Click to open network menu
Join or Log In
Mobafire logo

Join the leading League of Legends community. Create and share Champion Guides and Builds.

Create an MFN Account






Or

's Forum Avatar

Using Maths in Builds: Proving why some things...

Creator: SpicyRice April 6, 2011 10:09pm
SatansFate
<Member>
SatansFate's Forum Avatar
Show more awards
Posts:
213
Joined:
Sep 30th, 2010
Permalink | Quote | PM | +Rep April 10, 2011 5:15am | Report
This would be better done as a histogram IMO, then you could use frequency density to compare effects which last different times.
SpicyRice
<Member>
SpicyRice's Forum Avatar
Posts:
92
Joined:
Apr 3rd, 2011
Permalink | Quote | PM | +Rep April 10, 2011 4:24pm | Report
i've just done black cleaver and decided to use armour instead of time as the x axis. these are pretty clear and useful i think.

in the guide i've put them in there is a lot of explaining these and detailing assumptions etc, (see middle of Item Analysis section: http://www.mobafire.com/league-of-legends/build/ranked-ad-sion-a-comprehensive-guide-70183 if you're interested)
Uploaded with ImageShack.us[/img]

there are two lines for Black Cleaver which show average DPS when there are 2 stacks on the enemy and when there are 3 hits. It's kinda difficult to model otherwise - since you reach the 2-stack average on your third hit, it's safe to say that actual DPS is somewhere inbetween the two lines for black cleaver.


i did these graphs on Open Office and thickened the lines on paint, what are your thoughts? i think they add a lot, especially to claims like "get Black Cleaver next because it has best DPS per gold and still works well with Infinity Edge despite giving no crit chance"
Searz
<Ancient Member>
Searz's Forum Avatar
Show more awards
Posts:
13418
Joined:
Jun 6th, 2010
Permalink | Quote | PM | +Rep April 12, 2011 6:35am | Report
Those graphs are all kinds of confusing too. You don't even say what the Y axis represents. I assume it's gold since the values are so big, but Y NEEDS to be affected by X to make any sense.
Y needs to be damage or damage per second. X needs to be armor/MR.
You're also using a bad program for graphs. I don't think you can even use formulas in Office..
Google has a job title called "Head of Black Community Engagement"..
I don't know whether to cry or laugh.. or both.
SpicyRice
<Member>
SpicyRice's Forum Avatar
Posts:
92
Joined:
Apr 3rd, 2011
Permalink | Quote | PM | +Rep April 14, 2011 10:09am | Report
Searz wrote:

Those graphs are all kinds of confusing too. You don't even say what the Y axis represents. I assume it's gold since the values are so big, but Y NEEDS to be affected by X to make any sense.
Y needs to be damage or damage per second. X needs to be armor/MR.
You're also using a bad program for graphs. I don't think you can even use formulas in Office..


i should have labelled the Y axis. it is the damage dealt over 8 seconds (i chose 8 secondss because this is the lenghth of youmuu's active)

i didn't include gold in the calculations, i think it's still reasonably easy to compare items by saying... hmmm black cleaver costs 200 more than PD but does much more damage. so it's better...

the program is the best i have, although it doesn't allow you to use formulae to make the graph, i go the long way around by populating a spreadsheet with data for damage done against different armour values by different items. i then go over the lines in paint to make them thicker..

i will label the axis next time, any more feedback?
Searz
<Ancient Member>
Searz's Forum Avatar
Show more awards
Posts:
13418
Joined:
Jun 6th, 2010
Permalink | Quote | PM | +Rep April 14, 2011 1:50pm | Report
SpicyRice wrote:

the program is the best i have, although it doesn't allow you to use formulae to make the graph, i go the long way around by populating a spreadsheet with data for damage done against different armour values by different items. i then go over the lines in paint to make them thicker..

Use this program and use formulas! >.<

"I love the dirty bomb tag because i get either

a) posts about the game

b) current world affairs" - steel-sentry
SpicyRice
<Member>
SpicyRice's Forum Avatar
Posts:
92
Joined:
Apr 3rd, 2011
Permalink | Quote | PM | +Rep April 27, 2011 3:40pm | Report
i accept that your program probably works and will look into using it next time i want to do some graphs, however there is nothing to be gained from using that program right now; the graph i have done is not incorrect, it arrives at the same end product using a different method.

As i have said, formulating DPS would be difficult because of things like Tri Force Procs, which can become more frequent with CDR etc etc. I have successfully formulated it using an elaborate series of formulae and spreadsheets, which are not easy to condense into functions for the graphs.

Because making the functions was so difficult, i took a short-cut and just recorded the DPS for each item against different armour for different time periods and then put this data into a spreadsheet which i made a graph from.

This is just a different approach and is faster for one-off graphs, slower for repeats but certainly not incorrect.
Searz
<Ancient Member>
Searz's Forum Avatar
Show more awards
Posts:
13418
Joined:
Jun 6th, 2010
Permalink | Quote | PM | +Rep April 27, 2011 11:47pm | Report
SpicyRice wrote:

the graph i have done is not incorrect, it arrives at the same end product using a different method.

It's not incorrect, but it's not a real graph. There is no variable so it can't be a real graph. What you have done is more like a bar chart, but with lines >.>
"Gold can't buy you rape." - Mr Sark
DEWO
<Retired Moderator>
DEWO's Forum Avatar
Show more awards
Posts:
5526
Joined:
Feb 15th, 2010
Permalink | Quote | PM | +Rep April 28, 2011 3:20am | Report
Kinda odd. How did you made the calculations for crit on PD and IE?
This is rather a random thing.
Did i help you? Click that +Rep

Thanks to jhoijhoi for the sig!
Searz
<Ancient Member>
Searz's Forum Avatar
Show more awards
Posts:
13418
Joined:
Jun 6th, 2010
Permalink | Quote | PM | +Rep April 28, 2011 3:53am | Report
@DEWO
Calculating average damage is not that hard. I do it in my graphs.
Average damage really is the only way to compare crit items to other items.
"I'm ebolakin. You can refer to me as Ebola-Chan." - Vurtax

"I also am Otherkin. Whenever autumn rolls around I can feel my kin-type slowly taking over my body. You must know, I identify as pumpkin. I can control it pretty good most of the time, but when September ends, I just cant hold it in anymore. Whenever Im outside I see my brothers and sisters being chopped into soup, coffee, donuts, cake, drinks, bread - ppl even cut human faces into their skin, making a mockery out of their noble appearance. When I see things like that I cant control my pumpkin urges anymore. My natural instincts kick in. I then sit down motionless, while getting bright orange." - Morgana L
SpicyRice
<Member>
SpicyRice's Forum Avatar
Posts:
92
Joined:
Apr 3rd, 2011
Permalink | Quote | PM | +Rep April 28, 2011 7:58am | Report
@ DEWO: modelling crit items isn't too hard, you just have to know crit chance, bonus crit damage, attack damage, attack speed and armour pen; when Tri-Force procs and CDR gets involved it gets a whole lot tougher.

i inferred how the items interact and used what info i could find on the net. If anyone knows of or is interested in compiling a description of the game mechanics esp item interactions i would like to know/contribute. i think a lot of guides would benefit from a well thought-through and constructed resource. i don't know of any and had to do my own. i am free in june and would be happy to work on this with some others if they're up for it.

@ Searz: armour is the independent variable, DPS is the dependent variable which changes as armour is varied.

You need to log in before commenting.

League of Legends Champions:

Teamfight Tactics Guide