13 replies to this topic

[Intcel]

I noticed the weapon info are all over the place.

The heat info seems to make no sense, being a flat value without any further explanation and weapons with same heat value actually generating different amounts of heat.

Weapon cooldown also seems to make no sense. I stopped the cooldown times after trigger as well as after shot duration and it is never what the weapon info claims.

Weapon damage I cannot reliably test, but I lost my faith in any of the numbers slapped on by PGI.

[stjobe]

http://mwo.smurfy-net.de/

One guy outdid PGI by leaps and bounds for vital information about PGI's own game.

Have a look, it's got all the numbers you'll ever need for MWO, and it's parsed directly from the game client.

[Y E O N N E]

One thing you should note is that lasers begin their cool-down as soon as you pull the trigger in Testing Grounds, but in a real match they don't begin cool-down until after they complete their burn duration.

[Lyoto Machida]

Yeonne Greene, on 21 August 2015 - 01:02 PM, said:

One thing you should note is that lasers begin their cool-down as soon as you pull the trigger in Testing Grounds, but in a real match they don't begin cool-down until after they complete their burn duration.

Oh, that's nice and consistent...

[Y E O N N E]

Lyoto Machida, on 21 August 2015 - 01:04 PM, said:

Oh, that's nice and consistent...

Isn't it?

My lovely 5 small laser Locust seems like it would be pretty awesome in TG, but the fire rate is too slow in a real match. Boo-hiss!

[Intcel]

Yeonne Greene, on 21 August 2015 - 01:02 PM, said:

One thing you should note is that lasers begin their cool-down as soon as you pull the trigger in Testing Grounds, but in a real match they don't begin cool-down until after they complete their burn duration.

WTH?
So testing stuff yourself in a neutral environment goes out the window too? Well at least I can still count on developers completely lacking common sense....

stjobe, on 21 August 2015 - 12:54 PM, said:

http://mwo.smurfy-net.de/

One guy outdid PGI by leaps and bounds for vital information about PGI's own game.

Have a look, it's got all the numbers you'll ever need for MWO, and it's parsed directly from the game client.

Yea I used to use that, but I am suspicious about 3rd party info. Especially when stuff got changed. But I guess even equipping my mech blindfolded would make more sense than listening to PGI's own numbers.
Though if they don't even know their own values you have to ask yourself how competent their meddling with the game can be in the first place.

[Light-Speed]

Intcel, on 21 August 2015 - 01:57 PM, said:

WTH?
So testing stuff yourself in a neutral environment goes out the window too? Well at least I can still count on developers completely lacking common sense....

Not so much on common sense as the lack of time/ability to fix the bug fast.
Many studios for other multiplayer games are larger then PGI by far. Also PGI is pretty new to the gaming industry. So.



As for the issue of weapon damage, that might likely just be hit reg or high ping.

[stjobe]

Nightingale27, on 21 August 2015 - 03:56 PM, said:

Many studios for other multiplayer games are larger then PGI by far.

No. Most game developers are smaller than PGI; only a handful are larger.

This PDF from the Canadian Entertainment Software Association shows that
54% of Canadian game studios are very small (5 people or less)
Only 13% are 100 people or more.

Nightingale27, on 21 August 2015 - 03:56 PM, said:

Also PGI is pretty new to the gaming industry. So.

15 years is "pretty new"?

Stop making excuses for them.

[Escef]

Intcel, on 21 August 2015 - 01:57 PM, said:

Yea I used to use that, but I am suspicious about 3rd party info. Especially when stuff got changed.

Smurfy pulls their data straight from the game files, it isn't 3rd party info.

[Vlad Ward]

Wait, other nerds still play this game? Sweeeeet.

Couple things:

1) Weapon heat generation is affected by both Mech Quirks and Pilot Perks. Be sure to take these into account when calculating expected values.

2) Heat Capacity is determined by the number and types of heat sinks equipped on your Mech. There are up to 10 Heat Sinks included in your Engine by default, with the exact function being Engine Heat Sinks = floor(Engine Rating / 25). After and including Engine Rating 275, the engine gains a slot for additional heat sinks every 25 Engine Rating.

3) Double Heat Sinks are weird. When a Mech is equipped with the Double Heat Sink upgrade, all Engine Heat Sinks grant 2 dissipation per second and 2 heat capacity. All additional Double Heat Sinks (including ones added to engine slots at ratings 275-400) only add 1.4 dissipation per second and 1.4 heat capacity.

The formula for Base Heat Capacity is thus as follows

For Standard Heat Sinks, C = 30 + X

For Double Heat Sinks, C = 30 + (2 * E) + (1.4 * D)

Where X is the total number of Standard Heat Sinks (Engine and chassis mounted), E is the number of Engine Heat Sinks (explained above), and D is the number of chassis-mounted Double Heat Sinks.

4) Lasers start their cooldown after they finish firing. Your Mech will also not dissipate any heat while a laser or other extended-fire weapon is firing. This includes Flamers, but not Machine Guns which generate no heat and do not halt heat dissipation. Side note: This is why Chainfire, a mode designed to reduce heat generation, is actually awful and people should really stop using it already.

The heat value given for lasers is the total heat the laser will generate over the course of its burn, whatever that burn time is. It's all a linear function. Got a 7 heat laser with a .75s burn time? It will increase your heat by 9.33 heat/second for .75 seconds.

Ballistics and Missiles apply their heat instantly (IS), or in the same manner as lasers (Clan).

5) Training grounds are borked. Don't bother with them for serious testing. The game does all of its calculations server-side while the Training Grounds are entirely client-side. That's not an excuse, but it's broken nonetheless.

6) Keep doubting the files. While Smurfy does pull data from game XMLs, that doesn't mean those XMLs accurately reflect what the server is calculating.

For example, there was an issue back in Beta where Medium Pulse lasers were generating 3/4 of their intended heat. While the game files clearly showed that they generated X heat, the server was treating them as if they generated X heat per second - and with a 0.75s burn time, well, you see where I'm going with this. The players actually found this one, doing exactly the kind of testing you're doing now. Keep it up!

Hit me up if you want to talk methodology or bounce ideas around. I used to test all this crap obsessively when the game was new, but at the end of the day no one else really cared about the results so I stopped bothering, lol. I know I had the order of operations for heat perks broken down somewhere, but it probably ended up purged in one of the forum moves.

Edited by Vlad Ward, 22 August 2015 - 12:26 AM.

[Davegt27]

what did you use to test?

[oneproduct]

stjobe, on 21 August 2015 - 10:33 PM, said:

No. Most game developers are smaller than PGI; only a handful are larger.

This PDF from the Canadian Entertainment Software Association shows that
54% of Canadian game studios are very small (5 people or less)
Only 13% are 100 people or more.

While that may be true, context is quite important here. Most of those game studios probably make tiny games, like web games or cell phone games. None with the same kind of complexity as MWO, and probably not games that need much tuning or games that they will bother to tune even if there are problems. They probably dump out a new games instead of spending much time with old ones.

Games that get a lot of fine tuning are usually either ones that were recently released, ones that have very large companies backing them, or ones that have exploded in popularity so that even small teams have motivation to keep tuning them.

Edited by oneproduct, 22 August 2015 - 04:53 AM.

[ScarecrowES]

stjobe, on 21 August 2015 - 10:33 PM, said:

No. Most game developers are smaller than PGI; only a handful are larger.

This PDF from the Canadian Entertainment Software Association shows that
54% of Canadian game studios are very small (5 people or less)
Only 13% are 100 people or more.


15 years is "pretty new"?

Stop making excuses for them.

In fairness, for the purposes of your specified stats, a person who releases a Flappy Bird clone that was built in a day onto an app store would constitute a "game studio." Thus, these folks and other small team studios building mobile and independent games rightly should form the majority of studios by bulk.

Of course, MWO is a large game with multi-year development cycle, thousands of concurrent players. It's a full-on PC game. Compared to other studios in the same arena, PGI has a mere handful of employees.

[Mcgral18]

Intcel, on 21 August 2015 - 01:57 PM, said:

Yea I used to use that, but I am suspicious about 3rd party info. Especially when stuff got changed. But I guess even equipping my mech blindfolded would make more sense than listening to PGI's own numbers.
Though if they don't even know their own values you have to ask yourself how competent their meddling with the game can be in the first place.

Check them yourself then:

Source directory (generally C:) >Piranha Games\MechWarrior Online\Game

Gamedata>Libx>Items>Weapons>Weapons.XML


There you go, all the weapon stats in the game. Same info Smurfy pulls out.

Edited by Mcgral18, 22 August 2015 - 09:33 AM.