19 replies to this topic

[NST]

I've wondered at the huge difference between damage scores in players.

It seems that the highest scores come from the players surviving through to the latter half of the battle. The interesting point is that the difference isn't just 2x; it's often 3-4x as much damage dealt.

Good players: live longer, more bang-bang, higher accuracy... all true... Except that doesn't explain it working the same with my own scores (I'm a mediocre player at best).

The other mystery is that I can't always reconcile ammo spent, damage per shell, and my average accuracy with the times that I get massive scores.

My theory:

Damage dealt isn't just armor and structure. It's including the damage dealt to weapons & equipment (crit damage). Furthermore, if a whole component (arm / leg / whatever) is destroyed, the health of all the equipment, now destroyed, is added to the damage value.

This, if true, makes damage dealt a pretty terrible stat for judging performance of players, weapons, or much of anything, really.

Any ideas on how to test this? Happy to hear refutations or other theories.

Edited by Egomane, 02 May 2014 - 02:41 AM.

[Hillbillycrow]

Applying damage is a fair metric...even split up and categorized by various ways it's still damage, mostly.
I mean what else would you use?

Edited by Hillbillycrow, 02 May 2014 - 02:55 AM.

[El Bandito]

NST, on 02 May 2014 - 02:34 AM, said:

I've wondered at the huge difference between damage scores in players. It seems that the highest scores come from the players surviving through to the latter half of the battle. The interesting point is that the difference isn't just 2x; it's often 3-4x as much damage dealt. Good players: live longer, more bang-bang, higher accuracy... all true... Except that doesn't explain it working the same with my own scores (I'm a mediocre player at best). The other mystery is that I can't always reconcile ammo spent, damage per shell, and my average accuracy with the times that I get massive scores. My theory: Damage dealt isn't just armor and structure. It's including the damage dealt to weapons & equipment (crit damage). Furthermore, if a whole component (arm / leg / whatever) is destroyed, the health of all the equipment, now destroyed, is added to the damage value. This, if true, makes damage dealt a pretty terrible stat for judging performance of players, weapons, or much of anything, really. Any ideas on how to test this? Happy to hear refutations or other theories.

Damage is a bad way to judge player performance. A meta player might get much less damage per match than a Lurmer or Laser boat but much higher kill ratio due to the nature of the weapons. 2xSRM6 has bigger alpha than AC20, but that AC20 can sure kill more effectively.

Hell, a good Artillery Strike can double the damage dealt by many players in PUGs. See some of those Light pilots bragging about 1000 damage? Strikes might have something to do with it.

Finally, I am not 100% certain but ammo explosion also pads your damage.

Edited by El Bandito, 02 May 2014 - 03:04 AM.

[NST]

If damage dealt is the only thing being counted, it's a great stat... Or at least as good as most of the others.

The problem, if this theory is true, is the artificial padding that's going on.

You're riding your DDC (or whatever), I shoot your torso with an AC2 and blow it off. I miss all the crit rolls, but the component is destroyed, along with the AC20 in it. My damage dealt goes up by 20 -- 18 (total health of the AC20) + 2 (the damage it took to finish off the torso structure).

Is there a decent way to test or find out if this is the case?

[Joseph Mallan]

El Bandito, on 02 May 2014 - 02:59 AM, said:

Hell, a good Artillery Strike can double the damage dealt by many players in PUGs. See some of those Light pilots bragging about 1000 damage? Strikes might have something to do with it.

Though true, it wasn't always the case El.

[El Bandito]

Joseph Mallan, on 02 May 2014 - 03:13 AM, said:

Though true, it wasn't always the case El.

I'm sure. Now though, with the availability of strikes, any 1000+ damage with non-meta/Light/Medium/bad variants etc...are much less meaningful.

Show me a SS of 1000+ damage with an Urbie and I'll be like, "Yep, good job landing that strike". Unless of course, there is a video of it that says otherwise.

Damn I miss that trashcan. Salvaged god knows how many back in MW2 Mercenaries.

Edited by El Bandito, 02 May 2014 - 03:26 AM.

[Joseph Mallan]

El Bandito, on 02 May 2014 - 03:23 AM, said:

I'm sure. Now though, with the availability of strikes, any 1000+ damage with non-meta/Light/Medium/bad variants etc...are much less meaningful.

Show me 1000+ damage with an Urbie and I'll be like, "Yep, strikes".

I was in a match a few weeks ago (Centurion Challenge) Where my Spotter scored 740(ish) damage to my Missile boats 700(ish) No Arty used. Now an Urbie scoring 1,000... Underachiever!

[Deathlike]

NST, on 02 May 2014 - 02:34 AM, said:

Damage dealt isn't just armor and structure. It's including the damage dealt to weapons & equipment (crit damage).

Yes, this was true when MGs were able to dish crit damage, but not able to deal actual damage to internals. People were "thinking" that it did something, when it actually did not. There were long drawn out discussions of that particular matter for a while because people didn't properly make the connection.

Quote

Furthermore, if a whole component (arm / leg / whatever) is destroyed, the health of all the equipment, now destroyed, is added to the damage value.

I'm not sure of that, but that could be true.

El Bandito, on 02 May 2014 - 02:59 AM, said:

Damage is a bad way to judge player performance. A meta player might get much less damage per match than a Lurmer or Laser boat but much higher kill ratio due to the nature of the weapons. 2xSRM6 has bigger alpha than AC20, but that AC20 can sure kill more effectively.

That's correct. That's why I explicitly say that missiles in general pad damage... particularly Streaks. LBX is no different either.

Quote

Hell, a good Artillery Strike can double the damage dealt by many players in PUGs. See some of those Light pilots bragging about 1000 damage? Strikes might have something to do with it.

A good arty strike deals a good ~200 pts of damage alone.... and it's better if it's dropped into a cluster****.

Quote

Finally, I am not 100% certain but ammo explosion also pads your damage.

It does. That is technically why weapons like the Gauss Rifle did 15+ pts of damage on average (in stat tracking) before "crit bonus conversion into internal damage" was instituted. Ammo explosions do inflate the damage scores.

Edited by Deathlike, 02 May 2014 - 03:33 AM.

[Screech]

El Bandito, on 02 May 2014 - 02:59 AM, said:

Hell, a good Artillery Strike can double the damage dealt by many players in PUGs. See some of those Light pilots bragging about 1000 damage? Strikes might have something to do with it.

Nah. It was just much more likely that they were just much better then everyone else. Artillery Boogieman is not the reason for everything. But whatever you do, don't say "Artillery Strike" five times in a mirror, he may come for you.

[Rhaythe]

The stats in the game are notorious liars. There's really no good way to judge how good a player is in this game via stats. Kills are meaningless. Kill assists are almost as meaningless (machinegun Locusts can get 10 or better KA's every match. Dakka dakka). Damage is a little bit of a better indicator, but only in context. High damage, low KAs/Ks means players are just hosing damage downstream and not really aiming. Low damage, high KAs/Ks means either the player aims very well, or is extremely opportunistic about when he engages. Hard to tell.

Stats in this game can prove a player is bad, but can't really prove the opposite conclusively.

[Karl Streiger]

NST, on 02 May 2014 - 03:12 AM, said:

Is there a decent way to test or find out if this is the case?

There is a way:
you have to note the shots that hit and the damage you dealt. bot values you find on the stats page.

After a battle with High Damage - you have to look to that page again - and you have to subtract the values.
in the end you may have for example LBX shots hit the target 20 - damage dealt 220...so you have 110% damage for the lBX that means you have dealt critical damage.

Damage is the worst way to get an idea about player performance...not to mention those 21C-Bills for dmg point

[Greyboots]

NST, on 02 May 2014 - 02:34 AM, said:

I've wondered at the huge difference between damage scores in players.

It seems that the highest scores come from the players surviving through to the latter half of the battle. The interesting point is that the difference isn't just 2x; it's often 3-4x as much damage dealt.

Good players: live longer, more bang-bang, higher accuracy... all true... Except that doesn't explain it working the same with my own scores (I'm a mediocre player at best).

The other mystery is that I can't always reconcile ammo spent, damage per shell, and my average accuracy with the times that I get massive scores.

My theory:

Damage dealt isn't just armor and structure. It's including the damage dealt to weapons & equipment (crit damage). Furthermore, if a whole component (arm / leg / whatever) is destroyed, the health of all the equipment, now destroyed, is added to the damage value.

This, if true, makes damage dealt a pretty terrible stat for judging performance of players, weapons, or much of anything, really.

Any ideas on how to test this? Happy to hear refutations or other theories.

This is EASY to explain.

Weapon.
Ranges.

An AC10 has an "optimal" range of 450m. At 900m it will do 5 points of damage instead of 10. At 1349 it will do 1. Each will register as a "hit" for your accuracy so a "hit" doesn't automatically mean you deal 10 points of damage.

All weapons except Missiles have this damage decay after their optimal range up to their maximum range (Ballistic = 3x optimal range, energy = 2x optimal range). Missiles deal the damage they deal at up to their maximum range at which point they detonate dealing no damage at all (both LRMs and SRMs).

In the Mechlab you can look at the weapon damage graph associated with each weapon and see the damage decay due to range.

GOOD players don't just spam, they know at what point to stop shooting because it's essentially a waste of ammo.

Weapon ranges MATTER in your loadout and, in all likelyhood, this is why you can't reconcile your stats with your damage (well, that and hit registration issues probably help it along on occasion).

Edited by Greyboots, 02 May 2014 - 05:34 AM.

[Sprouticus]

Why wouldnt you count critical damage? Or removing arms (if that does bump damage values).

Both of those things make a HUGE difference in the combat effectiveness of the enemy mechs. A mech without an arm/torso or missing weapons is far less likely to perform anywhere near optimally. Sure, some really good players can use a shield arm/torso to extend their lives and not lose damage, but that is relatively rare, especially at the mid and lower tier Elo's.

I do agree that very good meta players end up with less damage compared to their effectiveness. again at the top Elo brackets or for folks with meta FLD builds damage is less of a direct indicator.

But again, for the majority of us, damage is usually correlative to effectiveness.

[AccessTime]

The key is survival, if you can survive mostly intact to the later stages where the damage you do (and absorb) will mean much more.

300 points of damage in the final stages of a close game can mean the match was won primarily by your efforts, 300 points of damage in the early stages of a game rarely affects the overall outcome.

Edited by AccessTime, 02 May 2014 - 06:32 AM.

[Greyboots]

Sprouticus, on 02 May 2014 - 05:59 AM, said:

Why wouldnt you count critical damage? Or removing arms (if that does bump damage values).

Both of those things make a HUGE difference in the combat effectiveness of the enemy mechs. A mech without an arm/torso or missing weapons is far less likely to perform anywhere near optimally. Sure, some really good players can use a shield arm/torso to extend their lives and not lose damage, but that is relatively rare, especially at the mid and lower tier Elo's.

I do agree that very good meta players end up with less damage compared to their effectiveness. again at the top Elo brackets or for folks with meta FLD builds damage is less of a direct indicator.

But again, for the majority of us, damage is usually correlative to effectiveness.

Component Destruction is accounted for by being a separate reward on top of the damage it took to achieve the destruction of the component (component destrcution = arm, leg, torso or head location destroyed).

Critical damage is accounted for in the damage you deal.

Both of these contribute to your earnings (in cbills and XP).

AccessTime, on 02 May 2014 - 06:32 AM, said:

The key is survival, if you can survive mostly intact to the later stages where the damage you do (and absorb) will mean much more.

300 points of damage in the final stages of a close game can mean the match was won primarily by your efforts, 300 points of damage in the early stages of a game rarely affects the overall outcome.

Just not true. That 300 damage you deal early on is 300 points someone else doesn't have to deal later. This is why Damage accounts for such a large percentage of Match Score.

Just because you get a lot of component destructions or kills doesn't actually make you any good. Live long enough and you'll get them simply because randomly firing at trashed mechs will see some die and others have parts fall off from sheer dumb luck. The guy that only did 250 but opened up an awful lot of left torsos while doing it is the "better player" IMO, being more accurate and creating weaknesses for the rest of the team to exploit before they go down and doing it while under heavy fire because everyone's still got all their parts attached and plenty of ammo for their weapons.

The stats in this game are rather meaningless and that's a huge part of the problem.

Edited by Greyboots, 02 May 2014 - 06:42 AM.

[Lostdragon]

I have had games whete I did less than 200 damage but contributed more to the match than some games where I have done 1000 damage. The amount of damage you do is far less important than putting that damage in the right spot. There are also a lot of things you can do, especially in a light or fast medium, that are just as or more valuable than damage when it comes to securing the win.

[Levi Porphyrogenitus]

NST, on 02 May 2014 - 02:34 AM, said:

My theory:

Damage dealt isn't just armor and structure. It's including the damage dealt to weapons & equipment (crit damage). Furthermore, if a whole component (arm / leg / whatever) is destroyed, the health of all the equipment, now destroyed, is added to the damage value.

I'm pretty sure your theory is correct, with the addendum that ammo explosions credit you with the damage they cause (though I'm not sure if it's actual damage that resulted, or if it's instead total potential damage from the ammo that popped).

[Bhael Fire]

I used to go straight for the kill (and still do if the enemy is aggressive and a high threat), but these days I tend to disarm my enemies first...stripping off one component at a time (mo money! mo damage!).

So...damage can be used as a means of judging player skill...if it's used in conjunction with their other stats like KDR, accuracy, wins/losses, etc.

[NST]

Finally! Got evidence that sort of shows what I'm talking about.

After several matches running around finding and firing the gauss until I got a single hit which destroyed a component, then running around not using the weapon for the rest of the match.

Weapon		Matches   Fired   Hit   Accuracy   TimeEquip	 Damage
GAUSS RIFLE   34		458	 244   53.28%	 03:38:50	  3,529
GAUSS RIFLE   34		461	 245   53.15%	 03:38:50	  3,572

One hit, but much damage (3529 to 3572) = 43 from a single shot.

Maybe that's possible with the crit system?

In other news, the stats aren't recording the match count or time equipped correctly.

I tried some other tests and had a match where I did 613 damage (2kill, 2assist), by destroying the side torsos of a heavily wounded atlas and a highlander using PPCs (2 assists were scribbling MLs across some fast mechs). I actually spent most of the match capping and don't think I could have fired more than 8 2xPPC shots (new mech, no basic efficiencies yet).
Maybe side torso destruction damage includes the attached arm + equipment too?

Someone should try blowing up a side torso (and attached arm) full of only single heatsinks and checking the score after that.

Not conclusive yet, will continue working on it.

There are lots of implications from this, particularly if it feeds the stats used by PGI to run balance. (Meta players live longer, destroy more components, weapons/mechs score higher damage, get nerfed).

As part of a scoring mechanism, this might be reasonable. Blowing stuff up is good for you!

But it's not "damage", and it's not consistent. The match I mentioned above was pretty horrible, I didn't do anything significant, yet that 617 damage score said I was the highest damaging player on the team. It was achieved because people had a lot of equipment in the body parts that I destroyed.

tldr;
Recorded damage might include the HP of all the items (heatsinks, weapons, everything) in a component when it is destroyed.

[Deathlike]

NST, on 02 May 2014 - 09:55 AM, said:

In other news, the stats aren't recording the match count or time equipped correctly.

Yea, telemetry is broken. Alert Paul ASAP.