CHANGE REQUEST to PUGs

  • Hey - turns out IRC is out and something a little more modern has taken it's place... A little thing called Discord!

    Join our community @ https://discord.gg/JuaSzXBZrk for a pick-up game, or just to rekindle with fellow community members.

X-Bomb said:
some are rated but the majority seem to be rated equally that's why my "Amount" is almost 1:1 with my "Rated Amount" on the 6v6 objectives taken line.

http://utassault.net/pugstats/?p=pexplrank&pid=120&gid=1
Click to expand...

Yes, it's fair to say there are maps in pug which were never played in the league 2-3 years ago, so all the objectives for these have a rating of 1 (e.g. on AsthenosphereSE). All the current objective ratings can be seen via the map stats page (you'd have to click into each map); I'm considering making a simple objective stats view later on.

All the league maps should have more representative ratings between objectives - of course as soon as I make the multiplier change across the board it will be more even.

In answer to DJ - yes I'm planning an archive/reset for January, but may not have the new system in place in time... Still working the mechanics out, but I may freeze the 2012 data tomorrow or Tuesday, then take a copy of it and see what the multiplier changes will do to it later on, so we can see both side by side (maybe), then look at re-calculating rank for the Jan stats (easy to do) onward.

/edit

ugh - Caught up with thread, and now I'm finding it hard to filter through the crap from the non crap... I don't mind people having opinions on what stats should look like, and yes I agree the "ranking" element is really putting an unnecessary skew on the game, so after the re-calc if people are still moaning, it seems the most sensible thing to do would be to just code it out... Will save a lot of CPU crunching!

But, just a side note, I'll be deleting any irrelevant posts just so I can see what I'm supposed to be doing! Please don't take it personally. If there's any relevant info in your post which has been deleted, please PM me and I'll sift it out.
 
Last edited:
In answer to DJ - yes I'm planning an archive/reset for January, but may not have the new system in place in time... Still working the mechanics out, but I may freeze the 2012 data tomorrow or Tuesday, then take a copy of it and see what the multiplier changes will do to it later on, so we can see both side by side (maybe), then look at re-calculating rank for the Jan stats (easy to do) onward.
__________________
Click to expand...

Nice 1 Timo and thanks for all time and effort you put in with the stats.
Including XBs ratings with obj.

happy new mega year UTA - sandy
 
you only wanna take stats into 2013, not ranking, or pugs will be prone to all kinds of stuff.


<m-prom> everyone setting hammers on deso to get stats pts ...
<m-prom> nice 8 mins attack
 
proms example is bad, rules for pug are listen to tactics ;) you don't listen to tactics because of stats then you get banned anyway. if i tell hudson and prom on TS to setup a launch on desolate then im pretty sure they will set it and no one else will.

you haven't really been playing since the stats came out seas, from what I have seen nothing has changed. the only thing which has changed is some people trying extra hard to get objectives on warmup! :rofl:
 
Hey Timo!

It looks like the stats has been reset. Nice 1 :):)

So i had a look at the stats now cause there has so far only 1 pug beeing played
But i dont understand all the numbers and stats and maybe you or anyelse can explain them for me. Its easier to watch statistics now also as there only been one pug pld.
After more matches are played i think it will be harder to analyse it.

Thats was from last night and stats u can see here.
http://www.utassault.net/pugstats/?p=rank&

Salva is nr1 but after watched the stats summary / explaination i dont really get whats stat shows. I compared sphere twnz and salvas and mercy and prom stats from this pug.

Salvas detail:
http://www.utassault.net/pugstats/?p=pexplrank&pid=119&gid=1

Twnz detail
http://www.utassault.net/pugstats/?p=pexplrank&pid=118&gid=1

Spheres detail
http://www.utassault.net/pugstats/?p=pexplrank&pid=124&gid=1

I thinks some things are weird here which i need explanation for.

1:
Salvas RANK 876.68 on Main Page "RANKINGS"
But hes details shows TOTAL 1252.40
Should not these 2 be the same? otherwise how is RANK figured out. I thought the Total was to be the same? How does this work?
Mercy and prom got same under Details as in "RANK" table but not rest of players. Why?
Do you understand what i mean? Just look at Salvas Details link above and see TOTAL. Then compare it to RANKING ladder. Mercy and prom got same amount but rest of players doesnt.

2:
Savals fragging shows under details:
Frags 190 300 57000
Deaths 180 -100 -18000
Taken from
http://www.utassault.net/pugstats/?p=pexplrank&pid=119&gid=1


But when u watch the Main Page Salva actually fragged 227
http://www.utassault.net/pugstats/?p=uta_match&matchcode=11pHySng

Why arnt these the same? Also sphere and twnz had wrong killing factor under details. Looks strange to me.

3:
twnz lost ALOT of point on a penalty? This penalty took almost half his points. He almost had same as salva before this penatly. What does this penalty mean? he was late on first map? U can see this under twnz details.

Total 111801
Divided by game minutes 97 1152
Penalty for playing < 100 minutes 1152 -0.5 -576

4:
WHy does mercy get so low ranking. he only pld 3 matches.
I can see now why under his details THis is cause he subbed and he also got a penalty. he lost almost all points due to this penalty. Cause he should be higher up in rankings due that we divide his score only with 17 minutes. This way he can end up as top 1 even if he plays only 3 matches/attacks/defends. But the penalty takes 90% of his points which make this bit weird. No1 will sub cause u lose so much of the stats. Otherwise mercy would be in middle rank table not at bottom. Makes no sense to me.

Multiplier * -0.9 is not good here. No1 will ever sub.
mercy should b middle table at least.

http://www.utassault.net/pugstats/?p=pexplrank&pid=114&gid=1


Total 15400
Divided by game minutes 17 905
Penalty for playing < 30 minutes 905 -0.9 -815
Total 90.59

5:
The weird thing with mercy also is that his TOTAL under details 90.59 is the same thats shown under RANK.
90:59 on both places. Thats should be the same case for every player.

6:
Also prom had same TOTAL under details as under RANK. He also lost 50% of his points due to penalty not pld all pug. Rank should NOT take this is mind imo the penalty multiplier.

Ulv had NOT the same TOTAL under details as under RANK.

7:
Salvas obj now whos this value on the RATED AMOUNT. Does this includes the new XBs rating of each obj summed up to a total?
THis looks really promising then. :)
Which makes the value more fair for obj.

Assault Objectives 3v3 23 20.9 3000 62701

8:
Are suicides that u make to get better spawns included as minus in the stats?

*************************

As i said i think its easier trying analyse the stats after 1 game now cause you can see whats looks weird. After 2-3 games its impossible. These was just some things that i think looked weird.
Plz explain them for me Timo or any1 else cause i think they look a bit dodgy at the moment. Or is it only me and there is an actual explanation for why its shown as it is.

//DJ
 
Last edited:
:lol: so many questions, Ill answer a couple so Timo can spend more time answering the rest :D

Dj_AL said:
But when u watch the Main Page Salva actually fragged 227
http://www.utassault.net/pugstats/?p=uta_match&matchcode=11pHySng
Click to expand...

This statement would only work if frags = kills, you are confusing the two, for benefit of stats they arent the same one is kills and one is kills minus suicides.
A better question would be what type of suicide or kill doesnt count towards frags, because kills-suicides doesnt actually = frags for everyone.

Penalty kicks in to stop ppl playing few games and falsely getting a high score, once he's played a few more games then he'll lose that off the total afaik. You are looking at total stats and applying the info to individual matches, when there has been only 1 match you expect it to be the same but itll show up some of the other elements of how stats tries to even out things.

Dont know about the total of the others, anyone without a penalty has a total score of about 30% short of their match total, so there is a factor there somewhere.
 
Last edited:
I kinda thought it would be interesting to make the analysis now after 1 game as its way easier understand how the individual score is linked together with "RANK".

Still feels weird with the penalty %age though imo.

After 2 games there will be way harder understand why its shown as it does.
Well see what Timo says about it,

Thanks.
:)
 
Last edited:
Originally Posted by MeetYourMaker
Penalty kicks in to stop ppl playing few games and falsely getting a high score, once he's played a few more games then he'll lose that off the total afaik.
Click to expand...

That explains all about penalties. Looks like this work as supposed to then.

Still some questions though.
 
Correct, I have only archived the stats, and begun work on making the views work with the archived data.. This is a LOT of work having to re-write all of the SQL queries to use variables, so will be a while before I move on to the next steps (as I previously illustrated).

DJ raises some good questions, so I'll do a first run at answering them:-

1: Rank differences between detailed and summary
- As far as I'm aware they should be the same, so I'll look at what's causing this.

2: Frags vs Kills
- Frags are Kills minus suicides and team kills (where appropriate), and calculated as such. Frags will never be less than 0.

3: <100min Penalties
- This is to prevent someone playing one match, getting lots of kills, with no/low suicides/deaths and therefore shooting right to the top of the table. By design. As soon as people play 2-3 pugs, they move out of the penalty zone.

4: As above. Remember, ranking is over all matches, not just this one.

5: As 1

6: As 3. Not going to be changing this behavior.

7: New ratings?
- Not yet, as per previous posts. Step 1: Archive, step 2: fix views, step 3: test new ratings, step 4: apply new ratings to 2013 data and recalc ranks.

8: Binded suicides counting in stats
- I'm not certain - I couldn't remember if the LAS tweak to stop a binded suicide from being counted in the scoreboard is being translated to stats. Need confirmation from MyM (or whoever has played recently) whether they are counted on the scoreboard, then I'll work out how to ignore them from stats (if that's possible).
 
Last edited:
If suicides > kills in a given map, then frags stays at 0, not go to minus figures.

This explains what you are seeing, and also suggests that binded suicides are indeed counted.

Example:

Map 1: NoMercy has 2 kills, 3 suicides. His frag count is 0.

This would then translate to 24 kills, 11 suicides = 14 frags.

Attaching raw data to illustrate.
 

Attachments

  • rawdata.png
    rawdata.png
    19.4 KB · Views: 15
Last edited:
Ok, so I just need to work out if we can see the difference in the server logs in order to alter the import routine.

/edit Looks like Cratos had already captured different suicide types.

We do indeed capture binded suicides, so I can probably just add some lines into the ranking bit to have a "Tactical suicide" count which takes from the normal suicide count.
 
Last edited:
Thanks for all answers and that clears things up a bit on my 8 diffrent questions.

Would be really nice to fix my point number 1 and 8

1: Rank differences between detailed and summary
just to make sure the system works. Atm dunno why it looks like it does.

8: Binded suicides counting in stats
Is it possible to change that binded sui is not included? Would be really nice. Should not get minus for wanting a better spawn tbh.

By the way there is a pug going on atm so individual stats + RANK will change so my old exemples will now be irrelevant. But my number 1 will still be shown strange i think.
:)
 
Last edited:
Well, I'm still digging, but after looking at our 7 year old development/change log thread, I picked out the following points:-

* Suicides removed from the ranking (no points reduce any more)
* Multiplier of 'Death' slightly reduced
* Assault Objectives added (hammers, launches, assists). They start logging if teamsize >= 2v2
* Objectives Multipiler now depends on Teamsize. Formula: (2*DefTeamsize - AttTeamsize)*1000
* Objectives can be rated. This is a multiplier to rate important/unimportant objectives

So it may be the case that suicides are correctly removed in the summary rank, but not in the explained rank. I'm still debugging stuff to confirm this, and whether it accounts for any/all the point differences.

/edit

Well, still unsure about the suicides, but after much debugging I've found why the extended view doesn't match... There seems to be a penalty all the way up to 5 hrs of game play which isn't shown on the extended view, and also why this was never ever spotted before:

Debugged salva and spotted this:
Code: 
4632-2d-rank:
 New Rank = (Rank/GameTime)*600 = (429.666670067 / 186) * 600 = 1386.02151635

--
4632-2e-gametime:
 Gametime = 186

--
4632-2f-penalty:
 Penalty = Rank - 30% = 970.215061442

So, now I know where the penalties are done, I can quite easily remove or tweak them and mirror the changes to the extended view... 5hrs seems a little harsh, so perhaps just mirror the extended view and have it for only <100 mins (just under 2 hrs)?
 
Last edited:
Oh nice 1 :cool:

1:
Btw Timo have u changed anything now cause it looks like Salva and twnz and pogo top 3 players atm has same TOTAL as it is under RANK.
It was 70% of the TOTAL earlier but now looks as it is the same. Perfect!

THis was not the case before. Total = RANK point as it should be !

2:
Also see that there is a penalty for less than 200 minutes. Is that something new also. Thats also a good thing though :)

Penalty for playing < 200 minutes 1386 -0.3 -415
Total 970.22
 
Last edited:
You must log in or register to reply here.
Top Bottom
Back