There was that whole period last year where people discovered the SG and AUG were actually quite good, and then Valve nerfed them into the ground again.

Several years ago, the UMP was insanely accurate when tapping/bursting, and it was a great half buy and anti-eco gun. That got nerfed, too. The MP5 behaves somewhat similarly, but it's not really worth buying over the MP9 in a serious match.

There was a recent discovery with the Nova, where the first shot spray was found to be consistent, and you could hit headshots consistently from distance by aiming to the right of the head. That one's pretty fun, although the situations where you'd want that over a deagle seem niche.

I think the biggest missing with this site is that it's only comparing raw DPS/rate of fire numbers and not spread recovery. A big part of it is the "feel" of bursting/tapping guns while momentum cancelling, when in scenarios where you can't full spray.

eg. Clicking on the head with a P250 at medium range isn't super accurate, but if you manage to momentum cancel perfectly and trace, you actually have a pretty decent chance vs full armor. But I could also buy a five seven, which has better armor pen, accuracy, and sprays better.

To elaborate, they didn't make it into the initial release as there's a substantial amount of work involved to tie that in dynamically with the data (as each shot is inherently dependant on all the shots before it and the time between them when firing multiple rounds).

I understand how much value this would add to the comparisons, so it's a top priority!

If I perfectly strafe and momentum cancel between every shot with a pistol, how often do I dink an opponent if my crosshair's on their head with a P2K, P250, CZ, and five seven? How different are the recovery periods where I'm strafing before I can tap again?

Edit: before I got into CS I played a lot of Delta Force and got pretty good at it. I was very proud of my screenshots where I won matches with a 30+ point lead. My weapon of choice there was a heavier machine gun with a model I don’t remember. It wasn’t scoped and didn’t have great accuracy but due to the size of the hit boxes I could still spray and pray pretty efficiently at snipers from a large distance. I was never fast enough at close quarters but the fact that you could see a dot that you knew was a sniper and shoot them down before they got the chance to move while you stay on the move really did help.

I feel there's enough variation there (especially for "eco rounds" were people can't afford the top weapons) to find value in comparing them.

Nonetheless, you are correct on the majority and for sure that will be reflected in the usage of the site - a lot of the comparisons are going to be "Can the M4A4 kill with 1 shot to the head at range X" or similar.

It externalises the action, and makes the match into a "show" to watch, which I would say would be naturally more appealing to viewers.

Another interesting option would be to give spectators a free camera to view, I don't think that anything technical difficulty ia there in most game engines from providing a free, unrestricted camera to spectators.

What sort of analysis of map positioning and strategies are you thinking about?