GG LSP
Our offense couldn't keep up the momentum in the second half.
Our offense couldn't keep up the momentum in the second half.
Forum > Goal Line Blitz 2 > Journeyman Top 15 Ladder Matches August 11th
Laggo
offline
offline
Originally posted by Thunderoo
Winning against a team ranked above you is worth more Elo points
which is what is broken with the random scheduling of inter-tier matches
Winning against a team ranked above you is worth more Elo points
which is what is broken with the random scheduling of inter-tier matches
NiborRis
offline
offline
Originally posted by USC_Trojans
also how does hawaii win vs vase breakers and go down a rank? lol
This would be a lot easier to explain if the elo ratings weren't hidden, because you'd see the rating go up after a win, even if your ranking slipped.
This is the same as how teams in college football can win their game that week and slip a spot or two down the polls.
also how does hawaii win vs vase breakers and go down a rank? lol
This would be a lot easier to explain if the elo ratings weren't hidden, because you'd see the rating go up after a win, even if your ranking slipped.
This is the same as how teams in college football can win their game that week and slip a spot or two down the polls.
NiborRis
offline
offline
Also hawaii was in a weird place in the rankings - before the ladder games there was a 3-way tie for 20th spot overall, Hawaii, Salty, and Warsaw. All three won their games, so the team that played the highest ranked opponent (Salty) is now 20th, and the team that played the weakest opponent (Hawaii) is now 22nd.
Originally posted by Laggo
which is what is broken with the random scheduling of inter-tier matches
Elaborate?
Not sure how getting higher ratings for beating higher ranked teams means something is broken.
which is what is broken with the random scheduling of inter-tier matches
Elaborate?
Not sure how getting higher ratings for beating higher ranked teams means something is broken.
Think he means because it introduces a major LUCK factor. Now, hopefully that luck falls evenly over the course of a season and ends up being roughly the same for everyone...
But if team A gets "lucky" and plays 16 ladder opponents all ranked higher than them....
and team B gets "unlucky" and plays 16 ladder opponents all ranked lower than them....
And both teams were to go 16-0 -- team A's schedule (and skill) would have gotten them much more ELO addition during the season - and there is nothing team B could have done about it.
Personally - I think it MOSTLY evens out over a season and that the system is pretty decent --- though showing the ELO would probably make things a lot clearer (and make the forums more active -- and the ladder threads more active -- as people could bitch, moan, complain and grab their pitch forks over their dislike of the formula!) so I'm for ELO showing.
(edit: somebody smart could probably due a study right now...or at the end of the season - looking at a tier (take Journeyman) - and calculating what the average differential between their ladder rank -- and their opponents' ladder rank (pre-game). It could be interesting and might show something like:
#1 Journeyman - on average played ladder opponents that were +9 to them
#2 - +4
#3 - -7
#4 - +1
#5 - -3
#6 - -9
Etc.
So #1 got lucky with tough matchups and took advantage of that opportunity
but #6 got unlucky with weak matchups...and though they did all they could - they just didn't get the same opportunity.
Again - I think the ladder system is pretty good (30+ rank window is too much) - and you should show the ELO - and you should quickly throw together a ladder matchups listing page.
But if team A gets "lucky" and plays 16 ladder opponents all ranked higher than them....
and team B gets "unlucky" and plays 16 ladder opponents all ranked lower than them....
And both teams were to go 16-0 -- team A's schedule (and skill) would have gotten them much more ELO addition during the season - and there is nothing team B could have done about it.
Personally - I think it MOSTLY evens out over a season and that the system is pretty decent --- though showing the ELO would probably make things a lot clearer (and make the forums more active -- and the ladder threads more active -- as people could bitch, moan, complain and grab their pitch forks over their dislike of the formula!) so I'm for ELO showing.
(edit: somebody smart could probably due a study right now...or at the end of the season - looking at a tier (take Journeyman) - and calculating what the average differential between their ladder rank -- and their opponents' ladder rank (pre-game). It could be interesting and might show something like:
#1 Journeyman - on average played ladder opponents that were +9 to them
#2 - +4
#3 - -7
#4 - +1
#5 - -3
#6 - -9
Etc.
So #1 got lucky with tough matchups and took advantage of that opportunity
but #6 got unlucky with weak matchups...and though they did all they could - they just didn't get the same opportunity.
Again - I think the ladder system is pretty good (30+ rank window is too much) - and you should show the ELO - and you should quickly throw together a ladder matchups listing page.
Edited by TxSteve on Aug 12, 2014 08:57:09
Originally posted by TxSteve
Think he means because it introduces a major LUCK factor. Now, hopefully that luck falls evenly over the course of a season and ends up being roughly the same for everyone...
But if team A gets "lucky" and plays 16 ladder opponents all ranked higher than them....
and team B gets "unlucky" and plays 16 ladder opponents all ranked lower than them....
And both teams were to go 16-0 -- team A's schedule (and skill) would have gotten them much more ELO addition during the season - and there is nothing team B could have done about it.
Well, I mean, you can't always play someone that's the same exact rank as you. While "bad" streaks like that can happen, I'd imagine it's pretty unlikely...and that problem only exists for as long as those teams are underrated.
Think he means because it introduces a major LUCK factor. Now, hopefully that luck falls evenly over the course of a season and ends up being roughly the same for everyone...
But if team A gets "lucky" and plays 16 ladder opponents all ranked higher than them....
and team B gets "unlucky" and plays 16 ladder opponents all ranked lower than them....
And both teams were to go 16-0 -- team A's schedule (and skill) would have gotten them much more ELO addition during the season - and there is nothing team B could have done about it.
Well, I mean, you can't always play someone that's the same exact rank as you. While "bad" streaks like that can happen, I'd imagine it's pretty unlikely...and that problem only exists for as long as those teams are underrated.
Originally posted by Corndog
Well, I mean, you can't always play someone that's the same exact rank as you. While "bad" streaks like that can happen, I'd imagine it's pretty unlikely...and that problem only exists for as long as those teams are underrated.
totally agree.
But could be a case where:
Top 10 seasoned faced journeyman on 8 of their 10 ladder matches
and another faced only 2 of 10 journeyman teams.
I agree with you though. Ladders are imperfect and messy...they have to be to function properly.
I like this year much better than last year in terms of tier preference - I just don't think Sophomores should continue playing only sophomores (it helps some and hurts others in ladder ranking)
Well, I mean, you can't always play someone that's the same exact rank as you. While "bad" streaks like that can happen, I'd imagine it's pretty unlikely...and that problem only exists for as long as those teams are underrated.
totally agree.
But could be a case where:
Top 10 seasoned faced journeyman on 8 of their 10 ladder matches
and another faced only 2 of 10 journeyman teams.
I agree with you though. Ladders are imperfect and messy...they have to be to function properly.
I like this year much better than last year in terms of tier preference - I just don't think Sophomores should continue playing only sophomores (it helps some and hurts others in ladder ranking)
Galithor
offline
offline
I think the ladder's working fine, and that's coming from a guy that's gone 3-9 playing all teams tier +1.
It's been a hell of an advantage to be able to see how our builds are shaping up against what are basically end-game builds.
It's been a hell of an advantage to be able to see how our builds are shaping up against what are basically end-game builds.
USC_Trojans
offline
offline
Originally posted by Corndog
Elaborate?
Not sure how getting higher ratings for beating higher ranked teams means something is broken.
While you're on can you check the Hawaii bulls vs beast town beasts and the roman testudo vs no luck bad rng? It says in progress but they haven't skimmed a single play yet.
Elaborate?
Not sure how getting higher ratings for beating higher ranked teams means something is broken.
While you're on can you check the Hawaii bulls vs beast town beasts and the roman testudo vs no luck bad rng? It says in progress but they haven't skimmed a single play yet.
Edited by USC_Trojans on Aug 12, 2014 10:14:40
Laggo
offline
offline
I think the luck factor is pretty broken, it's kind of worrying that the response seems to be "eh, I don't think it matters too much, it should all balance out right"? Because that's not really how chance works. If you flip a coin 100 times that doesn't mean you are going to get 50 heads and 50 tails. Why does that seem to be the logic here?
It should just be more deterministic when you play a tier play up and when you get to play ahead your rank and when you play below. Like having ELO thresholds where after you pass it your next game is a tier above at a set elo range, so every team that reaches that level (or higher levels) has the same opportunities to test themselves at around the same level of competition.
You should also alternate between playing below and playing up, or something a little better than "it's random". If it wouldn't be a huge pain to see snapshots of teams ladder positions through the season (actually maybe Trojan's threads would be really good for this, I'll check this out) you could probably identify a huge luck advantage for same teams versus others (Hawaii Bulls being the obvious "we got screwed" team).
It should just be more deterministic when you play a tier play up and when you get to play ahead your rank and when you play below. Like having ELO thresholds where after you pass it your next game is a tier above at a set elo range, so every team that reaches that level (or higher levels) has the same opportunities to test themselves at around the same level of competition.
You should also alternate between playing below and playing up, or something a little better than "it's random". If it wouldn't be a huge pain to see snapshots of teams ladder positions through the season (actually maybe Trojan's threads would be really good for this, I'll check this out) you could probably identify a huge luck advantage for same teams versus others (Hawaii Bulls being the obvious "we got screwed" team).
Edited by Laggo on Aug 12, 2014 10:29:26
Originally posted by Laggo
I think the luck factor is pretty broken, it's kind of worrying that the response seems to be "eh, I don't think it matters too much, it should all balance out right"? Because that's not really how chance works. If you flip a coin 100 times that doesn't mean you are going to get 50 heads and 50 tails. Why does that seem to be the logic here?
Because the only thing you're doing is racing to a .500 season.
I think the luck factor is pretty broken, it's kind of worrying that the response seems to be "eh, I don't think it matters too much, it should all balance out right"? Because that's not really how chance works. If you flip a coin 100 times that doesn't mean you are going to get 50 heads and 50 tails. Why does that seem to be the logic here?
Because the only thing you're doing is racing to a .500 season.
Laggo
offline
offline
Don't understand how that is a relevant response to what I just pointed out.
The logic of who you get scheduled against has nothing to do with whether you lose or win those games.
The logic of who you get scheduled against has nothing to do with whether you lose or win those games.
You are not logged in. Please log in if you want to post a reply.
