14 replies to this topic

[BigFatGator]

Seen several threads to the general theme: I'm doing xxx damage per match, is this any good?
I also wondered this myself when I started out, figured that if I was doing 200+ point per match then I wasn't sucking. Read that somewhere, but didn't know where that number came from. Decided I should see if I can come up with a reasonable metric for players to shoot for.

While there are many many other measure of effectiveness, and the point that damage per match isn't nearly a be-all and end-all measure is often highlighted, what it does measure is "are you hitting your opponent with ordnance before you get killed or match ends, and if so to what amount". With that caveat, I think it is a reasonable measure of non-scouting, non-spotting effectiveness.

So how much damage is 'good'? To get a measure for this, I screenshotted the final damage report from my last 30 matches and then recorded the damage numbers per mech per class (Light, Med, Heavy, Assault) on the WINNING team- since the team won, I figured this means that a good amount of the damage was effective damage.

The overall damage per mech on the winning team was 310 Points

On a class basis this worked out to:
Assault- 400
Heavy- 340
Medium-290
Light- 210

Looking deeper into the number, some interesting trends showed up. If we divide each class of mech into thirds and look at average damage in each 1/3 group, we see:

Assault 670 / 350 / 180
Heavy 650 / 280 / 70
Medium 560 / 230 / 60
Light 450 / 140 / 60

What do we make of this? The first point is that, if you score as much as or more damage than the middle third of the winning team by class, you probably are at least carrying your weight on the team for that match in terms of damage. The class average is skewed by some top performances in each class, outliers of ownage. To highlight, the average middle third of damage by class on the WINNING teams was:

Assault- 350
Heavy- 280
Medium-230
Light- 140

These seem like reasonable numbers to shoot for in your matches- if you are doing that much damage consistently then you are probably not hurting your team in terms of your damage number.

The second point (remember that all the numbers above came from the WINNING team) is that the lighter the mech, the more that outstanding performance on the winning team deviates from average performance on the winning team. The damage difference between top 1/3 and mid 1/3 damage scores rises from 1.9x with the assault class to 3.3x with lights. Mediums and Heavies were both around 2.3x.

The difference between an average vs. good light pilot is measurably greater than the difference between an average vs. good pilot of other classes.

[Appogee]

Thanks, science is good.

There is an important consideration though which your analysis isn't taking account of...

More damage isn't necessarily better. A good pilot may do less damage but be of higher value to the team because his accuracy means he gets more kills than a bad one.

To use a simple example:

Pilot A can take a 60LRM Stalker and reliably do 600+ points of damage per match. However, that damage will be splattered all over many opponents. He will probably get some kills because in some instances one of his missiles will do the lucky last point of damage which puts an opponent down.

Pilot B can take a twin-Gauss Jager and do 200 points of damage, but he'll kill five Mechs through headshots in that 200 points of damage.

Pilot B is the better pilot from a team perspective, even though he dealt only a third as much damage.

Still, I found your analysis very interesting and it certainly sheds new light on a question we have all asked ourselves. So, thanks again.

Edited by Appogee, 31 August 2014 - 12:17 PM.

[Lukoi Banacek]

It's ONE metric among many and has some significant value imo. Obviously I do not think OP is portraying the data as any more than it is.....that being said, I think 30 matches is a really small sample size AND significant in it's finding only in relation to your Elo, whatever it might be.

Still, it's interesting stuff nonetheless and not dissimilar to things others of us have been doing.

[JigglyMoobs]

Great data.

It also shows that the upper tier of medium to assault players now consistently get >=550 damage per game. Not a surprise given the recent increase in weapons power. Even good lights average around 450.

So it's not a myth folks - good damage now = 450 or above for all weight classes.

Edited by JigglyMoobs, 31 August 2014 - 12:48 PM.

[Lukoi Banacek]

I don't think it shows that at all JM. It shows those scores in singular instances, not player aggregates.

If in his limited sampling of 30 matches, the same thirty players were playing, you could point to those "top third" numbers as averages, but they are not for any one player. Not that I'd argue 450 isn't a reasonable indicator of good damage if one can approach it consistently. I'd say given the vagaries of PGI's connectivity and simple bad luck though would conspire against most people holding that as an "average" if they've played triple digit games in a given chassis.

[JigglyMoobs]

Lukoi, on 31 August 2014 - 12:51 PM, said:

I don't think it shows that at all JM. It shows those scores in singular instances, not player aggregates.

If in his limited sampling of 30 matches, the same thirty players were playing, you could point to those "top third" numbers as averages, but they are not for any one player. Not that I'd argue 450 isn't a reasonable indicator of good damage if one can approach it consistently. I'd say given the vagaries of PGI's connectivity and simple bad luck though would conspire against most people holding that as an "average" if they've played triple digit games in a given chassis.

Lukoi, no you are right. The data alone doesn't show that. However, in my experience, there are a number of people who end up in the top tier in almost every game and frequently have outliers well into the 1000s. It's just a reflection of the new game dynamic after recent changes.

Fore example, if I go into a solo game, during most of the matches match maker would put me in with a bunch of players who are much less experienced, and then on the other team it would be a similar story with one or two experienced players and the rest being new. If you have someone even better like JagerXII, then you'll see that the MM puts Jager by himself on a team with noobs and then puts maybe two or three JigglyMoobs equivalents on the other team against him along with new players. What then happens is that the game becomes a contest between the few experienced players on each team, with all of the newbies acting as basically targets.

But then that creates even more problems because the "good" players now find themselves in a carry hard situation every single game where they have to do lots of work to get that win, so they end up being forced into taking the angriest, most lethal builds around vs the tons of newbie pilots in their LBX Locusts, etc.

Meanwhile, a lot of medium level experienced players have just abandoned the solo queue unless they are grinding mechs because it's very hard to do well when you are stuck between a bifurcated combination of tryhard pilots and clueless newbies.

So the combination of clan mechs, match maker mechanics, division of the solo and group queues dynamics has basically turned the solo queue into a gaming reserve where armed to the teeth veterans come out to shoot newbies barely able to pilot their mechs, before finally having that "boss confrontation" somewhere at the end of the match.

Yesterday I played 5 games on the solo queue. Never fell below 800 damage. Had 3 games where I broke 1000 and 1 game where I also killed 7 players on the other team. There might have been one game where I killed less than 3 players? I don't remember. But, I know that my KD has shot from 1.9 to 2.9 in the space of like 2 weeks because of how the new game dynamics are working out post-ER PPC nerf. So it's gotten crazy.

If you follow JagerXII's stream, last week he had a crazy losing streak where he lost like 10 solo games in a row doing near or above 1000 damage each game.

The real numbers right now for some of the more experienced people playing medium-assault clan mechs in solo queue probably looks a lot MORE than 650 a game. I'd wager it's up there in the 700s or more.

[Nightmare1]

Only 350 for an Assault? That's pretty low, I think. Still, a large part of it depends on your elo; i.e. - your own skill level and with whom you are dropping. For new solo puggers, 350 in an Atlas should be considered decent, I think. For myself, because I know how well I can (and have) played in the past, I am unhappy if I don't score at least 300 in my Lights, 400 in my Mediums, 500 in my Heavies, and 600 in my Assaults. Those are the lowest acceptable numbers for me. Anything below those, and I become irritated with my performance.

All that being said, the damage scores as statistics are a good initiative. What would be interesting, is if PGI provided an elo score for each pilot on our pilot stats page. This would enable us to match our match/damage averages to our respective elo scores to acquire a better representation of "good damage" for each level of skill.

[On1m]

As I general rule of thumb I consider doing your armor value in damage to be a good starting point for measuring how well you are doing.

[dragnier1]

Ping plays a part in player performance as well. I appear to be hitting the target, but with my high ping (some have wondered if the game was even playable with that ping) i am not, unless my opponent and i are walking straight at each other.

[LauLiao]

You also need to take into consideration the composition of the enemy forces. If the OpFor consists entirely of bottom-of-the-weight-class mechs (20 ton lights, 40 ton mediums, 60 ton heavys, 80 ton assaults) and/or the builds are stripped of armor, you have a much smaller hard-cap of potential damage done than if you were fighting against top-of-the-weight-class mechs that have full armor. 100 damage per winning mech could be all it takes to defeat a company of locusts.

[BigFatGator]

Thanks for the comments. Several have said damage isn't everything, and to that I agree. A player who picks his shots and takes out mechs by cleanly burning through side torsos of known likely XL sporting mechs, or hitting the CT repeatedly and not just firing ordnance as the opponent is shielding with an arm, is doing more effective damage than someone spraying LRMs or SRMs all over the place and hitting on random locations. However, I think that in the aggregate with damage numbers for about 80-90 mechs per class, it balances out. That I took the numbers only from the WINNING team helps in that respect I think as well. I never said that damage is the ultimate measure of performance, just that of the imprecise measurements available to us (w/l, k/d, or damage) it is the best of what we have.

The main reason I posted the stats was to dispel the below thoughts for newer and low/mid-experienced players that might be reading this or the may other 'I'm doing xx damage, is that any good" threads.

On1m, on 31 August 2014 - 02:12 PM, said:

As I general rule of thumb I consider doing your armor value in damage to be a good starting point for measuring how well you are doing.

Nightmare1, on 31 August 2014 - 01:57 PM, said:

I am unhappy if I don't score at least 300 in my Lights, 400 in my Mediums, 500 in my Heavies, and 600 in my Assaults. Those are the lowest acceptable numbers for me. Anything below those, and I become irritated with my performance.

You certainly have the right to hold yourself to that high standard, and I salute you for it and hope you drop on my team in solo matches. But honestly, those reflect a very high level of performance compared to the average mech pilot on the average winning team. We should not hold these out there as expectations to people. Newer pilots expecting to score in that range once they 'learn the game' are only going to get frustrated. Very clearly, for average pilots on average winning teams, the mid-200 to low-300 range in damage is pretty solid. For lights, anything above 150 is probably a reasonable contribution that isn't hurting your team.

OTOH, for those of us with more experience that are in games with less experienced players, I completely agree with Jiggly:

JigglyMoobs, on 31 August 2014 - 12:48 PM, said:

It also shows that the upper tier of medium to assault players now consistently get >=550 damage per game. Not a surprise given the recent increase in weapons power.

Even within a ELO bracket that is determined by W/L, there will be a wide variety in skill. If you think you are really good but aren't consistently in the 500+ range with Medium to Assault classes (aside from the occasional brain fart, lolroll, or bad luck) then you are probably not as good as you think. The data for the top 1/3 of the winning team is probably a good measure of excellence vs. being a solid contributor.

Edited by BigFatGator, 01 September 2014 - 01:35 PM.

[Nightmare1]

BigFatGator, on 01 September 2014 - 01:33 PM, said:

You certainly have the right to hold yourself to that high standard, and I salute you for it and hope you drop on my team in solo matches. But honestly, those reflect a very high level of performance compared to the average mech pilot on the average winning team. We should not hold these out there as expectations to people. Newer pilots expecting to score in that range once they 'learn the game' are only going to get frustrated. Very clearly, for average pilots on average winning teams, the mid-200 to low-300 range in damage is pretty solid. For lights, anything above 150 is probably a reasonable contribution that isn't hurting your team.

OTOH, for those of us with more experience that are in games with less experienced players, I completely agree with Jiggly:


Even within a ELO bracket that is determined by W/L, there will be a wide variety in skill. If you think you are really good but aren't consistently in the 500+ range with Medium to Assault classes (aside from the occasional brain fart, lolroll, or bad luck) then you are probably not as good as you think. The data for the top 1/3 of the winning team is probably a good measure of excellence vs. being a solid contributor.

Right, which is why you'll notice that I agreed with the OP's 350 estimate for a new pilot in an Assault Mech. Everything in this game is relative to a pilot's own personal skill. Another reason why I said it would be interesting if PGI provided our elo scores - we could have a skill metric to compare to the average damages and obtain a better idea of what "average" damage is for each level of expertise.

[Gangnail]

Are you having fun playing? Are you trying to get better? If you answered yes then your doing it right

[Asher420]

Match score is important. Lately I've won and lost escort missions with no kills besides the vip mech. Crazy right? Well, if I have my atlas loadout, his main and only job, imo, is to deal as much damage as possible, while soaking up damage for the vip. Even if I die, I just ate 300 or so damage for the vip. If you're ranking in the top half of the winning team, kudos. You're ready for fw.

[Rogue Jedi]

Asher420, on 01 January 2017 - 02:59 PM, said:

Match score is important. Lately I've won and lost escort missions with no kills besides the vip mech. Crazy right? Well, if I have my atlas loadout, his main and only job, imo, is to deal as much damage as possible, while soaking up damage for the vip. Even if I die, I just ate 300 or so damage for the vip. If you're ranking in the top half of the winning team, kudos. You're ready for fw.

I know if you use the search function it brings up old threads so just in case you did not realise, the last post prior to yours was a few months before the introduction of Faction play.

you are replied to a thread created in August 2014, with the last post prior to yours from September 2014.