Jump to content

Match Score Breakdown (And Help Wanted) For Science!

General

88 replies to this topic

#41 Bilbo

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 7,864 posts
  • LocationSaline, Michigan

Posted 24 August 2015 - 09:20 PM

View PostMcgral18, on 24 August 2015 - 06:31 PM, said:



Each "Capturing Base" tick is worth 2 points, and the Capture Assist is worth 1.

How many ticks are there? Why would anyone cap a base for anything other than a win? I certainly wouldn't bother for the score. It kind of puts the lie to "match_score = your_contribution_to _win"

#42 riverslq

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 443 posts

Posted 24 August 2015 - 09:23 PM

View PostLynx7725, on 24 August 2015 - 08:35 PM, said:

Not enough data at this point really. Scout looks to be good, and we don't have info on TAG and NARC, and in some sense those also count towards Counter ECM and Counter ECM Locked Damage. So there's still a point to it.

I still enjoy narcing , tagging, and lighting everything up in sight, but if i'm not getting points for it. guess i believed the 'role-playing' we were promised ;)


is there a better way of submitting data from scouting after matches besides a screen cap or a manual entry?

#43 Mcgral18

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • CS 2019 Top 8 Qualifier
  • CS 2019 Top 8 Qualifier
  • 17,987 posts
  • LocationSnow

Posted 24 August 2015 - 10:00 PM

View PostBilbo, on 24 August 2015 - 09:20 PM, said:

How many ticks are there? Why would anyone cap a base for anything other than a win? I certainly wouldn't bother for the score. It kind of puts the lie to "match_score = your_contribution_to _win"


I think the tick is roughly every 10 seconds? A little under that.

#44 Mcgral18

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • CS 2019 Top 8 Qualifier
  • CS 2019 Top 8 Qualifier
  • 17,987 posts
  • LocationSnow

Posted 24 August 2015 - 10:28 PM

View PostKevjack, on 24 August 2015 - 07:00 PM, said:

Saw a guy, in a loss, get 1 damage and 1 assist.

Still got 23 points.


All of the Proximity Protection (every 6 seconds?) give 2 points, which add up fast if he's around his dying allies without shooting.

Here's the early version of a table, with the likely stats from our testing:
Posted Image

The Kill Most Damage Dealt seemingly being bugged, in that it actually removes 1 point from a Hit and Run (but not from a Brawling).

If someone could explain what's going on here, it would be welcome. In regards to the KMMD taking away one of the Hit and Run.
In this set, you see 20 points from a win, 40 points from damage, and 5 points split between a KMMD, 3 Hit and Runs, and an Assist
Spoiler


The Assist is worth 3, which leaves 2 points left. 3 H&Rs should give you 3 points, yet here it only gives you 2 (because the KMMD removes 1).


Similarly, these two examples should be clearer:

Spoiler


Note the Win (20) +Damage(10) and the Assist(3), that comes out to 33. The Hit and Run adds the last point, to 54.

Spoiler


Note this test is essentially the same, but adding a bit more damage and a KMDD.
20 for the Win, 30 for the Damage (to 50). And the Assist to 53...but where did the Hit and Run go? And the KMDD does nothing? In fact, it seems to remove the 1 point from the H&R.



Seems odd to me.

Edited by Mcgral18, 24 August 2015 - 11:55 PM.


#45 Lynx7725

    Member

  • PipPipPipPipPipPipPipPip
  • 1,710 posts

Posted 24 August 2015 - 10:50 PM

View PostMcgral18, on 24 August 2015 - 10:28 PM, said:

In this set, you see 20 points from a win, 40 points from damage, and 5 points split between a KMMD, 3 Hit and Runs, and an Assist

Wait there's something wrong. Your Team screen says you did 80 damage but your Individual screen says you did 118 damage.

#46 kapusta11

    Member

  • PipPipPipPipPipPipPipPipPip
  • Little Helper
  • Little Helper
  • 3,854 posts

Posted 24 August 2015 - 10:57 PM

So the main chunk of score is damage? I hope it's not used in PSR calculations.

#47 Mcgral18

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • CS 2019 Top 8 Qualifier
  • CS 2019 Top 8 Qualifier
  • 17,987 posts
  • LocationSnow

Posted 24 August 2015 - 11:01 PM

View PostLynx7725, on 24 August 2015 - 10:50 PM, said:

Wait there's something wrong. Your Team screen says you did 80 damage but your Individual screen says you did 118 damage.


Fixed

#48 Lynx7725

    Member

  • PipPipPipPipPipPipPipPip
  • 1,710 posts

Posted 24 August 2015 - 11:26 PM

View PostMcgral18, on 24 August 2015 - 11:01 PM, said:

Fixed

1st one I managed to overshoot to 68 exact, using the following:

Win +20, Assist x3, Dam x0.5, KMDD x2, H&R x1.

So, 20+3+40+2+3 = 68. 2nd one accurate on 34; 3rd one went out, I got 56 when PGI called 53.

But with my Assist at x3, I can build a case for KMDD to replace the Assist score rather than H&R.. has that been tested?

EDIT: Actually, have we firmed on the Assist? You have at x2, I have at x3.

Edited by Lynx7725, 24 August 2015 - 11:28 PM.


#49 Mcgral18

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • CS 2019 Top 8 Qualifier
  • CS 2019 Top 8 Qualifier
  • 17,987 posts
  • LocationSnow

Posted 24 August 2015 - 11:44 PM

View PostLynx7725, on 24 August 2015 - 11:26 PM, said:

1st one I managed to overshoot to 68 exact, using the following:

Win +20, Assist x3, Dam x0.5, KMDD x2, H&R x1.

So, 20+3+40+2+3 = 68. 2nd one accurate on 34; 3rd one went out, I got 56 when PGI called 53.

But with my Assist at x3, I can build a case for KMDD to replace the Assist score rather than H&R.. has that been tested?

EDIT: Actually, have we firmed on the Assist? You have at x2, I have at x3.


Assist is supposed to be 3 (I'll go fix that, typed it in wrong).

Posted Image
Posted Image

Here, a Match Score of 98, 79 from damage+win, leaving 19.


Solo Kill(8)
KB(4)
Component (2)
Brawling(1)
H&R(1)
Protection (2)2

20, so perhaps a number is wrong, but it's consistently only when there's both H&R and KMDD, when tested with Brawling , this didn't happen.

Can you explain the second pair of pictures? Where there's the missing 1 point (1 point less when removing win+damage) with KMDD added?

#50 Lily from animove

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • The Devoted
  • The Devoted
  • 13,891 posts
  • LocationOn a dropship to Terra

Posted 25 August 2015 - 12:14 AM

View PostMcgral18, on 24 August 2015 - 05:24 PM, said:


Only difference I see is Assist (and I haven't tested all of the Protected family, but Protection Proximity was worth 2, and likely all), which we found was to be 3.

Solo Kill being worth 8, but Kill Most Damage Dealt being worth 0 (and for each KMMD, it removed a Hit and Run bonus, costing 1 Match Score).


to be tested: make 2 KMDD + 3 H&R, see how many of the hit and runs get cancelled, maybe its generally onyl 1 maybe truly for every KMDD one H&R does not count.

View PostLynx7725, on 24 August 2015 - 08:29 PM, said:

Take note I'm distrusting displayed Damage values, both Damage Done and Team Damage.

This is as we know that the game can go to fractional damage, but displays always show in integer form. We don't know whether the rounding is done before or after formula. So as long as I get close enough (within 1 match score either side) I'm "happy".

Team Damage is a problem under this scheme though.


That si why you make private matches where you have a fixed amount of HP given to your emch and you deliver a specifc amount of damage as well. This way youc an check the damage you gave (6xAC20 on a 120HP HBR CT) is correctly applied 120
on the scoreboard.

View PostBilbo, on 24 August 2015 - 09:20 PM, said:

How many ticks are there? Why would anyone cap a base for anything other than a win? I certainly wouldn't bother for the score. It kind of puts the lie to "match_score = your_contribution_to _win"


ticks have been 23. on a untouched base until compelte capture:

yet given the "events" PGI has given:
Capture Win Capture Assist Capture 50 2500 Capture Pulse 0
25
First Capture

a capture pulse is what I think is a single tick, worth 2.
a capture winseems to e 5. sinc ein one tets we had one guy capping the entire base and the other doing nothing. The one doing nothing got 25 score. When you win an Assault scenario by battling it out, the score is 20. So the capture win is either 25. or ir calculates as 5, by adding 20 for just a win and 5 for the win being due to capping.

what is a bit unknown now is the real value of the "cap assist", and if the "first Capture" as well as what just "capture" means. it needs further testing.

also strange was the situation where mcgral capped the base alone, and the end screen showed not a "capture win" for him, but his teammate Saxie not having done any captures gotten it

And yes Bilbo. matchscore is very low for capping. If someone wants to game the Scores he better goes for damage instead.
PGI should rate action differently in different modes. like Assault/Conquest, damage*10% instead of 50%. And increase score for caps, and people being close to the cap. and cap defending.

but the current way is truly a bit sad in Scores, because a Solokill is worth 14 (solo kill 8, comp. destr. 2, killing blow 4) which equals out to be 28 damage. So you ar eprobably better jsut stripping other mechs than efficiently killing them. especially since this will still grant you an assist worth 3. making the delta only 11 (22dmg).

I am not sure if I have screened the single boni of that match:

Posted Image

But accordig to the score we hae so far

605 score was due to damage.
20 a win
at leats 3 kills were solokills, and some KMDD was on the assists.
so thats
at leats 15 due to assists
12 by kills wihtout solo
42 for the solokills.

so 694 scores and probably a bit more came just by damagign and killing. 113 came form others (there were more component detsructions involved and I remember quite a lot flankings.

but 85% of my score here in this extreme example of a matchs was just by blowing up stuff. The entire protections, uav detection, scouting and whatever by this makes just a small amount for the score. basically negliable.

It still stans to see how tags and narcs add, but since they only should go wth lrm's wich hardly anyone uses, I doubt its going to be much.

So gettign T1 plying the scores means: racking up damage, without killing stuff in firts place but trying to finish it off.

A better system would be granting everyone involved in the kill additional score for the HP a mech has left after being destroyed to judge an efficient and quick kill correctly.

Edited by Lily from animove, 25 August 2015 - 12:32 AM.


#51 Mcgral18

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • CS 2019 Top 8 Qualifier
  • CS 2019 Top 8 Qualifier
  • 17,987 posts
  • LocationSnow

Posted 25 August 2015 - 12:17 AM

These numbers don't seem to add up either, but it's close.

KMDD could use more testing, I guess. Here's a match from tonight that had 2 unknown variables (Flanking and Lance Formation).

Nothing spectacular, but it has mainly basic variables:

Spoiler



Quote

484-damage(377 rounded up)
107-win(20)
87-Prox 22
65 (?LANCE FORMATION?)
65 (?Flank?)
65(Scout 8)
57-Compo 12
45-KB(12)
33-assists(21)
12-SK(8)
4

Leaves 4 Lance Formations, H&R, 4 KMDD


If we assume KMDD also eats up Flanking, this would mean the Lance Formation are 1 each, and make up each point in this match.



Still doesn't make much sense, and if someone could get a logical number for KMDD that doesn't eat other variables, that would be great...testing can be accommodated.

Edited by Mcgral18, 25 August 2015 - 12:18 AM.


#52 Lily from animove

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • The Devoted
  • The Devoted
  • 13,891 posts
  • LocationOn a dropship to Terra

Posted 25 August 2015 - 12:42 AM

View PostMcgral18, on 25 August 2015 - 12:17 AM, said:

These numbers don't seem to add up either, but it's close.

KMDD could use more testing, I guess. Here's a match from tonight that had 2 unknown variables (Flanking and Lance Formation).

Nothing spectacular, but it has mainly basic variables:

Spoiler





If we assume KMDD also eats up Flanking, this would mean the Lance Formation are 1 each, and make up each point in this match.



Still doesn't make much sense, and if someone could get a logical number for KMDD that doesn't eat other variables, that would be great...testing can be accommodated.


best way to test is again private:

get 5 people, 1vs 4(those having same class, to prevent other protection side bonusses)

get a lance formation bonus, on the team of 4 then suicide the one guy. Should show a clear number of what a lance formation is worth.


repeat this, but let one guy get a KMMD.

now you can compare the scores of

regular KMMD only
Lance formation only
lanceformation + KMMD

If the Score of Lance+KMMD is off from what the "onlys" showed it would explain it similar to the KMDD + H&R discrepancy.

Edited by Lily from animove, 25 August 2015 - 01:19 AM.


#53 ApolloKaras

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,974 posts
  • LocationSeattle, Washington

Posted 25 August 2015 - 12:44 AM

View PostLily from animove, on 25 August 2015 - 12:42 AM, said:


bets way to test is again private:

get 5 people, 1vs 4(those having same class, to prevent other protection side bonusses)

get a lance formation bonus, on the team of 4 then suicide the one guy. Shoudl show a lear number of what a lance formation is worth.


repeat this, but let one guy get a KMMD.

now you cna compare the scores of

regular KMMD only
Lance formation only
lanceformation + KMMD

If the Score of Lance+KMMD is off from what the "onlys" showed it would explain it similar to the KMDD + H&R discrepancy.

That would definitely isolate it.

#54 Revis Volek

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 7,247 posts
  • Google+: Link
  • Facebook: Link
  • LocationBack in the Pilots chair

Posted 25 August 2015 - 10:20 AM

I will be around tonight if you guys need more test subjects....Hit me up.

#55 Deathlike

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 29,240 posts
  • Location#NOToTaterBalance #BadBalanceOverlordIsBad

Posted 25 August 2015 - 03:02 PM

View PostLily from animove, on 25 August 2015 - 12:14 AM, said:

a capture pulse is what I think is a single tick, worth 2.
a capture winseems to e 5. sinc ein one tets we had one guy capping the entire base and the other doing nothing. The one doing nothing got 25 score. When you win an Assault scenario by battling it out, the score is 20. So the capture win is either 25. or ir calculates as 5, by adding 20 for just a win and 5 for the win being due to capping.

what is a bit unknown now is the real value of the "cap assist", and if the "first Capture" as well as what just "capture" means. it needs further testing.

also strange was the situation where mcgral capped the base alone, and the end screen showed not a "capture win" for him, but his teammate Saxie not having done any captures gotten it


With regards to capping (and I'm not talking about the match score), the rewards are scrutinized a little differently.

Capping Bonus - Just touching a cap that is able to change its status (in your favor) gets you that bonus. This repeats over 5 or 10 second intervals though IIRC to accumulate the bonus again. I believe this applies to Assault on the same way.

First Capture -
When a cap point is unowned/neutral (based on the initial starting caps), anyone on the cap when the conversion is successful gains this bonus.

The other time this might happen I think is if you cap the opfor's initial cap... this needs testing, but I'm kinda sure on this.

You can max out on 3 (or 4, if I'm correct) on the First Capture bonus, but most likely on average, you'll get 1 or 2 when driving a Light/fast mech.

Capture -
Basically, after the First Capture bonus is removed, you get a smaller/regular bonus.


There's also a Defensive Kill, where you kill someone that is on the cap recently (applies to Conquest and Assault) while you still have ownership of the cap (so it's always true on Assault, but not so much on Conquest if a cap was converted to neutral or owned by the opfor - I'm unsure if that's totally true on Conquest).


So, these are other potential match score factors.

Edit:

I forget Capture Assist... it's just a cap win for either Conquest+Assault as long as you've actually touched a cap point once (as long as it accomplished something - however little - it excludes fully owned/cap locations).

Edited by Deathlike, 25 August 2015 - 04:07 PM.


#56 Tarogato

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Civil Servant
  • Civil Servant
  • 6,557 posts
  • LocationUSA

Posted 25 August 2015 - 04:36 PM

Am I drunk or are you referring to "Kill Most Damage Dealt" as KMMD instead of KMDD?

#57 Mcgral18

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • CS 2019 Top 8 Qualifier
  • CS 2019 Top 8 Qualifier
  • 17,987 posts
  • LocationSnow

Posted 25 August 2015 - 04:39 PM

View PostTarogato, on 25 August 2015 - 04:36 PM, said:

Am I drunk or are you referring to "Kill Most Damage Dealt" as KMMD instead of KMDD?


Sometimes


I haven't edited all the wrong ones out, nor do I intend to. 3 in the morning might not be the best time to type these.

#58 Deathlike

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 29,240 posts
  • Location#NOToTaterBalance #BadBalanceOverlordIsBad

Posted 25 August 2015 - 04:41 PM

View PostMcgral18, on 25 August 2015 - 04:39 PM, said:


Sometimes


I haven't edited all the wrong ones out, nor do I intend to. 3 in the morning might not be the best time to type these.


For Science, errors will be made.

#59 Mcgral18

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • CS 2019 Top 8 Qualifier
  • CS 2019 Top 8 Qualifier
  • 17,987 posts
  • LocationSnow

Posted 25 August 2015 - 10:37 PM

In regards to the CW non-dying score...it seems normal.

View PostMcgral18, on 25 August 2015 - 10:31 PM, said:


Looks to me like it's the regular formula...almost. I made sure to use 1 mech, then calculated everything from the testing I'd done earlier.
Posted Image

I did the arithmetic...and found two issues. The Match Score came out to 1677, but my calculations to 1676. Pretty close, BUT, that didn't include the 20 points from the Win (and the 11 Hit and Runs cancelled out from the Kill Most Damage Dealt).

Otherwise, the match used the regular match scoring system. It seems death is the thing that changes the formula, significantly.
Posted Image
Spoiler


As for death...maybe I'll do the math another time. Normal match is -2 (and you only get 1 death), but it's without a doubt more significant in CW.


Only abnormality is the 1 point difference in my Math, and the 20 points from the win.

Anyone else have single death CW samples?

#60 Lily from animove

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • The Devoted
  • The Devoted
  • 13,891 posts
  • LocationOn a dropship to Terra

Posted 26 August 2015 - 12:17 AM

hey Mcgral, I probabyl won't be online in MWO the next days, because Starbound got an update

if oyu ahve stema, you can add me so I can hop over to MWO if I am online.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users