I didn't update the config this time around, I just never benchmarked after the gun mettle update.

2639 frames 11.816 seconds 223.34 fps ( 4.48 ms/f) 13.420 fps variability

Last update^

was using same settings and PC as before

2639 frames 11.478 seconds 229.91 fps ( 4.35 ms/f) 13.286 fps variability

about to release a benchmark for the no-explosion smoke script

2639 frames 11.524 seconds 229.00 fps ( 4.37 ms/f) 13.135 fps variability

2.5% Improvement for me, would recommend the no-explosion script.
I used the .vpk one here http://www.teamfortress.tv/post/449092/no-explosion-smoke-script

https://i.imgur.com/P2TOEZO.png

I'll test again when I get home, but their's no way performance would've dropped by 20%. It would've been noticeable in everyone of my configs.

No. This is exactly how I do it.

Pls move to this thread.
http://www.teamfortress.tv/thread/25328/comanglia-s-config-fps-guide

I really hope some people sign up to do this.

Would love to see video about when to offclass, and when to stop offclassing.

2639 frames 15.886 seconds 166.12 fps ( 6.02 ms/f) 7.576 fps variability

2639 frames 16.154 seconds 163.37 fps ( 6.12 ms/f) 12.263 fps variability

2639 frames 15.901 seconds 165.96 fps ( 6.03 ms/f) 11.996 fps variability

You don't have real proof of that. (you could argue maybe on the price point, but it's certainly not the best in pure performance)

I have tested with the exact same settings as infinity on my stability config with an i5 AND had a lower clock rate, with the same plugins, the same GPU even, but on Win 7 an OS that get's slightly less FPS than Win 8.1.
Only difference that I can immediately discern between our 2 tests is that I made sure Google chrome was closed, and even then that wouldn't account for a 40fps difference especially when I have a lower clock rate.

http://www.teamfortress.tv/post/431875/comanglia-s-config-fps-guide

-edit-
Hell you can just look at Phunk's result. http://www.teamfortress.tv/post/467937/tf2-benchmarks
i7, with a lower clock rate, using a Max Quality config with shadows on with -dxlevel 98
get's basically the exact same FPS. The pentium is by and far not the best in terms of performance.

I have a detailed step by step guide on that thread.

You have A LOT of redundancies in your cfg.

"cl_localnetworkbackdoor "1"
cl_muzzleflash_dlight_1st "0"
cl_muzzleflash_dlight_1st "0"  
cl_localnetworkbackdoor "0"

"
^ just 1 example, I also noticed ragdolls were turned off twice.

hud_deathnotice_time "1"

Was I wrong in my assumption people wanted to be able to see who died for more than 1 second?

r_queued_post_processing "1"

as has been noted in one of my thread this command causes issues for people who do use post processing and have AMD cards

cl_threaded_bone_setup 0 

You have this under audio, this command is related to adding a thread for model bone animations. I'm also curious as to how you had better fps with this off than my FPS config for quality PCs.

-edit-

showhitlocation "1"

about 90% sure this is for CSS and CSGO only

I've only played once on a server that does this, but prec didn't exec on round start.

$100 for a new PSU? LOL, hmm actually he might have to pay shipping but still

[redacted]

[redacted]

[redacted]

[redacted]

You could also make the case that playing on Red makes blue players more likely to play overly aggressive, or that red players might play to passively against Blue players.