• 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Rankings should be win based
#1
Yes, I am biased because my team was hurt by the current system prioritizing match points over win-loss record. However, even if my team was not affected by this system, I would still disagree with the practicality of it. Competitive TF2 has always been about winning more matches instead of winning fewer matches but rolling the teams that you beat. That's why teams try to win every match instead of focusing on matches that they have the potential to roll the other team. 

Fast Forward is currently 3-2, yet we sit near the bottom of the rankings at 5th seed. The 3rd and 4th seed are both 2-3, with about a 1.2 point gap between the 3rd seed and 4th seed. Less than 2 rounds on KOTH. Even though The Empire won fewer matches than Fast Forward, my team will have to win 2 matches in a row and maintain a 5-2 record just to get the opportunity to play The Empire for 3rd place. That doesn't make any sense to me; wins matter in playoffs, so why should rounds matter more than wins in regular season? Teams shouldn't be punished for not 3-0'ing every bad team they play while still winning a majority of matches. It's unfair that my team only lost twice in a 6 team division yet we are below 4 teams. Please consider going back to UGC's seeding style in which win-loss record matters more than round difference.
#2
I can agree with Spu that wins should matter more than match points even if I am on a team that benefited from the current rules. Match victories are inherently more important than match points so I see no reason to keep the current system.
#3
I too agree, I was thinking of making this thread earlier in the season but I kinda forgot. If anything I've only heard complains about this match point system and the only people that seem to defend it are the admins themselves. Maybe I'm not talking to the right people but the vast majority of what I've seen do not like the match point system.
Dog is my supplier, talk to him if you want the stuff
#4
Sort by wins then sort by mps
#5
I would also rather see wins out weight points
Inspired — Whenever Pain Seer becomes untapped, reveal the top card of your library and put that card into your hand. You lose life equal to that card's converted mana cost.

Heavy main for the IM team - team
#6
I agree 
Edit: to clarify, I think its extremely dumb that a team that is going 0-5 has even a chance of knocking a 3-2 team out of playoffs, with 2 other teams who are 2-3 playing in the bracket above us?
#7
+1 .
#8
SORT BY WINS AND THEN USE MATCH POINTS AS A TIEBREAKER

if i win more games i am better than you, at the end of the day a win is a win
plug walk
#9
I've been trying to advocate for a change in either match points calculation or just switching to pure W/L to other admins since the season started. There will and should be changes for next season.
#10
I'm going to preface this by saying that I'm for the current system, but it's not a hill I'm willing to die on. I'm typing this as my own opinion on why the current system was put in place/devil's advocate, but if it got changed back I wouldn't be overly affected.

What does winning a regular season game mean? Not much, honestly. It means that on a given day, a given roster of players beat another given roster of players on a given map. I don't think that means all that much, especially if you take the position that regular season only exists to seed playoffs (which is debatable in and of itself, but that's a different topic). Can you say, definitively, that a team is better than another team after a 5-4 victory on a 5cp? Of course not, it'd be stupid to assume that a BO9 that came down to one single round is enough to place a clear gap between teams. It could've come down to a 1% misplay, a bad lag spike, one player having had a shitty day and being in a bad mood, or a freak accident that may never happen again if you replayed the entire BO9 out another 100 times. You could argue that given you'll never play the same match again, you can never definitively say one team is better than another, but the point is that you want to get as close to that as possible, and that's where match score comes in.

Every one of the instances I listed above have the ability to tip a game between two otherwise exactly even teams in favor of another, obfuscating who the 'true winner' should've been. Over an entire match, though (especially a BO9), the overall impact of those occurrences are minimized. A bad lag spike might cause one round to be a fluke, but the chances of that happening multiple times get successively and exponentially lower the more rounds you have. It's a reason why stopwatch is seen as a better 'upset' gamemode, because 1-2 plays of a team over/underperforming could potentially flip a match. Conversely, koth/5cp are harder to upset because it means a team must be consistently better than another team in order to win multiple rounds. This is partially offset by the fact that PL rounds are longer and take more to win them as compared to 5cp where one death on a mid fight can end a round. Regardless, though, I don't think it's incorrect to say that the more rounds a team wins by, the more likely our assumption of 'this team is better than another' is correct. More likely isn't absolute, though, and in the case of very close games, taking a win in the regular season doesn't have any grand implications about playoffs. All of this is supporting evidence for the claim that

cloud Wrote:if i win more games i am better than you

really isn't true, in terms of what the league is looking for out of the regular season (seedings for playoffs, and by extension the chances that a certain team will finish above another team). Consider two teams that are exactly similar to each other: across any number of games, they will inevitably split all of them, with the difference never being more than 1. If they were forced to play a BO3, one team would win; if they play a second, by definition, the other team will win. Can you make any definitive judgment about which team is better off of one game? It'd be foolish to do so. You can make an educated guess after the first game, but you have no way of knowing.

In the case of >> this season, the fact that they are ranked below space force is a fluke. They got lucky and scored a win off KND no other team had the opportunity to get, and if the admins had reversed your seeding in light of this (which would have been a whole new can of worms, granted) I wouldn't have had any complaints. With that in mind, they should have been 1-4 and finished below >> and Empire regardless of wins or mp. I'm mainly going to be looking at your placement vs Empire's, because I think the case of who's actually a better team is a little hazier there.

>> beat Empire, 2-1, the week after Empire lost a significant number of players from their roster. Roster issues are something the league can't account for as much as we would like (in KND's case too, hence why space force is 4th), so we'll ignore Empire's circumstances for now. All three halves of this game were won by the team that pushed first (and were then able to hold their time). Moreover, cap times were all within normal ranges, and the team that failed to push last in all 3 rounds never had more than 2 minutes to push it. Each match has its own flow, and it's very possible that if Empire started pushing instead of defending they'd take the 2-1; we'll never know. We know that >> was, again, more likely to win a potential 4th half off of the fact that they won the 3rd, but payload matches are inherently biased due to side selection (>> was home team) and whether or not >>'s margin of skill over Empire is enough to account for side selection, random chance, and other factors is something we'll never know. So rather than making a definitive judgment about >>'s skill level being absolutely greater than Empire's, RGL chooses to let the match points speak for themselves. >> won the match, and they're rewarded with double the match points as Empire, but they didn't convincingly win the match. Especially once you take Empire's roster issues into account, it supports the fact that the 'better team' is still, for the most part unknown.

----------------------------------brainlets skip to here-----------------------------------------------

Even if you've read all of this, it's possible that you don't agree with the notion that wins aren't absolute skill indicators. So instead of focusing on a single match, let's compare >>'s and Empire's other matches to try to see a bigger picture, and incorporate a larger sample size:

vs. <<: Empire 0.6-2.4, >> 0-3 (empire-favored)
vs. KND: Empire 0-3 >> 0-3 (tie)
vs. Space Force: Empire 3-0 >> 2-1(empire-favored)
vs. The Flow: Empire 3-0 >> 2.4-0.6 (empire-favored)

Against every other team in the league, Empire had equal or stronger games than >> did, and I don't think that's arguable. If we assume, for a second, that Empire had side selection in their match against >> and that was indeed enough to tip the match (which it very well may not have been), and look at the standings afterwards, you see Empire as 3-2 with 8.6 match points, and >> with 2-3, 5.4 match points, more than a full win away. In this situation, there would be no doubt in anyone's mind that Empire was the better team. It is a hypothetical, so we can't glean too much from it, but if you buy into the belief that a 2-1 match could go the other way if it was played again (and >>/Empire definitely could), you can see the drastic changes one round can make.

If >> really is the stronger team, why did Empire outperform them in almost every week of the season? Is the fact that >> took one round, against Empire, one time, enough to account for the 2 koth and 1 pl round that Empire had "over" >> against other teams? Is being the better team for 15 minutes (arguably 7.5) worth being the better team for a whole 3 weeks of the season?

You could make a strong argument saying that yes, it is, and there's a ton of historical evidence on your side if you do feel that way. That being said, though, if can't at least acknowledge the merit that the MP system has in terms of accounting for statistical randomness and rewarding consistent performance, you're either being close-minded or salty, proof of which can be found below:

[Image: MMvzPzY.png]

As a bit of a closing remark, I think the current system punishes dropping rounds on koth far too strongly, and despite the massive wall of text otherwise I do think there is value in being able to pull a definitive win over an opponent, flukes included. I think the emphasis that the current system places on strong performance over winning regular season games is a good thing, but it can still be retooled.
#11
(02-14-2019, 05:46 PM)exa_ | hl.rgl.gg Wrote: I've been trying to advocate for a change in either match points calculation or just switching to pure W/L to other admins since the season started. There will and should be changes for next season.

You know you run this league right
Dog is my supplier, talk to him if you want the stuff
#12
(02-14-2019, 06:04 PM)fruitcup Wrote:
(02-14-2019, 05:46 PM)exa_ | hl.rgl.gg Wrote: I've been trying to advocate for a change in either match points calculation or just switching to pure W/L to other admins since the season started. There will and should be changes for next season.

You know you run this league right

Match points are built into the league system. I have no control over that. With the season still ongoing, it was better to put that discussion for the offseason, rather than enforce changes midseason.
#13
(02-14-2019, 06:04 PM)fruitcup Wrote:
(02-14-2019, 05:46 PM)exa_ | hl.rgl.gg Wrote: I've been trying to advocate for a change in either match points calculation or just switching to pure W/L to other admins since the season started. There will and should be changes for next season.

You know you run this league right


You can't just enact changes in the middle of the season because it would require all the previous matches to be replayed. The importance of an upcoming match depends on the last. 

We have tweaked our MP system to wipe records of dead teams that have taken rounds off you before they died. A dead team should have no impact on teams that are still playing in the season. But if you lose to a dead team, then you lose to a team.

In the next season, lowering the max MP a team can take from you from a loss would be a good thing to look into. So that would mean capping the max MP a team can take from be lower than 1.7MP
#14
bro fuck MP. a win is a win fam
#15
(02-14-2019, 06:23 PM)Aad | hl.rgl.gg Wrote:
(02-14-2019, 06:04 PM)fruitcup Wrote:
(02-14-2019, 05:46 PM)exa_ | hl.rgl.gg Wrote: I've been trying to advocate for a change in either match points calculation or just switching to pure W/L to other admins since the season started. There will and should be changes for next season.

You know you run this league right


You can't just enact changes in the middle of the season because it would require all the previous matches to be replayed. The importance of an upcoming match depends on the last. 

We have tweaked our MP system to wipe records of dead teams that have taken rounds off you before they died. A dead team should have no impact on teams that are still playing in the season. But if you lose to a dead team, then you lose to a team.

In the next season, lowering the max MP a team can take from you from a loss would be a good thing to look into. So that would mean capping the max MP a team can take from be lower than 1.7MP

I forgot the lower divs were not round robin, but as of now the only people to defend the mp system were a few admins (which I mentioned in my earlier post). I don't see why even if almost everyone wants the win loss system you still want the mp system.
Dog is my supplier, talk to him if you want the stuff
#16
people for this change:
spu
spotlight
fruitcup
duck
pain seer
bowl
nyxi
me
exa

people not for this change:
aad

alto was kind of for and against, i really dont feel like reading the virgin essay again
plug walk
#17
(02-14-2019, 06:38 PM)fruitcup Wrote:
(02-14-2019, 06:23 PM)Aad | hl.rgl.gg Wrote:
(02-14-2019, 06:04 PM)fruitcup Wrote:
(02-14-2019, 05:46 PM)exa_ | hl.rgl.gg Wrote: I've been trying to advocate for a change in either match points calculation or just switching to pure W/L to other admins since the season started. There will and should be changes for next season.

You know you run this league right


You can't just enact changes in the middle of the season because it would require all the previous matches to be replayed. The importance of an upcoming match depends on the last. 

We have tweaked our MP system to wipe records of dead teams that have taken rounds off you before they died. A dead team should have no impact on teams that are still playing in the season. But if you lose to a dead team, then you lose to a team.

In the next season, lowering the max MP a team can take from you from a loss would be a good thing to look into. So that would mean capping the max MP a team can take from be lower than 1.7MP

I forgot the lower divs were not round robin, but as of now the only people to defend the mp system were a few admins (which I mentioned in my earlier post). I don't see why even if almost everyone wants the win loss system you still want the mp system.

Personally, I don't really care which system we go with. You can pretty easily exploit both to get the matches you want in a Swiss system.

The only thing MP system has going for it is that it prevents teams that know that they will roll Team X, so they throw rounds & off class to prolong the matches, similar to pugs.

Now that can already unsportsmanlike under both MP/WL systems since you're already "throwing" so MP punishes you for throwing without admins stepping in and WL just requires it be reported to admins.

(02-14-2019, 07:09 PM)cloud Wrote: people for this change:
spu
spotlight
fruitcup
duck
pain seer
bowl
nyxi
me
exa

people not for this change:
aad

alto was kind of for and against, i really dont feel like reading the virgin essay again

If you read my message again I did not explicitly say I was against the change. I really do enjoy throwing games that I'm not going to win so I can get an easier match up the next week when I played in UGC.
#18
(02-14-2019, 07:09 PM)cloud Wrote: people for this change:
spu
spotlight
fruitcup
duck
pain seer
bowl
nyxi
me
exa

people not for this change:
aad

alto was kind of for and against, i really dont feel like reading the virgin essay again

ah yes
the classic 'i'm going to out myself as a fool by making this a popularity contest rather than arguing anything based on merit'

If you just want to state your opinion, you're free to do so, but the admins already know the system isn't popular. I posted what I did because if you think the W/L system is better, you (the collective you, not cloud as an individual) should be able to at least refute some of the points I listed and support your case with literally anything else than 'it's always been done this way', which is bar-none the shittiest argument anyone can make for ANYTHING, followed by 'more people like it.' Spu's actually giving me a competent argument over steam chat to make his case, which is more than anyone here crying that the big bad rgl admins made you do something different.
#19
(02-14-2019, 08:07 PM)Alto Wrote: support your case with literally anything else than 'it's always been done this way', which is bar-none the shittiest argument anyone can make for ANYTHING

excellent strawman, you're really showing yourself to being the intellectual we know you are
plug walk
#20
(02-14-2019, 08:15 PM)cloud Wrote:
(02-14-2019, 08:07 PM)Alto Wrote: support your case with literally anything else than 'it's always been done this way', which is bar-none the shittiest argument anyone can make for ANYTHING

excellent strawman, you're really showing yourself to being the intellectual we know you are

????????? do you even know what a strawman is? I made a claim that the merits of the MP system outweigh the flaws (you'd know this if you actually read I wrote, instead of throwing around buzzwords), to which only one person has even attempted to make a merit-based argument against. no one in this thread has made an actual point about why the W/L system is better, but you apparently think counting the riders on your bandwagon gives you some kind of power. a strawman would imply i misrepresented your argument when your post was literally counting the people who agreed with you. what else would your argument be?


Forum Jump:


Users browsing this thread: 1 Guest(s)