Take off Forward Grip from your Pulsar C (thx Wrel!)

Good finds. NC and VS ARs are probably the part of the game I know the least. With AFG Carnage is also moved to the list of weapons that are supposedly made less accurate by grip.

Carv S seems to be fine with grip (in theory).

I'll test Corvus and LSW later today.

I haven't included auraxium weapons in the sheet for this very reason - they've been changed around so many times I have no idea what they're supposed to be, let alone what stats they have. And seeing as there's no choice of using or not using a forward grip on them, it doesn't really matter.

Keep in mind that issues with inconsistency can be caused by spreadsheet rounding numbers. As Nakar once said to me, the real numbers in the game files easily can be 0.39995 or something alike. The spreadsheet rounds it up to 0.4, which makes a huge difference.

Sadly, there's no real way to find out other than try datamining this stuff once again by ourselves. I tried it once but gave up halfway.

I'm unsure what the real numbers are, but what we do have are the numbers from the API, which i tend to look when in doubt. Not that they're necessarily right either.
Oddly enough for the Corvus, H.recoil is 0.149999999999 and HRT is a flat 0.5 in the API, but in game as i said i did get the result i expected (2 shots on either side) on the foregrip (so 0.1125 and not 0.112499999999).
It's very possible that the data is rounded at some point either by API or game, so even the numbers we may have through it aren't quite trustworthy.

Also some data is obviously not properly reported: the FSRM for the Gauss Rifle Burst is still reported as "2.14999999" in the API, when that's obviously not the case in-game.

If you need to check anything about a gun with the API, the way i do it is as follow:

1. First go to http://ps2.oraclestats.com
2. Notice that in the menu options each weapon has a specific ID, this is what you'll need. In the case of the Darkstar it's 1909
3. Then use this kind of link: http://census.daybreakgames.com/get/ps2/fire_mode?c:lang=en&item_id=1909&c:join=fire_mode_2
4. What you have now is a JSON object string which you can format in a more readable fashion, by copy-pasting, or even putting the direct link, in an online formatter like: http://jsonformatter.curiousconcept.com/

There are some fields i don't know anything about in there like "recoil_increase"/"recoil_increase_crouched" which might have an effect on our numbers, it's hard to fathom.

Oh that's great. I knew about the census query, but I heard the API was down, and assumed census wasn't an option.

You're correct about data being not trustworthy; as theorycrafting often is. That's why I said it's important to properly in game test all questionable cases.

By properly I mean recording a video and watching it in slow-mo. I'll do it soon.

I've tested Corvus, and it seems that you were right, I was only able to get it to bounce 2 times.

The explanation could be this: maybe I was wrong about accumulated horizontal recoil being able to exceed horizontal tolerance once.

Using these calculations, we can assume that accumulated horizontal recoil can bounce twice, which would produce 0,224999999999999985 accumulated recoil, which is less than half of the HRT = 0.225. But the next shot would exceed it, so the recoil goes to another direction.

The problem I have with that is the system becomes less clear because it requires an exception when HRT < horizontal recoil.

We should be able to confirm / disprove this theory by a new round of theorycrafting, using numbers from census query. I'll do it, but that will require me some time.

Too bad I don't know a thing about web programming, and I can't automate the process of importing data from census.

I'm actually a web developper myself, and have thought of doing this on a scripted basis lately, but the issue here is knowing the guns' original identifiers, like the "1909" i took for the Darkstar in my example above.
In theory it might be possible to make a self-updating google site with a spreadsheet based on the API.

There might be a way through the API, but i've not found it at this point, after trying some stuff through the DA site's census query page.
So basically i would need some insight from people who know about the API to automate it, but i don't know who/where to ask.

Why can't we take them from ps2oraclestats? It's a pain to do it manually, but it needs to be done only once.

Then it's just the matter of automatically quering the census server for each Item ID and parsing the returned JSON string.

Using third party data is more of a workaround than a solution, which a developper typically doesn't like much. In that case i would have 2 alternatives: 1) inputing all the ids i've gathered from ps2oraclestats manually in the code, which would make the script autonomous, but not dynamic or 2) parse the html code from ps2oraclestats and get those ids which is significantly more complex.

After a quick glimpse, i don't believe i could do the latter easily: there's a lot of html code generated with JavaScript(on page loading in browser), so the base page is rather empty. My best bet for that would be to parse their JS code and find if they query the census API client-side or server-side. If they update it client-side i can copy what they do, if server-side i can't do anything.

I meant the first way. It's not like they release new weapons every day, and unlikely DBG will change their weapon ID system. Like I said, it's a pain, but for the most part needs to be done once.

You just did.