Jump to content

Mech Performance Metrics


6 replies to this topic

Poll: Best Performance Metrics (12 member(s) have cast votes)

Best Performance Metrics

  1. Kills / Death (4 votes [18.18%])

    Percentage of vote: 18.18%

  2. Kills / Match (1 votes [4.55%])

    Percentage of vote: 4.55%

  3. Damage / Match (5 votes [22.73%])

    Percentage of vote: 22.73%

  4. Damage / Kill (7 votes [31.82%])

    Percentage of vote: 31.82%

  5. Damage / Time (2 votes [9.09%])

    Percentage of vote: 9.09%

  6. Kills / Time (2 votes [9.09%])

    Percentage of vote: 9.09%

  7. Other (1 votes [4.55%])

    Percentage of vote: 4.55%

Your personal best Damage / Kill ratio

  1. < 200 (4 votes [33.33%])

    Percentage of vote: 33.33%

  2. 200 - 249 (2 votes [16.67%])

    Percentage of vote: 16.67%

  3. 250 - 299 (1 votes [8.33%])

    Percentage of vote: 8.33%

  4. 300 - 349 (1 votes [8.33%])

    Percentage of vote: 8.33%

  5. 350 - 399 (1 votes [8.33%])

    Percentage of vote: 8.33%

  6. 400 - 449 (0 votes [0.00%])

    Percentage of vote: 0.00%

  7. 450 + (3 votes [25.00%])

    Percentage of vote: 25.00%

Your Best Time / Kill (Min:Sec)

  1. < 2:30 (6 votes [50.00%])

    Percentage of vote: 50.00%

  2. 2:30 - 2:59 (2 votes [16.67%])

    Percentage of vote: 16.67%

  3. 3:00 - 3:29 (0 votes [0.00%])

    Percentage of vote: 0.00%

  4. 3:30 - 3:59 (0 votes [0.00%])

    Percentage of vote: 0.00%

  5. 4:00 - 4:29 (1 votes [8.33%])

    Percentage of vote: 8.33%

  6. 4:30 - 4:59 (3 votes [25.00%])

    Percentage of vote: 25.00%

  7. 5:00 - 5:29 (0 votes [0.00%])

    Percentage of vote: 0.00%

  8. 5:30 - 5:59 (0 votes [0.00%])

    Percentage of vote: 0.00%

  9. >= 6:00 (0 votes [0.00%])

    Percentage of vote: 0.00%

Vote Guests cannot vote

#1 veri745

    Member

  • PipPip
  • 31 posts

Posted 05 April 2013 - 08:26 AM

I ran some numbers based on my per-mech stats calculated the following metrics:
  • * Damage per kill
    • Personal range: 352 - 696
  • Damage per match
    • 200 - 525
  • Kills per match
    • 0.4 - 1.1
  • * Time played per kill
    • 4:48 - 18:00
  • Damage per time played
    • 15 - 40 dmg/min
Based on how much I like playing each of my mechs and how effective I think they are, I think the best performance indicators are:
  • Damage / kill (lower is better)
  • Time / kill (lower is better)

Thoughts from others?

Edited by veri745, 05 April 2013 - 08:29 AM.


#2 veri745

    Member

  • PipPip
  • 31 posts

Posted 05 April 2013 - 11:23 AM

View PostForestGnome, on 05 April 2013 - 09:52 AM, said:

So you made a poll to circlejerk heavy/assault kd's?

What about those lights who get 8 assists per match and >600 damage?


You may be right. I don't really play any lights beside my commando 2D. I don't see how damage/kill is inherently a heavy/assault metric, but I can see why you would thing that about kills/time.

One of the reasons I created the thread/poll is that I don't think the classical Kill/death ratio is a particularly good metric for MWO.

What metric do you think is important for lights?

I personally hate how an "assist" is just any amount of damage on a target that dies. I would much prefer a "kill assist" to mean a significant amount of damage to a mech or particular component.

Edited by veri745, 05 April 2013 - 11:24 AM.


#3 Phaesphoros

    Member

  • PipPipPipPipPipPipPip
  • 513 posts

Posted 05 April 2013 - 11:27 AM

Next time you produce statistics, please provide (weighted) average and standard deviation (as well as sample size), if not raw data :)

#4 TruePoindexter

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,605 posts
  • Facebook: Link
  • Location127.0.0.1

Posted 05 April 2013 - 11:55 AM

I don't think any of those are useful metrics. My most successful mech has the lowest Damage/Kill ratio simply because it kills with precision burst and doesn't net much overall damage. An SRM boat by contrast would have a very good ratio due to the massive damage dealt across an enemy from missile spread.

#5 Iron Frost

    Member

  • PipPipPip
  • Legendary Founder
  • Legendary Founder
  • 87 posts

Posted 05 April 2013 - 11:57 AM

Damage Done / Damage Taken, Kills / Damage Taken, Damage Taken / Win, Damage Taken / Loss would be much more interesting to me.

#6 veri745

    Member

  • PipPip
  • 31 posts

Posted 05 April 2013 - 12:01 PM

View PostTruePoindexter, on 05 April 2013 - 11:55 AM, said:

I don't think any of those are useful metrics. My most successful mech has the lowest Damage/Kill ratio simply because it kills with precision burst and doesn't net much overall damage. An SRM boat by contrast would have a very good ratio due to the massive damage dealt across an enemy from missile spread.



I don't know if you followed my post. Damage per kill is a "lower is better" metric. i.e. 10 kills with 2000 damage dealt is much better than 10 kills with 4000 damage dealt.

I consider this metric to be about accuracy and damage efficiency.

Edited by veri745, 05 April 2013 - 12:02 PM.


#7 aniviron

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,752 posts

Posted 05 April 2013 - 02:43 PM

damage/kill is also only a valid metric if you are doing Zellbrigen; with teammates, it gets pretty fuzzy. My HBK-4P has an amazing damage/kill ratio; this is partly because I have good tracking aim and can shoot for 120 damage without overheating, meaning that one or two shots to the ct can take someone down. But it's also because I scan for targets in it, and if I see anyone with an exposed torso I immediately find them and blow that section out. High damage, high precision shots makes for excellent kill-cleaning/poaching.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users