Talk:Cheese Effect

Cheese effect in MH3

As seems currently combaing insanely stale base + trap gives insanely stale effect rather then uber stale. Could someone confirm this. --Nux 13:47, 12 June 2010 (UTC)

• It does that for me with Chillbot/Magma, ACRONYM/Magma, and OoI/Magma, so I'd say yes, it did change. Whether or not the change was intentional, I don't know. --Dreamwalker 13:51, 12 June 2010 (UTC)

I guess the update from Draobyek is right regarding the averaging... What is a little surprising to me is this: pairing the "Über Fresh" from The Gingerbread House Surprise withe the "Insanely Fresh" from either Polar, Dehydration or Gingerbread Bases should yield to "Ultimately Fresh" but only results in "Insanely Fresh" as stated in the article. Any ideas or comments? ____m. 12:32, 14 June 2010 (UTC)

I personally am stumped as to why the Gingerbread House is providing an Insanely Fresh. From checking many setup options, I'm sure of Cheese effect= floor((BassEffect+WeaponEffect)/2). When using it, assume 0 = Uber Stale, 12 = Uber Fresh. However, without the Dev-numbers, working out these values is practically impossible. Although, it is possible that Ultimately Fresh/Ultimately Stale has been removed. Draobyek 14:04, 14 June 2010 (UTC)

The average is close, but it's not completely accurate. For example, try an Insanely Stale base (Magma) and any No Effect weapon - you get Extremely Stale, which is not expected (Very Stale) in either 10- or 12-scale models. There are a few other discrepancies as well. From what I can tell, if you consider No Effect as 0, the final cheese effect tends to be biased away from 0. In any case it's definitely something more complicated than we have right now. For now I will calculate using a lookup table of verified effects until the true formula is found. --Hyperchao 13:07, 15 June 2010 (UTC)

Trying a squared scale, where the effect is round up to the next square number. Assuming 144 = Uber Stale, 0 = Uber Fresh. The Insanely Stale + No Effect would fit in perfectly. Other examples seem to fit this scale quite well. Gonna test this model some more with a bit of coding emulation. Draobyek 19:04, 16 June 2010 (UTC)

Tested the model. It seems to fit a lot better in calculating cheese effects. Tested with Uber Fresh = 0, on a 12 stage model. Appears to be calculated exactly the same. Setting Uber Fresh to 1 made calculations inaccurate. Going to test a 10-scale model now. Draobyek 19:33, 16 June 2010 (UTC)

On a second look, it seems this only works for some of the higher values. I'm going to have to recalculate this one, but I think I'm getting closer. Draobyek 19:39, 16 June 2010 (UTC)

For future reference, this is the table I used to determine freshness. This is based on pairings from the game and recorded, so the only errors would be typos. Rows are Base freshness and Columns are Weapon freshness. Please update as needed when new weapon/base freshnesses appear. Hopefully this will make it easier for other people to verify formula candidates. --Hyperchao 21:13, 16 June 2010 (UTC)

"O" = Uber (as in oo-ber lol), "U" = Ultimate, "I" = Insanely, "E" Extremely, "V" Very, "NE" = No Effect:

B\W

OS

US

IS

ES

VS

S

NE

F

VF

EF

IF

UF

OF

OS

US

IS

IS

IS

ES

ES

ES

ES

VS

S

F

ES

VS

S

ES

VS

VS

S

S

NE

NE

F

EF

NE

ES

VS

S

S

NE

NE

F

VF

EF

F

ES

S

S

NE

NE

F

F

VF

EF

VF

VS

NE

NE

NE

F

F

VF

VF

IF

EF

IF

NE

F

VF

VF

VF

VF

EF

EF

IF

UF

OF

The only thing I cannot work out, is exact formula. I'm getting closer. There is a curve in the values, I just cannot nail it. There is a slight jump after Insanely Fresh, which means that Über Fresh is at a peculiar value, since it doesn't fit in. The closest I've come is nailing the Insanely and inside box.

I'm pretty sure Ultimately Fresh has been removed completely right now, as plotting a chart for data, under Ultimately Fresh, I'm getting an error margin 1.5 between Über + Very fresh values and a 1.0 error margin between Insanely Stale and Über Fresh.

Otherwise, my formula is proving perfect. I may have to use a switch to try to emulate better results. Will report next run soon. Draobyek 22:27, 16 June 2010 (UTC)

I give up with trying to do it by formula. I'm assuming the devs use something *close* to this.

00 Über Stale (?)

11 Ultimately Stale (?)

20 Insanely Stale

27 Extremely Stale

32 Very Stale

35 Stale

36 No Effect

37 Fresh

40 Very Fresh

45 Extremely Fresh

52 Insanely Fresh

61 Ultimately Fresh (?)

72 Über Fresh (?)

Using these numerical values, then the add together, divide by 2 and then rounding down has proven pretty accurate. The difference between the numbers moving away from No Effect are all square numbers. However, Über Fresh doesn't fit in at all. Feel free to experiment further if you like, since I'm completely stumped other than the fact I know the values curve both sides.

However, I'm beginning to wonder. Does Freshness work as No Effect = around 50, Über Stale down to 0 and Über Fresh up to 100. Then it averages hidden values, and gives a label to say approximately where the freshness is? But working that out, would require testing too many stale setups. x_x Draobyek 23:12, 16 June 2010 (UTC)

In the end I can say for sure this new system makes no sense. I add a "No Effect" component and the cheese effect changes. Wut? --Hyperchao 04:14, 17 June 2010 (UTC)

Square numbers are fairly accurate, but still very inaccurate. The closest I've come, is trying binary bits, which is pow(2,n). I can't quite tweak it to nail the accuracy on stale cheese with formulas, but my coding's likely rusty. I'm gonna try some more methods, but this is one headache. Draobyek 13:45, 17 June 2010 (UTC)

After trying a different approach, based of average freshness, I managed to get a table almost matching. I'm pretty sure now cheese effect reflects average freshness in terms of a percentage. There's no way otherwise, that the Über Fresh cheese effect could fit in. No matter my calculations, it's always been way out. Draobyek 16:11, 17 June 2010 (UTC)

Update makes me want to rubber mallet a dev. It's back to the V2 style. Draobyek 19:50, 17 June 2010 (UTC)

¡EPiC FAiL! – Dev's wise. : (( ____m. 22:08, 17 June 2010 (UTC)

History

History of changes needs to be included. Changes in version 3 should not just be directly edited into the article without mention of the old version 2 information. This is to provide players with a connection between old and new to avoid confusion. -- Grexx 11:49, 15 June 2010 (UTC)