Weapon "Epicness" Chart

[AngleWyrm]

Hmmm... Well... How would you do it? If it were your chart?

Implementing the suggestions for a revised and expanded version
Already present in the column 'H' for Warmup time is one of the scaling weights for importance we were looking into, incorporated as (Warmup * 10). A further gain in usability can be had by separating out that 'x 10' importance multiplier into another table, and then referring to it as a sort of look-up reference; it makes a definition for that warmup scaling factor and any changes to it will get reflected into the main table. The second sheet would look something like

Code Select Expand

Accuracy Damage TBS Burst Range Dmg*Burst Warmup
1 1 1 10

If we name the cells in the second sheet, then then the main sheet's column 'H' can read something like "=WarmupImportance * D2" and the importance weights can be tweaked separately without re-writing the main table. Those scaling values can be seen as 1 is it's normal size, 2 is making it twice as important, 1/2 is making it half as important, and so on.

And the Epicness column could be "=(AccuracyImportance * B2)^2 + (RangeImportance * F2)^2 + (DmgBurstImportance * G2)^2 - (WarmupImportance*H2)^2"

[InfinityKage]

Hmmm... Well... How would you do it? If it were your chart?

Implementing the suggestions for a revised and expanded version
Already present in the column 'H' for Warmup time is one of the scaling weights for importance we were looking into, incorporated as (Warmup * 10). A further gain in usability can be had by separating out that 'x 10' importance multiplier into another table, and then referring to it as a sort of look-up reference; it makes a definition for that warmup scaling factor and any changes to it will get reflected into the main table. The second sheet would look something like

Code Select Expand

Accuracy Damage TBS Burst Range Dmg*Burst Warmup
1 1 1 10

If we name the cells in the second sheet, then then the main sheet's column 'H' can read something like "=WarmupImportance * D2" and the importance weights can be tweaked separately without re-writing the main table. Those scaling values can be seen as 1 is it's normal size, 2 is making it twice as important, 1/2 is making it half as important, and so on.

And the Epicness column could be "=(AccuracyImportance * B2)^2 + (RangeImportance * F2)^2 + (DmgBurstImportance * G2)^2 - (WarmupImportance*H2)^2"

I think i'm starting to feel you here. Let's say we wanted to do this all on one sheet. We could leave rows B-G as is. Row H or Warmup*10 is simply row D times 10 in order to give it some weight. TBS=Time between shots. It does not have to be times 10. It can be times whatever we think fits in the equation best. We can then turn rows I-K into Accuracy*AccuracyImportance, Range*RangeImportance and DmgBurst*DmgBurstImportance. Of course row H will already be Warmup*WarmupImportance(TBS*TBSImportance). We can then make row L the Epicness which would I+J+K-L. Is that right?

[AngleWyrm]

yes, that works.

If column 'M' is Epicness that looks like "=I + J + K - L"
Then I'd like to recommend a column 'N' right next door to it that looks like "=I^2 + J^2 + K^2 - L^2"

Then you can sort the sheet on the two different scores, and judge for yourself which ranking of the weapons most closely resembles what feels right to you.

[InfinityKage]

yes, that works.

If column 'M' is Epicness that looks like "=I + J + K - L"
Then I'd like to recommend a column 'N' right next door to it that looks like "=I^2 + J^2 + K^2 - L^2"

Then you can sort the sheet on the two different scores, and judge for yourself which ranking of the weapons most closely resembles what feels right to you.

Sounds good. I'll do it. Now here is the big question. What should AccuracyImportance, RangeImportance, DmgBurstImportance, and WarmupImportance actully be? Like what do you think would be good values?

[AngleWyrm]

Is version 2 out yet?

[InfinityKage]

Is version 2 out yet?

First off. That's a dope picture.

And no. Not yet. I can't decide what to set the values as for AccuracyImportance, RangeImportance, DmgBurstImportance, and WarmupImportance. I'm thinking accuracy should be the most important thing so put that at a times 1. And you have to put TBS at a times 10 for it to even factor since it's such a small number it won't make an impact otherwise. Maybe put range and DmgBurst at a times 0.8? I'm not sure.

[AngleWyrm]

I can't decide what to set the values as for AccuracyImportance, RangeImportance, DmgBurstImportance, and WarmupImportance. It took the work of many interested lives to find ways to answer the kind of question you've asked, and for all their toil, blood, sweat and tears, they've hidden the answers away in magical tombs of an uncommon tongue, gathering dust in the archives. And so we present magicians break those seals and take the knowledge by force.

You have to put TBS at a times 10 for it to even factor since it's such a small number it won't make an impact otherwise.

Looks like a problem with different units of measure: The range of values for one item doesn't have anything to do with the range of another item. And so the purpose of the scaling multiplier shows itself as tool to bring about equal footing for each of the things on the list. So how about starting with giving them exactly equal footing, and then work from there?

There is a set of weapons, and within that set of weapons are the only values we are interested in. Not the theoretical set from which they are drawn, but the ones we actually have. And for each measurement such as accuracy or range, there is a minimum and a maximum value represented in our collection. Normalization is to stretch/squeeze each of those ranges to fit on a new scale of 0..1, and in this way they will have the equal footing that can at least serve as a desirable starting place.

Let's say I have some variable x that goes from 55..180, and I want to fit it into the range 0..1.
First subtract the minimum 55, so the range is 0..125.
Then divide by the maximum 125, so the range is 0..1.

So for every value in my column of x's, I subtract 55 and then divide by 125. They still bear the same relative relationships to each other; we have merely changed our perspective by zooming out and shifting over a bit.

[InfinityKage]

I can't decide what to set the values as for AccuracyImportance, RangeImportance, DmgBurstImportance, and WarmupImportance. It took the work of many interested lives to find ways to answer the kind of question you've asked, and for all their toil, blood, sweat and tears, they've hidden the answers away in magical tombs of an uncommon tongue, gathering dust in the archives. And so we present magicians break those seals and take the knowledge by force.

You have to put TBS at a times 10 for it to even factor since it's such a small number it won't make an impact otherwise.

Looks like a problem with different units of measure: The range of values for one item doesn't have anything to do with the range of another item. And so the purpose of the scaling multiplier shows itself as tool to bring about equal footing for each of the things on the list. So how about starting with giving them exactly equal footing, and then work from there?

There is a set of weapons, and within that set of weapons are the only values we are interested in. Not the theoretical set from which they are drawn, but the ones we actually have. And for each measurement such as accuracy or range, there is a minimum and a maximum value represented in our collection. Normalization is to stretch/squeeze each of those ranges to fit on a new scale of 0..1, and in this way they will have the equal footing that can at least serve as a desirable starting place.

Let's say I have some variable x that goes from 55..180, and I want to fit it into the range 0..1.
First subtract the minimum 55, so the range is 0..125.
Then divide by the maximum 125, so the range is 0..1.

So for every value in my column of x's, I subtract 55 and then divide by 125. They still bear the same relative relationships to each other; we have merely changed our perspective by zooming out and shifting over a bit.

I don't want to sound insulting. You've had a lot of great ideas. But this latest one seems... needlessly complicated. Wouldn't just choosing a good multiplier be faster and easier?

[AngleWyrm]

Wouldn't just choosing a good multiplier be faster and easier? Multiplication alone will not resolve the difference between two ranges. Each range has to start at a minimum value of 0 in order for scaling to work. Gonna have to break out a double-barrel answer to your query: Addition as well as multiplication. I know it's a toughie, but I feel you're up to the challenge.

Addition, shifting the range to start at zero
Multiplication, the scaling factor

Also, aren't there several range accuracy bands?

[InfinityKage]

Wouldn't just choosing a good multiplier be faster and easier? Multiplication alone will not resolve the difference between two ranges. Each range has to start at a minimum value of 0 in order for scaling to work. Gonna have to break out a double-barrel answer to your query: Addition as well as multiplication. I know it's a toughie, but I feel you're up to the challenge.

Addition, shifting the range to start at zero
Multiplication, the scaling factor

Ok. I'm starting to see what you are talking about. The problem is how do I know what the maximum is? For accuracy it's easy 1-100. But for Damage, Range, and TBS I don't know what the maximum is. The maximum is infinite. I mean I could go with the highest I've ever seen but that's still not really the max. It could change if I find a weapon better. Or do you think the highest I've seen will be good enough?

[AngleWyrm]

how do I know what the maximum is? ... do you think the highest I've seen will be good enough? What you've seen thus far is a collection of observations; this is field research data. As your encyclopedia grows the ranges will fill out, approaching theoretical limits. And as new mods come and old ones go, the ranges may change as well. The highest/lowest in the set remains true to the set even as things change over time.

[InfinityKage]

how do I know what the maximum is? ... do you think the highest I've seen will be good enough? What you've seen thus far is a collection of observations; this is field research data. As your encyclopedia grows the ranges will fill out, approaching theoretical limits. And as new mods come and old ones go, the ranges may change as well. The highest/lowest in the set remains true to the set even as things change over time.

Ok. I feel you. New chart created. Check it out and see if I'm on the right track. I still need to add in the multiplication to decide importance of each stat. I just have not decided yet what the importance levels will be.

[AngleWyrm]

Oh yeah, that train is definitely on track! Two things to help: 1). minor bug fix: min gets shifted to zero to set up for multiplying, but max isn't yet scooting over an equal amount before the multiplier kicks in. It's getting pre-stretched before the multiplier gets to do the stretching. 2). The spreadsheet functions MIN and MAX can prevent re-writes if a weapon with new largest/smallest value is added to the table. MIN(F:F) is minimum of column 'F' and can be put into a formula just like any other value. For example "= D2 - MIN(F:F)"

[InfinityKage]

Oh yeah, that train is definitely on track! Two things to help: 1). minor bug fix: min gets shifted to zero to set up for multiplying, but max isn't yet scooting over an equal amount before the multiplier kicks in. It's getting pre-stretched before the multiplier gets to do the stretching. 2). The spreadsheet functions MIN and MAX can prevent re-writes if a weapon with new largest/smallest value is added to the table. MIN(F:F) is minimum of column 'F' and can be put into a formula just like any other value. For example "= D2 - MIN(F:F)"

1. I'm not entirely sure if I follow. Are you saying that amounts are being rounded? I think I might see that. How would I fix this? I could have all amount rounded to the nearest but that would make the sheet less accurate. Any ideas?

2. Thanks for the info. Added to the sheet. Check out the new file and see if it's looking right.

[AngleWyrm]

It turned out to be some in-betweeny convenience feature stuff; the multipliers are fine.