 1 point of damage reduction with the various items. While two Rings may give you better protection, they give you worse protection for your gold than a Stout shield. This is the important part, having the information allows you make the most efficient use of your starting gold.
The two Yes/No columns at the end are if the Rings are more gold efficient than a Shield, not if two Rings provide better damage reduction.

I know. That's what I meant. Either way: http://www.reddit.com/r/DotA2/commen...pms_an/cm9r3rf. There you go. You guys are too stuck on semantics to see what's important here.

- - - Updated - - -

You should discard your reading comprehension and/or alt + F5 the page couse we have no idea what are you talking about.

I'm quite aware of what the author has said and how he changed his results. That doesn't change the fact that he didn't know that PMS blocks 100% of the damage whne he created the thread of that he used 60% on Stout. I'm also aware that he says 2 RoP<Stout in terms of mitigation/gold but his results are incorrect which you can easily see if you actually knew how the game worked and some simple math.

[Jester Joe]

I love all these garbage PA players just giving me free items ^_^

How is the item thing even working? I've been playing WoW in my spare time to keep up more than Dota lately.

[StrawberryZebra]

I know. That's what I meant. Either way: http://www.reddit.com/r/DotA2/commen...pms_an/cm9r3rf. There you go. You guys are too stuck on semantics to see what's important here.

The end results in the second spreadsheet match, almost exactly, the ones in the first. There are a few minor changes, but for the most part things are the same. It gives us more information about a single Ring of Protection vs a Stout Shield. The results are surprising to say the least, but that wasn't the discussion we were having in the first place.

I'm also aware that he says 2 RoP<Stout in terms of mitigation/gold but his results are incorrect which you can easily see if you actually knew how the game worked and some simple math.

His results match the ones in the second spreadsheet you linked. He might not know all about the game mechanics, but the math is correct for the circumstances that were outlined.

[Deleted]

The end results in the second spreadsheet match, almost exactly, the ones in the first. There are a few minor changes, but for the most part things are the same. It gives us more information about a single Ring of Protection vs a Stout Shield. The results are surprising to say the least, but that wasn't the discussion we were having in the first place.



His results match the ones in the second spreadsheet you linked. He might not know all about the game mechanics, but the math is correct for the circumstances that were outlined.

It doesn't match at all. Just look at AA, Bat, Bane, Dazzle, Necrophos, Storm or about any ranged where author says NO on 2 rings>Stout and compare. If you sort and see all ranged heroes first in my spreadsheet that I linked then you'll see 2 rings is better than 1 stout per gold spent on ALL ranged heroes. The first spreadsheet linked by Sorcereria doesn't say this at all.

[Sorcereria]

If you read carefully why it does say NO. For instance, afromentioined heroes will have 2 Rings > PMS in most cases.
The reason solely due to the nature of gold difference and that in short time a salve or tangos for the price difference are better if you compare 2 rings or Stout. If you compare 2 Rings / PMS the gold difference is much lower and 2 rings scale much better than PMS will ever do.

"This is based on cost efficiency. Suppose each enemy hero attack does 50 damage instead of 51.16 for simplicity. Suppose a hero has 6 armour. 2 rings increases the hero's damage reduction to 41.9% from 26.5% for 400 gold. That 50 damage now does 29.05 damage. The gold per overall damage reduction is 19.093 gold per damage reduced. The gold per increased damage production from the rings' 6 armour only is 400/7.7 = 51.95.

With PMS, the hero takes approximately 21.255 damage per hero attack. The gold per overall damage reduction is 19.133.

With stout shield, assuming the 60% chance to block is true (the actual value is around 52% due to pseudo-random chance), the hero takes on average 27.93 damage per hit. The gold per overall damage reduction is 250/22.07= 11.328"

And trying to discredit someones work becouse of small mistake that was picked and fixed shortly. Seriously.

[Scummer]

I love all these garbage PA players just giving me free items ^_^

Most are pretty bad. Granted I've only come across 2 that rush BF which is actually pleasant surprise.

[Deleted]

If you read carefully why it does say NO. For instance, afromentioined heroes will have 2 Rings > PMS in most cases.
The reason solely due to the nature of gold difference and that in short time a salve or tangos for the price difference are better if you compare 2 rings or Stout. If you compare 2 Rings / PMS the gold difference is much lower and 2 rings scale much better than PMS will ever do.

"This is based on cost efficiency. Suppose each enemy hero attack does 50 damage instead of 51.16 for simplicity. Suppose a hero has 6 armour. 2 rings increases the hero's damage reduction to 41.9% from 26.5% for 400 gold. That 50 damage now does 29.05 damage. The gold per overall damage reduction is 19.093 gold per damage reduced. The gold per increased damage production from the rings' 6 armour only is 400/7.7 = 51.95.

With PMS, the hero takes approximately 21.255 damage per hero attack. The gold per overall damage reduction is 19.133.

With stout shield, assuming the 60% chance to block is true (the actual value is around 52% due to pseudo-random chance), the hero takes on average 27.93 damage per hit. The gold per overall damage reduction is 250/22.07= 11.328"

And trying to discredit someones work becouse of small mistake that was picked and fixed shortly. Seriously.

Where did you get that text? It isn't in the original post so I don't see how I can read carefully. All I see is an introduction, changelog and the table.

I also think that math is for a melee hero which is not interesting at all. No one goes double RoP on a melee, it's ranged heroes we are talking about. Why do you take a text about melee damage reduction as exmaple? I already said Stout is better for melee way before that reddit post was made. If you look at the table he is simply doing division by doing Stout shield cost/damage reduced. Then he checks the same for 2 RoP and then decides if he should use NO or YES.

He didn't even fix the mistake because he is still using 60% block on Stout which invalidates his entire table.

[veiledy0]

Why does it actually matter at all which one's better? You wont be using either for more than 5 minutes.

[Deleted]

Why does it actually matter at all which one's better? You wont be using either for more than 5 minutes.

It does because these two guys seem to be unable to do simple high school math and are listening to a guy who has made wrong calculations because he assumed the wrong things and now they are arguing for that the data is still correct even though he is using the wrong values and they are drawing the wrong conclusions from it.

RoP is also indefinately more useful than a Stout shield on ranged because you can build a Basi which is almost always good. With stout shield you can't do shit.

@Sorcereria and StrawberryZebra
Look at my link and that table and tell me what's wrong with it and why it differs so wildly from the first link. Also tell me why if you just look at the gold/delta damage reduction the difference is so huge between 2 RoP and 1 stout? You two seem to struggle with high school math to be honest. That's fine if you're in middle school I guess.

[Sorcereria]

Ariadne 60% on stout is perfectly fine. The difference between 52/53 and 60% is negligible and the % difference doesnt occur constant.
Do you really want to compare 400g 2x RoP vs 250g Stout? This is the reason why the delta is so high for a ranged hero. Couse you pay almost TWICE MORE.

On a hero that has low starting armor, a RoP or 2 + salve will grant enourmous EHP. Not to mention heroes with high HP regen and good str gain.
Whole this argument is being done from a bad side. Count EHP within certain propable encounters.

Everyone forgets, that going by an argument how long you can sustain the lane, important factor is the regen on top of your ability to take hits.
A good example is alchemist and his ulti. Given not enough EHP and his regen means nothing if you die before regen can kick in and enemy can burst you.
Give a hero with high dmg reduction but no regen and its the same story.

[veiledy0]

RoP is also indefinately more useful than a Stout shield on ranged because you can build a Basi which is almost always good. With stout shield you can't do shit.

Does RoP being better than Stout on ranged heroes even have to be discussed tho? I don't think you need maths to know now to buy Stout on ranged heroes, and that Basi is a decent purchase on every ranged carry in the game (Drow, Weaver, Luna, etc). If someone doesn't know/understand that, I don't see much point in arguing it.

[Deleted]

Ariadne 60% on stout is perfectly fine. The difference between 52/53 and 60% is negligible and the % difference doesnt occur constant.
Do you really want to compare 400g 2x RoP vs 250g Stout? This is the reason why the delta is so high for a ranged hero. Couse you pay almost TWICE MORE.

On a hero that has low starting armor, a RoP or 2 + salve will grant enourmous EHP. Not to mention heroes with high HP regen and good str gain.
Whole this argument is being done from a bad side. Count EHP within certain propable encounters.

Everyone forgets, that going by an argument how long you can sustain the lane, important factor is the regen on top of your ability to take hits.
A good example is alchemist and his ulti. Given not enough EHP and his regen means nothing if you die before regen can kick in and enemy can burst you.
Give a hero with high dmg reduction but no regen and its the same story.

It does matter, lol. That's the entire difference between the table I linked and the one you linked. I said gold cost per delta damage reduction. It doesn't matter how much an item cost when you are calculating how much gold you pay PER damage reduction, lol. Are you serious?

Again, tell me what's wrong with my link and why the conclusion differs from your link? The only thing changed is that the guy assumed 53% on Stout instead of 60%. Why did you link that thread if you want to look at EHP and account for regeneration and stuff? You just linked that link because you wanted to show that 2 RoP is worse than 1 stout on many ranged heroes but that isn't the case at all. If you were interested in EHP or any dicussion like that you'd mention it in your original post.

[Jester Joe]

Good god this thread is turning into the whole .9999...=1 thread now.

[Deleted]

Good god this thread is turning into the whole .9999...=1 thread now.

Except this is basic middle school math and these two guys can't even do one calculation to prove their point but instead link to someone else's math which they don't even understand.

[Sorcereria]

You just linked that link because you wanted to show that 2 RoP is worse than 1 stout on many ranged heroes but that isn't the case at all. If you were interested in EHP or any dicussion like that you'd mention it in your original post.

Ok bro. You really do have comprehension issues. I cant be arsed to even bother to quote myself anymore from now on.

[Deleted]

Ok bro. You really do have comprehension issues. I cant be arsed to even bother to quote myself anymore from now on.

Ok bro, nice math there to prove that your link wasn't completely incorrect. LOL.

[StrawberryZebra]

It does because these two guys seem to be unable to do simple high school math and are listening to a guy who has made wrong calculations because he assumed the wrong things and now they are arguing for that the data is still correct even though he is using the wrong values and they are drawing the wrong conclusions from it.

RoP is also indefinately more useful than a Stout shield on ranged because you can build a Basi which is almost always good. With stout shield you can't do shit.

@Sorcereria and StrawberryZebra
Look at my link and that table and tell me what's wrong with it and why it differs so wildly from the first link. Also tell me why if you just look at the gold/delta damage reduction the difference is so huge between 2 RoP and 1 stout? You two seem to struggle with high school math to be honest. That's fine if you're in middle school I guess.

Wrong assumptions aside, the actual math is correct. I'm not sure how many other ways I can try rephrasing this. I even admitted that the whole thing was totally useless outside of lab conditions. Recalculating with 53% block did not even affect the outcomes for fringe cases which were already very close anyway. I'm not sure how closely you've looked at the data in both, but the value of a Stout Shield isn't significantly changed.

The fact that you can build a Ring into a Basi, Aquila or Vlads is totally irrelevant to the discussion. We were talking about how much survivability it gives in comparison to a Stout Shield, PMS or two Rings. Obviously you would buy at least one if you planned on building something out of it.

It does matter, lol. That's the entire difference between the table I linked and the one you linked. I said gold cost per delta damage reduction. It doesn't matter how much an item cost when you are calculating how much gold you pay PER damage reduction, lol. Are you serious?

And this is the other reason why the whole thing is useless - The author used the total of the Heroes armour with two Rings of Protection, not how much the value of additional armour was. But again, the conclusions drawn are the same. Shields are better for melee, Rings are better for Ranged. The Reddit author was just off in how much better they are.

Good god this thread is turning into the whole .9999...=1 thread now.

We have one of those? Why was I not aware of this!? But I think you're right enough Math for now, someone needs to derail things with a complaint about how OP Bloodseeker is or something.

[Deleted]

Wrong assumptions aside, the actual math is correct. I'm not sure how many other ways I can try rephrasing this. I even admitted that the whole thing was totally useless outside of lab conditions. Recalculating with 53% block did not even affect the outcomes for fringe cases which were already very close anyway. I'm not sure how closely you've looked at the data in both, but the value of a Stout Shield isn't significantly changed.

The fact that you can build a Ring into a Basi, Aquila or Vlads is totally irrelevant to the discussion. We were talking about how much survivability it gives in comparison to a Stout Shield, PMS or two Rings. Obviously you would buy at least one if you planned on building something out of it.



And this is the other reason why the whole thing is useless - The author used the total of the Heroes armour with two Rings of Protection, not how much the value of additional armour was. But again, the conclusions drawn are the same. Shields are better for melee, Rings are better for Ranged. The Reddit author was just off in how much better they are.

What? Conclusion is not the same. That's the whole point.

There are very few cases in which those values are close enough for them to be incorrect - In some cases the difference is as high as 9 gold per point damage reduced. I can't at a quick glance see any fringe cases which would be swung one way or the other by recalculating for PRND and in that regard, the math is accurate. It might not be 100% accurate of course, but the outcome would be the same even if recalculated.

What do you mean with that the outcome would be the same? In the reddit table there are on many ranged heroes where from his calculations he comes to the conclusion that 2 Rings is worse than one stout shield in terms of gold per damage mitigated. In my table it's the opposite.

By not using 53% instead of 60% you get for many ranged heroes that 2 rings are worse than Stout shield, which isn't true if you use 53%. You're so stuck on semantics. The values used by the author leads to wrong conclusions.

You also said this:

The end results in the second spreadsheet match, almost exactly, the ones in the first. There are a few minor changes, but for the most part things are the same.

which is completely false. Have you revisited and looked?

You also said this:

Looking through the table, it appears that 2 Rings are only better than a shield for ranged heroes with low base armour.

which you now changed to this:

Shields are better for melee, Rings are better for Ranged. The Reddit author was just off in how much better they are.

[StrawberryZebra]

What do you mean with that the outcome would be the same? In the reddit table there are on many ranged heroes where from his calculations he comes to the conclusion that 2 Rings is worse than one stout shield in terms of gold per damage mitigated. In my table it's the opposite.

By not using 53% instead of 60% you get for many ranged heroes that 2 rings are worse than Stout shield, which isn't true if you use 53%. You're so stuck on semantics. The values used by the author leads to wrong conclusions.

Compare columns AA-AG for the table you linked to the one on Reddit. The AB column is different due to the changed block value on the Stout Shield, but AA and AC are the same. Furthermore, the outcomes of AF and AG show very similar results to the table on Reddit, the differences arising from cases where the incorrect value of a Stout Shield pushed it closer to the value of two rings, (Like AA, Bane and Batrider).

The value of Rings of Protection has remained identical across both tables. This leads to the conclusion that the original authors math for it is correct, and that they haven't done anything stupid like try dividing by zero. I have no idea what it is you are looking at if you are finding different values for the same fields. I do not really see how you can claim that someone else proving this math correct somehow invalidates all of its results?

You cannot compare the delta values because the original author either has not included them or has not done them. They should have, but they have not. That is why the results in AD-AG are inconsistent with the Delta results - They are including the Heroes base armour in the value calculations. And that is fine if you are looking for the best gold in to defences out.
Columns AH-AO are the value of what the Hero gains from the items without their base armour in the calculations. It tells us how much of a change in their defences they get per gold spent, not their total value of defences per gold spent. Both of those give us useful information in their own ways.