Talk:Shotgun

Shotgun Similarities

As I mentioned on the TF2 Wiki, the Shotgun bears a serious similarity to the pistol grip version of a wood Smith & Wesson M3000 with a sawn off barrel. Here's a picture of an airsoft version ([1]), a video of a synthetic M3000 with a stock and normal barrel (http://www.youtube.com/watch?v=k7TzvXRj2aQ), and a wooden version with a stock and normal barrel ([2]). Please consider this. --Takeiuchi 07:39, 20 July 2010 (UTC)

Shotgun Reload Rates

Interesting, I never noticed the difference. I can't watch the video (Dial-up :/) so could you please fill out a 4-column table with each differing load speed then put it on the page? Thanks! -Firestorm 17:33, 1 July 2010 (UTC)

	Heavy/Engineer	Soldier	Pyro
Attack Interval	0.625	0.625	0.625
Base Reload	1	2	3
Additional Reload	4	5	6

Heavy and Pyro base reload has changed as of October 6 2010. Pyro no longer double reloads the first shell and Heavy initial reload is sped up.--Mechlord 22:52, 8 October 2010 (UTC)

Shotgun + Gunslinger

There isn't much difference between the picture of the shotgun with and without Gunslinger.At least make them with higher FOV to see the difference. --- Mitko1239 (Talk page) --- 07:02, 12 September 2010 (UTC)

Renamed Shotguns

The renamed ones reload slower? Why is that? It's like discouraging people from renaming a gun used by 4 classes.--Awesome Weegee 19:43, 27 October 2010 (UTC)

It's most likely not intentional. It's a bug and valve will get around to it eventually. And tbh, I use a renamed shotgun the reload isn't noticeable for me.Flying Pyro

Huh, you expect Valve to fix it. Okay, you listening? Buggy game < Hats. SpaceAnimator 17:46, 11 June 2011 (GMT)

Reload Glitch

When Holding down the 'R' key (or what ever your reload key is) will make your class continue to reload. You do not gain or lose any extra shells for doing so, it is only visual. This can only be done when reload is available. Don't remember if this works for the frontier justice, but i do know for a fact that this glitch occurs as a pyro. M-NINJA 02:24, 3 November 2010 (UTC)

• Does not happen at all with the engineer. M-NINJA 17:45, 7 November 2010 (UTC)
• Does happen with the pyro shotgun. M-NINJA 17:45, 7 November 2010 (UTC)
• Does happen with the heavy shotgun. M-NINJA 22:54, 7 November 2010 (UTC)
• Does not happen with the soldier shotgun. M-NINJA 22:54, 7 November 2010 (UTC)

The infinite reloading glitch was patched, I'm still testing the other things out, but so far nothing's changed below except for the infinite reloading glitch for pyro and heavy shotguns. --Pb 03:41, 20 November 2010 (UTC) This is especially annoying when using auto-reload with the heavy and pyro. One way around it is to hold down Alt-Fire (Right Click) immediately after firing. This simulates a regular shotgun and prevents the weapon from reloading [No Reloading Animation --Pb 23:07, 9 November 2010 (UTC)] until the mouse button is released or until the clip runs out. To clarify, after being held down, Alt-Fire (Right Click) must be released to fire. One way to shoot goes something like this: Shoot-HoldAltFire-ReleaseAltFire-Repeat. The Alt-Fire (Right Click) will be held down for the majority of the time, only releasing it briefly to fire. An alternate and simpler method is to: HoldAltFire + HoldPrimaryFire-ReleaseAltFire (Weapon Fires)-HoldAltFire. When the shotgun runs out of ammo in the clip, the shotgun will automatically reload, Alt-Fire (Right Click) cannot be used to prevent an empty clip from reloading (if there is more ammo available). Also, once the shotgun fully reloads the entire clip (6 Bullets), the Alt-Fire (Right Click) cannot be used to stop the reload animation, a shot needs to be fired first. Additionally, holding Alt-Fire (Right Click) does not stop the reload process once it starts. The Alt-Fire (Right Click) needs to register before the reload process occurs - right after a shot is fired. The only way to have a full clip without the infinite reload (after having fired a shot) is to switch to another weapon and switch back. I found that if you HoldAltFire -> PrimaryFire (Left Click) in quick succession, the weapon will still fire. The Alt-Fire blocking mechanism has a short period of delay before the "blocking" effect occurs (in all cases) [For Consistency, in these posts I will refer to this as Alt-Fire Delay - --Pb 22:17, 9 November 2010 (UTC)]. It would be interesting if anyone can find out more about this/look at code/do more testing. I've only done tests on the heavy's shotgun and the shortstop, but I assume all reloadable weapons (without an alternate fire ability/animation) share these traits. Spy Alt-Fire (Right Click/Cloak) also has the same property of preventing auto-reload. Spy is the only class that has an animation for Alt-Fire (That I have found), all other classes do not show an animation for Alt-Fire for reloadable weapons. Latest Update: --Pb 03:41, 20 November 2010 (UTC)

:Interesting finds that a search on reloading turned up:

• "Clip-based weapons that reload a full clip at a time can now have their reloads aborted by firing" - [3] - This leads to the reloading property: Shooting/Firing has priority over reloading. (Seems obvious, but this patch made weapons such as Scout's Shortstop, Sniper's SMG, Medic's Blutsauger, Medic's Syringe Gun, and Scout's Force-A-Nature, all of which reload a full clip, have this same property.) --Pb 06:58, 9 November 2010 (UTC)
• "Fixed reload exploit that allowed plays to fire instantly after switching weapons" - [4] - I am not sure how the reload exploit worked, so I'm not sure if this is one of the causes of the reload glitch. --Pb 06:58, 9 November 2010 (UTC)

Also, I found another interesting property. Weapons that reload a full clip at a time *can* stop the reload when ammo reaches 0. I've tested using Alt-Fire (Right Click) on Scout's Shortstop, Scout's Force-A-Nature, Medic's Blutsauger, Medic's Syringe Gun, and Sniper's SMG, the weapons remain at 0 ammo and do not reload as long as Alt-Fire (Right Click) is held down. The same does not apply to weapons that reload one ammo at a time - Weapons such as Scout's Scattergun, Heavy's Shotgun, Soldier's Black Box (Are the ones I've tested so far) stick to what I've written in the previous blurb, that is, upon reaching 0 ammo in the clip, the weapon is forced to reload and Alt-Fire (Right Click) is unable to prevent the weapon from being reloaded. A side note - If a Spy uses Alt-Fire (Right Click/Cloak) right after he reaches 0 ammo, auto-reload will not occur and he will stay in stealth with 0 ammo in his revolver (Tested with L'Etranger, The Ambassador, and Revolver). The Spy is unable to auto-reload in stealth. Also, if timed right, the Spy can be made to look like he has three hands - Fire a shot->Wait for the right moment->Reload. You can see the Spy's left hand with his watch, and an extra left hand reloading (I think I've seen this mentioned before, I'm just adding it here for completion) --Pb 06:58, 9 November 2010 (UTC)

Found that my previous statement that Alt-Fire (Right Click) cannot be used to stop the auto-reloading process once it has started is not entirely true (but it's not completely wrong). The auto-reloading process *can* be stopped (with Alt-Fire), but the *animation* cannot. Let me explain: Empty the clip in a Heavy's Shotgun. The Heavy will automatically start reloading his weapon. If you Alt-Fire (Right Click), the reload process and animation will both continue unimpeded. However! If you hold Alt-Fire (Right Click), then hold PrimaryFire after the Alt-Fire Delay, the process of reloading will be stopped. The animation continues, but no ammo is being reloaded into the clip. I assume this is because Shooting/Firing takes priority over reloading: Shooting/Firing is preventing the weapon from reloading, and Alt-Fire (Right Click) is preventing the weapon from shooting. I also found that PrimaryFire does not need to be held down. The process of preventing a weapon from auto-reloading goes something like this: Have the weapon Reload->HoldAltFire->Wait for Alt-Fire Delay->PressPrimaryFire. If PrimaryFire is released, the weapon continues to have the property of not reloading, but the animation for reloading continues. --Pb 22:17, 9 November 2010 (UTC)

I found that manually reloading (ReloadButton) takes priority over Alt-Fire (Right Click). For example, you've stopped the auto-reload by Shooting/Firing then holding Alt-Fire. If you press/hold the ReloadButton (R Key), it will override the Alt-Fire (Right Click) and reload the weapon. I'd like to remind whoever reads this section that for all of the above, I am referring to auto-reloading, the option found in Option->Multiplayer->Advanced. --Pb 22:39, 9 November 2010 (UTC)

Found an interesting (but useless) property for Medic's Syringe Gun and Blutsauger (And Scout's Shortstop). If Alt-Fire (Right Click) is used with right timing, each of those weapons (Syringe Gun, Blutsauger, Shortstop) can have the auto-reload process stop right after each piece of ammo. For example, you can halt the auto-reload process at 39, 38, etc. for the Syringe Gun and Blutsauger if you use the process described above, the one where you shoot (quickly tap PrimaryFire [Must Release Mouse Button]) and hold down Alt-Fire (Shoot-HoldAltFire-ReleaseAltFire-Repeat). However, the second process described (HoldAltFire + HoldPrimaryFire-ReleaseAltFire (Weapon Fires)-HoldAltFire) reveals interesting (but useless) details. The Blutsauger and Syringe Gun fire 6 syringes before Alt-Fire stops the weapon from firing. The Shortstop fires 2 bullets before Alt-Fire stops the weapon from firing. I am not sure what to make of this information, but it *might* tell us when the Alt-Fire button press registers and executes (If the weapon firing speed for each weapon is considered, an exact time can be estimated). --Pb 23:07, 9 November 2010 (UTC)

Found a shotgun glitch on youtube: Shotgun Glitch. This is a counter bug, I just tried it and reproduced it. The video shows: Engi empties clip, switches weapons, holds PrimaryFire when switching back to shotgun, Engi looks like he fires directly from reserve ammo without the clip counter moving from 0. To reproduce this bug, switching weapons is unnecessary. Hold down PrimaryFire without releasing it. Once the clip reaches 0, the reload animation will show, but the clip counter will remain at 0, and 1 ammo is deducted from the reserve ammo. Sometimes the counter will show what is really happening: the weapon actually reloads because the clip increases to 1 before being fired, but sometimes the counter fails to register the reload. Again, a useless reload counter bug, but I guess I'm compiling them here. --Pb 07:09, 10 November 2010 (UTC)

Please let me know if I'm writing too much about a topic nobody cares about, I couldn't find a "Reload" wiki page, so I'm just adding it all here. Also, everything I have written about I have not copied from anyone else, all of this is through testing and in-game experience. If I have made any errors, please let me know. If I find out more I'll keep adding stuff, but please try it out for yourself. --Pb 06:58, 9 November 2010 (UTC)

Anyone have a resonible explanation as to why this not only occurs but only happens to the pyro and heavy M-NINJA 22:44, 8 November 2010 (UTC)

My guess is that it's tied to the reload speed fix. Valve must've left a loose end somewhere leading to this bug, since it only began after the fix. - Mistah Phoenix 23:12, 9 November 2010 (UTC)

Just out of curiosity, is the reload speed fix you're referring to, "Fixed an exploit where you could reload The Huntsman faster than intended" in the Classless Update? --Pb 00:03, 10 November 2010 (UTC)

No, I'm referring to the recent update that made shotgun reload times the same for all shotguns. It was quite recent, but I don't think it was announced in the patch notes. - Mistah Phoenix 00:39, 10 November 2010 (UTC)

Ok. If you find a source somewhere, please let us know, thanks. Nvm, the info's updated on the shotgun's wiki page, haha. --Pb 07:09, 10 November 2010 (UTC)

• The info isn't completely true. it says the glitch is exclusive to the pyro when the glitch also occurs with the heavy's shotgun M-NINJA 11:37, 12 November 2010 (UTC)

Patched!--Pb 03:41, 20 November 2010 (UTC)

Thats mildly disappointing, i enjoyed the endless reloading sequence :3 -= M-NINJA =- 23:16, 26 November 2010 (UTC)

Heavy's Baby Gun has no trigger

just want to know if I should put the Heavy's Shotgun lacking a trigger and trigger guard in the article. Seems like it'd be a bug, but it's a minor graphical thing. which I remember not supposed to be in the bug or trivia listing. Thoughts? — The preceding unsigned comment was added by BraveTriforcer (talk) • (contribs) 2010-12-07T14:10:48

I don't know what you mean. The w and c models have both triggers and guards. If you're referring to the animations when used by the Heavy, it does have a guard but the trigger cannot be seen due to his large finger.--Focusknock 11:49, 19 February 2011 (UTC)

Engineer's glove is gone in the first person view

I noticed you can see his bare right hand in the first person view when it should be a glove. Don't if this should be noted. -HimitsuNoRyu 02:18, 6 August 2011 (PDT)

Nevermind. It may just be the picture they used on the wiki is wrong or it just looks that way from the lighting. Looking at it in-game, he clearly has his glove

-HimitsuNoRyu 02:27, 6 August 2011 (PDT)

It's not an Ithaca37 at all.

It'll be easier to use images but first and foremost: the Ithaca37 is a bottom ejecting shotgun, it doesn't have an ejection port on either side.

[[5]] [[Inconsistencies1]] [[Inconsistencies2]]

Foxyb (talk) 13:43, 16 March 2014 (PDT)

If you read the trivia correctly, It says it "Appears to be" Doesn't mean it IS. That and the description even mentions of some modifications. Remember, Some guns aren't direct replicas but instead are used as reference. Also the above uploads really aren't necessary. Ashes (talk) 13:46, 16 March 2014 (PDT)

Remington 870 Shotgun, Not an Ithaca M37

Something I feel needs to be changed is the basis behind the shotgun. The Shotgun shows no similarities to the Ithaca M37 other than the pump. On close examination, the Shotgun shows more visual similarities to the Remington 870 Shotgun than the Ithaca.

Reference Pictures for Comparison:

-TF2 Shotgun:

-Remington 870:

-Ithaca M37:

Comparing the two models, it seems that the Remington bears more similarities to the Shotgun, with a sawed off barrel and modified stock, than that of the Ithaca, and could be changed in the Trivia section of the Shotgun to replace the Ithaca M37 or alongside it..

-ColinTheAwesomest (talk) 11:12, 24 July 2014 (PDT)

The shotgun is based off Itacha 37, as can be seen on this image which the concept art is very similar to.

Spython (talk) 00:16, 27 July 2014 (PDT)

Missing Trivia

There are two things about the Shotgun that I think should be in the trivia:

1. The fact that the shells eject from the right side even though the ejection port is on the left side and why that may be.

2. The fact that even though the shells are loaded forward, they eject backwards and why that may be, too. — The preceding unsigned comment was added by Shotgun Dude 37 (talk) • (contribs)

This is just how Valve did the models, it's not notable trivia. --Omolong (talk) 09:29, 16 August 2014 (PDT)

Okay, well was the ejection port being on the wrong side a mistake or a purpose? --Shotgun Dude 37 (talk)

Shotgun Reload Animation

i just noticed a bug of the shotgun, the Engineer animation does not match with the Heavy/Pyro/Soldier.

Links below as proof, see the difference at that black hole on the shotgun of engineer and the Soldier/Pyro/Heavy reload animation:

http://steamcommunity.com/sharedfiles/filedetails/?id=523280541 http://steamcommunity.com/sharedfiles/filedetails/?id=523280793

does this fit on the bugs section Villager3056 (talk) 05:50, 23 September 2015 (PDT)

not really notable. Just a visual thing that doesn't affect gameplay aside from inconsistency. I wouldn't note it. Ashes (talk) 16:33, 23 September 2015 (PDT)

For the recent changes to the damage and function table

Can someone confirm if it is correct?

⇒ Anyar (talk/contributions/giveaway) 10:14, 27 March 2017 (PDT)

I am assuming you meant to change it. See Talk:Scattergun for my explanation. || twolfe's page || twolfe's talk page 10:21, 27 March 2017 (PDT)

Damage and function times tables missing subsequent reload speed times

Right now the table has 1 second for "Reload (first)", but what about subsequent reloads? ⇒ Anyar (talk/contributions/giveaway) 11:50, 28 August 2017 (PDT)

Bullet spread on either this page or the page for the Scattergun (and potentially a lot of other weapon pages) is WRONG

I noticed the page for the Shotgun lists the spread as 15:1 and the page for the Scattergun lists it as 30:1. At first I figured that the Scattergun's spread was actually just narrower for the sake of making Scout less useless at medium range, but I did the testing myself and they're actually identical.

https://youtu.be/eTaPmB5igSA (sorry for the lack of audio and editing)

Other pages for similar weapons, like the Force-A-Nature and Reserve Shooter, list the spread as 30:1. So either the spread for the Shotgun page is wrong, or someone measured the spread for the Shotgun specifically, found out it was 15:1 instead of 30:1, and then either didn't test other weapons or did and then didn't update them.

Can someone who understands Hammer Units better than I do verify this? HitlineMiamo (talk) 12:20, 12 March 2021 (UTC)

Updating to add a list of affected pages.

On these pages, the spread is listed as 30:1. Based on their listed stats, these weapons should all have identical spread to the stock Shotgun, or to the Scattergun, which I tested above and saw it had the same spread. So these ones are either 15:1 or 30:1 but I don't know which.

• Scattergun
• Force-A-Nature
• Soda Popper
• Baby Face's Blaster
• Reserve Shooter
• Frontier Justice
• Widowmaker

In addition, the pages for the Back Scatter and the Shortstop list the spread as 24:1 and 50:1 respectively, but these numbers seem like they're calculated based on the possibly-incorrect spread of the stock Scattergun (30 x 0.8 = 24 for the Back Scatter; the wiki page for the Shortstop calls it about 40% more accurate, but 30 x 1.4 = 42, not 50, so I'm not sure what's going on there).

The page for the Panic Attack has no spread listed at all; although that weapon has the whole increased-spread-on-consecutive-shots thing going on, I think it's still worthwhile to at least list the initial spread (as well as the growth per shot if possible).

If anyone knows a precise enough way to test spread, input would be appreciated so this can get sorted out. HitlineMiamo (talk) 21:52, 15 March 2021 (UTC)

Shotgun spread proof. Stop reverting.

Weapon scripts state that the spread pattern is 0.0625 for both shotgun and scattergun. Plenty of other people have provided proof of this on the talk page. I shouldn't have to provide this proof when the person who made the 15:1 edit in the first place didn't. Shotgunspread.png Blobbles (talk) 01:24, 26 August 2021 (UTC)Blobbles

Could you provide the source of these files? Are they part of the TF2 source code? If so, then this image should be deleted · Ashe (talk) 02:42, 26 August 2021 (UTC)

The source of these files are not leaked source code. They are in the tf/scripts folder and can be decompiled with VICE, valves's own tool. Blobbles (talk) 03:13, 26 August 2021 (UTC)

Furthermore, these weapon scripts were stored uncompiled as plain text until only a few years ago. they have been accessible since the launch of this game, and anyone who installed the game since before this change still has them in .txt format. They were never leaked, because they were always available. Blobbles (talk) 03:42, 26 August 2021 (UTC)

Weapon spreads in general

On another note, it's difficult to imperically determine weapon spread in terms of hammer units deviation from crosshair. I've been considering changing them to the values actually used within the scripts, which is measured in degrees from crosshair, or from projectile spawn in the case of projectiles like the beggar's bazooka.

Blobbles (talk) 01:37, 26 August 2021 (UTC)Blobbles

Spread Clarification

Each weapon is in \steamapps\common\Team Fortress 2\tf\scripts\items\items_game.txt with attributes and the weapon's Item_Class. Item_Class corresponds to a ctx file in \steamapps\common\Team Fortress 2\tf\tf2_misc_dir.vpk. The vpk can be extracted with GCFScape, and the 78 ctx files can be decoded with VICE (both tools via developer.valvesoftware.com/wiki). These are also exposed in wiki patch notes (last known ctx changes visible in December_17,_2015_Patch#Files_changed). The ctx has Spread, and items_game.txt sometimes has attributes for a spread modifier, but how it's all actually used is up to the game.

The wiki Damage table includes pellet spread that is "distance from target" for one unit of "spread on target." Assuming "spread on target" means distance from aim center, it describes a right triangle with length "pellet spread" and width 1 such that rotating around the length edge defines a cone of pellet spread (with a base diameter of 2).

Assume the ctx spread value is the angle of that cone (in radians). This lets us derive the "pellet spread" of the weapon. The calculation is pellet spread = round(1/tan(spread/2)). Specifically: ctx for shotguns says Spread = 0.0675. 1/tan(0.0675/2) = 29.618378775240068116408174736672, rounded that's 30, so "30 : 1" - it works.

That calculation is based on how the right triangle uses half the spread cone angle (spread/2) where the "pellet spread" value is the length of the aim center edge adjacent to the angle (and the other edge is width 1). Trigonometry says tangent(angle) = otherEdge / adjacentEdge AKA adjacentEdge = otherEdge / tangent(angle). Consequently, "pellet spread" = 1 / tangent(spread/2)

Given this clear formula, we don't need the oversimplified calculation Void1702 relied on (which was incorrectly using the value 0.0625 - a typo in Blobbles' discussion which showed ctx image proof of 0.0675. I think Blobbles made the same mistake from the typo when originally setting a value on the pistol, although I don't know what "4% less of the shortstop" came from in the change description).

I extracted the pedigree of all spread weapons this way. Shotguns & scatterguns are 30 (.0675), backscatter is 25 (.0675*1.20), shortstop is 50 (.04), pistols are 50 (.04), miniguns are 25 (.08), tomislav is 31 (.08*.80), SMGs and revolvers are 80 (.025). I plan to list all relevant values in each change summary.

Beggar's includes "projectile spread angle penalty: projectile_spread_angle=3" clearly not the same units - 3 degrees in radians is 0.0523599, but using 38:1 with the label "pellet spread" is a little weird. Panic Attack's worst case appears only known via patch notes "March 28, 2018 Patch #1 Reduced worst-case shot pattern spread by 40%" which doesn't sound like a modifier to its base spread, but a change from before. If we instead interpreted it as 40% increase in spread worst-case, then the 6th shot would be 0.0675*1.40=0.0945 with a "pellet spread" of 21 (a little worse than backscatter).

rotatcepS (talk) 02:22, 25 February 2023 (UTC)

I was digging in to how random spread works and I came up with a setup to measure the angles produced using random spread. I looked through the code and the ctx files and found the function random spread calls the tf2 base github to also come up with a theoretical spread function to compare my results with. An issue I'm having is my results are disagreeing with what is written here and there are a few questions I had that might be able to reconcile the differences.

Are you taking the spread variable to represent the angle between each edge of the cone or is it the angle from the center line to the edge? If it's the former I'm assuming that is why you divide spread by 2 in the pellet spread formula? If it's the latter, why are you dividing spread by 2?

My reading of the ctx file assumed that the value represented the angle from the center line to the cone edge. In the code, the game calls two random floats for x and y that have minimum values of -1 and max values of 1. I'm assuming these are used as scaling factors for the angle, and since the value can be negative the spread variable needs to only cover the angle from the center line to the edge. I automated the measurement system I used so that I could get a large number of trials (a little over 2,000 shots recorded, which came out to about 21,000 individual pellets) and the maximum angle from the center line I was able to find was 0.0675 radians. Assuming that is the case that means that the spread value as it's listed is approximately double what it actually is.

That said, it is entirely possible I missed something since information about this topic is scarce. I want to share my results but I'd rather make sure my theoretical basis is correct, because if its not all of my results are flawed in some major way.

Ceezyr (talk) 18:44, 16 May 2023 (UTC)

Yes, that's the reason I divided by 2. I don't doubt your tests (but video or screenshots will be needed, see below), and they clearly correlate to the number as the right triangle's angle, not the cone's angle.  All my figures and assumptions were based on the assumption that the original 30:1 was a reasonable number, but I have no knowledge of its history.  I mainly just wanted to create consistency on all the pages based on some solid evidence.

If some old-timers could help out with the history of this ratio, that'd be great.  Barring that, I agree, it must be that the shotgun is twice as inaccurate, so the ratio would become 15:1 and the others would be halved, too (must be re-calculated and re-rounded, not just halved).

This, in turn, supports the other controversy where User:Gearhart edited the shotgun to 15:1 without presenting any evidence (or updating related pages), they simply said "^{tested shotgun spread and found out that shotgun spread is actually 15:1, which is twice as much spread as the stated 30:1. And yes, I did go by spread radius instead of diameter.}"  If you or somebody else can provide the evidence of Ceezyr's angle tests, we can put this to rest.

On a related note, I personally think that the 30:1 ratio is a super confusing value to represent spread.  For years I thought it was an "odds that a pellet will hit" number but it's the opposite.

I have been working on a better representation _{(I guess I should discuss on the main discussion page)}: "accurate at 36% range" because with this spread ALL pellets should hit a 25 Hu diameter target (typical hitscan body width https://www.reddit.com/r/tf2/comments/90qklv/comment/e2tc3kw/?utm_source=reddit&utm_medium=web2x&context=3 ) at a range we can calculate: 184.9 Hu which is 36% of mid range 512 Hu.  I believe this example uses the new assumptions (code I was using for accurate range in Hu: "1.0 / Math.Tan(spread) * (0.5 * 25)").

rotatcepS (talk) 22:12, 29 May 2023 (UTC)

I'm not sure if links will work on here, but I put together a google drive folder https://drive.google.com/drive/folders/1vFadC98NJKTdTYBPiJVP2uDY7OOtt7_5?usp=drive_link. In there is a quick writeup (which definitely needs some editing), all of the images, the raw data in a spreadsheet, and the various files and scripts I used to gather the data. There is something weird going on because I was getting repeat shots, which should not happen. I actually did this twice and threw out the first round of data because of that problem but it happened again. I discuss it in the writeup, but basically I either have some flaw in the data collection and processing or there is something weird happening with TF2's random number generator. Assuming the data doesn't need to be thrown it (which I'm fairly certain it's useful despite its flaws) the important values are the min and max angles which came out to just below 0.0675 radians.

As for recalculating, because the angles are small and we're rounding the result anyways the small angle approximation of tan(2 * x) = 2 * tan(x) should work. Once you algebra it out like that it is effectively the same as dividing the spread by 2. I haven't checked any other shotgun weapon or automatic weapon values though so I have no idea if the rest need correcting or not.

The only better representation I can think of is to use the actual angles. I also think it should call out that the spread refers to the random spread, not fixed spread, which was another source of my confusion when I was trying to figure this out.

Ceezyr (talk) 01:38, 31 May 2023 (UTC)

You beautiful bastard, your honorary Masters degree is in the mail.

I especially appreciated "Importantly the max number of decals was set to 9, which I found would cause all random pellets to be shown without showing the accurate pellet" and the RNG overlap weirdness. Good theories.

When I opened the "data" spreadsheet, I found google sheets claiming there are different min/max values in the radians than in your write-up? Regardless, it still supports your conclusion.

I would have calculated the angle based on the position's distance-from-center instead of handling x and y as separate angle calculations. It seems likely it would even better support your conclusion.

If this seems clear enough, we'll want to either update all the ratios to be about half what they show now, or take the opportunity to use a new spread representation.

While the actual angle is the most definitive representation, it doesn't seem super intuitive for readers (maybe why the ratio was introduced). Perhaps a compromise using tooltips like "Spread: 0.0675 radians"

rotatcepS (talk) 04:08, 31 May 2023 (UTC)

Long range pellets

Most pages were listing a range of e.g. 3-30 at long range when the damage per pellet is e.g. 3.168. I think this slipped by mainly because we don't usually shoot at giants. But a giant, or some really rare RNG luck, would hit with all (10 in this example) pellets, rounding to 32 damage total.

1 Pellet	2x	3x	4x	5x	6x	7x	8x	9x	10x
3.168	6.336	9.504	12.672	15.84	19.008	22.176	25.344	28.512	31.68
rounds to: 3	6	10	13	16	19	22	25	29	32

https://youtu.be/Hz3GmGv6YKg shows visible long range damage of 3, 6, 10, 13, 16, and 22 - consistent with this table (instead of multiples of 3).

• I get the same results for Reserve Shooter, Frontier Justice, and Widowmaker
• Panic Attack results are consistent with 38 max at long range
• Family Business results in 27
• Scattergun results in 32 and the same for Soda Popper, Backscatter, and Baby Face's Blaster.
• Force a Nature results in 34
• Shortstop results in 25

I'll list video evidence for each.

rotatcepS (talk) 05:35, 28 February 2023 (UTC)