You know if Luffy is hurting then we're all hurting :(
if you look at my recent posts, it literally is like 2 pages of applied and my updates,
i have 0 interest in this site rn its hard to keep updating and be a whore
01-06-2020, 12:41 PMInf1d3l Wrote: You know if Luffy is hurting then we're all hurting :(
if you look at my recent posts, it literally is like 2 pages of applied and my updates,
i have 0 interest in this site rn its hard to keep updating and be a whore
01-06-2020, 11:34 AMgolden_apricot Wrote: So the tests I did to see if this was dire in the SMJHL showed that a team of all snipers had about a 55ish win rate vs the rest of the league as they are right now. I just swapped PA and SC for players that had higher passing to keep tpe roughly the same. So its not like they win everygame. I also used the lines as they were for the S50 playoffs not optimized to beat a specific team. This was done for the better teams, i didnt test to see if this was the case for some of the lower tpe teams in juniors sadly.
Still not quite sure what you mean by all Snipers, PA 40 or PA 90, or something in between? But assuming you kept the rest of the league as it was, a 55% win rate is hardly game breaking and, depending on how big the sample was, could well be within the realm of statistical randomness. Hell, it's worse than a lot of teams have been doing this season even without the exploit. But then again SMJHL is probably is probably quite different than SHL when it comes to this as the TPE-differences both between individual players and between teams are less pronounced.
Quote:I agree it would be great to see that info, the one that got leaked I believe was chicago going on some stupid win percentage or they led the league or something and people freaked out, well they were a playoff team this year so this might have been an over reaction. (I think they were the team that were used as the "bad team" for testing but im not positive). That being said most of it didnt really show much besides "yup turns out NOLA has like a 65% win rate with everyone on auto lines and high scoring" or a bunch of minute slider tests on each of the versions of sth. Most of the test results are in goalie discord which i left because im not a goalie and i think it would be great to show some of the results. In the end each version added more complicated issues and scoring was still by far the most dominate attribute. I would add images if i could :(
Yeah that's my main issue, all this panic seems to be based on a select few tests with a limited scope or highly experimental setups whereas the more moderating ones are largely ignored. Hell we had three teams with a win rate above 65% this season and one did much better than the team which is using the exploit, with others on equal footing or not far behind. Hamilton didn't break the league this season, exploit or not. I know it's not particularly attractive to give them the benefit of the doubt at the moment and I sure hope they don't end up winning the cup, but the playoff sample size is way too small to draw any conclusions. Sweeping a Wildcard team with a low TPE goalie and now leading 3-1 in a close series against a team you finished ahead of in the regular season? That's also not gamebreaking, no matter how much we might dislike them at the moment.
Quote:Ill add a link here to something I posted on the STH forum. Now I was not involved in the Hamilton tests that led to us lowering passing, I decided on a scorer on my own and lowered passing upon suggestion and reading this link, but I believe this is where this testing started. This took me about 3 minutes to find. Discussion on attributes in STH
Interesting as a starting point for sure, but this is exactly the kind of limited and experimental setups that I mean, with an aboslute Superstar line on a single team, but absolutely no one else in the league adjusting to it. It shows that the exploit does indeed exist, but not if and how it would actually play out in a realistic league scenario where teams will adapt and rules can be made to prevent certain extremes. Hell you could even argue that a line with these three players should probably be scoring at such a torrid pace, and that they are undervalued when using their original attributes.
Quote:In the end sth isnt great for this league as a whole and i think the hive mind of the league made up their mind on hamilton and the passing. Does it infalte their win rate a few percentage points? yes. Does it mean they win every game? i mean they came in third in the league so you tell me.
Yup, agreed.
Quote:As far as the FHM worries thread i get it. The people working on it who were commenting in the thread more or less made it seem like it would be a seamless transition and great for everyone. That is not the case. There are issues with it for sure and soemwhere in that thread a list of them was outlined, mainly
1. Cost of game
2. Complexity of lines
3. Updating system
4. Getting an index out for users to look at
Of these i think 4 has been more or less solved and we are using this as a tool to update the index system from what I can see. I would articulate more but dont want to over promise and under perform.
for 1. this is a major issue but unless a newer version solves problems 3 and 4 for good, i see us using fhm 6 for awhile meaning the cost will decrease over time as is the case for PBE.
Number 2 is still an issue that will separate teams a bit based off how much time they are willing and able to put in, but for the majority of the league i dont think this will affect them too much. I would love for us to have a tactics sub forum on this site to talk things through, but I do not think 90% of the leagues GMs and coaches would use it.
Number 3 is a major time sink IMO. there is no "import function" in FHM so even though you can get all players and their attributes as .csv you cant change that file and re import it. This will increase update time considerably and we might have to go into a system with designated updater for teams that is strictly upheld.
I actually think that those four concerns are all manageable as well. But i think you missed the one point that was actually the most important one for me at least: A new sim engine would almost definitely be better but even if it was, I'm not sure if the league could survive the switch itself. Plus more than the arguments itself, it often was the dismissive tone in that thread that had me concerned. It didn't seem like constructive criticism was actually wanted, but that the purpose of the thread was to tell everyone who has doubts that they are wrong.
Quote:I am more than willing to keep people in the loop if they ask but dont want everything to be out in the open, as i hate the hive mind and I believe it is more counter productive. Should be be giving more updates on this? Yes and I think I will do my best to do that as i work through and finish the SMJHL file and hope that the testers of the SHL file do their best to update the site on the move and what we are doing on that end. The base SHL file with all players and teams in it is almost done from what I can tell (I havent seen it so im not positive on this) but this will then need to be tested in all sorts of ways to see if there is anything game-breaking and that out update system will work.
I quite strongly disagree here. I don't see how opening the actual info and data up to the "hive mind" would be bad in this case, as we are not talking about sensitive league information (which is limited to things like HO for good reason), but tests and numbers. This is exactly the kind of stuff that as many people as possible should be looking at to poke holes into it or make suggestions, which is where a hive mind is actually useful. And if it is true, as you suspect, that the league has already decided to unfairly hate on HAM no matter what, then don't you want the information out in the open that would prove them wrong?
I think it would be good to have one central thread with a main post that the league keeps updating, where links to the various test sims and other sources of information are collected, so that people can conduct their own analysis if they chose to, or take a critical look at the data that is already available. So much of this seems to be floating around somewhere, in private conversations, subforums or Discords, but if you don't know the right people or don't have the time to scan everything, you never get the full picture and are either forced to live in the dark and just believe what you are told, or rely on very incomplete information when you wanna participate in the discussion, and risk becoming a part of the problem by spreading misinformation. It's a very unhealthy situation for a league to be in at a time where, as evidenced by the other responses in this thread, frustrations are high and we need to build consensus and elicit as much public support as possible for the significant and risky changes that will likely need to be taken.
Quote:Again if you have more questions feel free to let me know. Ill do my best to keep you and anyone else informed.
01-06-2020, 01:34 PMRomanesEuntDomus Wrote: Thanks for the explanations and the insight!
01-06-2020, 11:34 AMgolden_apricot Wrote: So the tests I did to see if this was dire in the SMJHL showed that a team of all snipers had about a 55ish win rate vs the rest of the league as they are right now. I just swapped PA and SC for players that had higher passing to keep tpe roughly the same. So its not like they win everygame. I also used the lines as they were for the S50 playoffs not optimized to beat a specific team. This was done for the better teams, i didnt test to see if this was the case for some of the lower tpe teams in juniors sadly.
Still not quite sure what you mean by all Snipers, PA 40 or PA 90, or something in between? But assuming you kept the rest of the league as it was, a 55% win rate is hardly game breaking and, depending on how big the sample was, could well be within the realm of statistical randomness. Hell, it's worse than a lot of teams have been doing this season even without the exploit. But then again SMJHL is probably is probably quite different than SHL when it comes to this as the TPE-differences both between individual players and between teams are less pronounced.
Quote:I agree it would be great to see that info, the one that got leaked I believe was chicago going on some stupid win percentage or they led the league or something and people freaked out, well they were a playoff team this year so this might have been an over reaction. (I think they were the team that were used as the "bad team" for testing but im not positive). That being said most of it didnt really show much besides "yup turns out NOLA has like a 65% win rate with everyone on auto lines and high scoring" or a bunch of minute slider tests on each of the versions of sth. Most of the test results are in goalie discord which i left because im not a goalie and i think it would be great to show some of the results. In the end each version added more complicated issues and scoring was still by far the most dominate attribute. I would add images if i could :(
Yeah that's my main issue, all this panic seems to be based on a select few tests with a limited scope or highly experimental setups whereas the more moderating ones are largely ignored. Hell we had three teams with a win rate above 65% this season and one did much better than the team which is using the exploit, with others on equal footing or not far behind. Hamilton didn't break the league this season, exploit or not. I know it's not particularly attractive to give them the benefit of the doubt at the moment and I sure hope they don't end up winning the cup, but the playoff sample size is way too small to draw any conclusions. Sweeping a Wildcard team with a low TPE goalie and now leading 3-1 in a close series against a team you finished ahead of in the regular season? That's also not gamebreaking, no matter how much we might dislike them at the moment.
Quote:Ill add a link here to something I posted on the STH forum. Now I was not involved in the Hamilton tests that led to us lowering passing, I decided on a scorer on my own and lowered passing upon suggestion and reading this link, but I believe this is where this testing started. This took me about 3 minutes to find. Discussion on attributes in STH
Interesting as a starting point for sure, but this is exactly the kind of limited and experimental setups that I mean, with an aboslute Superstar line on a single team, but absolutely no one else in the league adjusting to it. It shows that the exploit does indeed exist, but not if and how it would actually play out in a realistic league scenario where teams will adapt and rules can be made to prevent certain extremes. Hell you could even argue that a line with these three players should probably be scoring at such a torrid pace, and that they are undervalued when using their original attributes.
Quote:In the end sth isnt great for this league as a whole and i think the hive mind of the league made up their mind on hamilton and the passing. Does it infalte their win rate a few percentage points? yes. Does it mean they win every game? i mean they came in third in the league so you tell me.
Yup, agreed.
Quote:As far as the FHM worries thread i get it. The people working on it who were commenting in the thread more or less made it seem like it would be a seamless transition and great for everyone. That is not the case. There are issues with it for sure and soemwhere in that thread a list of them was outlined, mainly
1. Cost of game
2. Complexity of lines
3. Updating system
4. Getting an index out for users to look at
Of these i think 4 has been more or less solved and we are using this as a tool to update the index system from what I can see. I would articulate more but dont want to over promise and under perform.
for 1. this is a major issue but unless a newer version solves problems 3 and 4 for good, i see us using fhm 6 for awhile meaning the cost will decrease over time as is the case for PBE.
Number 2 is still an issue that will separate teams a bit based off how much time they are willing and able to put in, but for the majority of the league i dont think this will affect them too much. I would love for us to have a tactics sub forum on this site to talk things through, but I do not think 90% of the leagues GMs and coaches would use it.
Number 3 is a major time sink IMO. there is no "import function" in FHM so even though you can get all players and their attributes as .csv you cant change that file and re import it. This will increase update time considerably and we might have to go into a system with designated updater for teams that is strictly upheld.
I actually think that those four concerns are all manageable as well. But i think you missed the one point that was actually the most important one for me at least: A new sim engine would almost definitely be better but even if it was, I'm not sure if the league could survive the switch itself. Plus more than the arguments itself, it often was the dismissive tone in that thread that had me concerned. It didn't seem like constructive criticism was actually wanted, but that the purpose of the thread was to tell everyone who has doubts that they are wrong.
Quote:I am more than willing to keep people in the loop if they ask but dont want everything to be out in the open, as i hate the hive mind and I believe it is more counter productive. Should be be giving more updates on this? Yes and I think I will do my best to do that as i work through and finish the SMJHL file and hope that the testers of the SHL file do their best to update the site on the move and what we are doing on that end. The base SHL file with all players and teams in it is almost done from what I can tell (I havent seen it so im not positive on this) but this will then need to be tested in all sorts of ways to see if there is anything game-breaking and that out update system will work.
I quite strongly disagree here. I don't see how opening the actual info and data up to the "hive mind" would be bad in this case, as we are not talking about sensitive league information (which is limited to things like HO for good reason), but tests and numbers. This is exactly the kind of stuff that as many people as possible should be looking at to poke holes into it or make suggestions, which is where a hive mind is actually useful. And if it is true, as you suspect, that the league has already decided to unfairly hate on HAM no matter what, then don't you want the information out in the open that would prove them wrong?
I think it would be good to have one central thread with a main post that the league keeps updating, where links to the various test sims and other sources of information are collected, so that people can conduct their own analysis if they chose to, or take a critical look at the data that is already available. So much of this seems to be floating around somewhere, in private conversations, subforums or Discords, but if you don't know the right people or don't have the time to scan everything, you never get the full picture and are either forced to live in the dark and just believe what you are told, or rely on very incomplete information when you wanna participate in the discussion, and risk becoming a part of the problem by spreading misinformation. It's a very unhealthy situation for a league to be in at a time where, as evidenced by the other responses in this thread, frustrations are high and we need to build consensus and elicit as much public support as possible for the significant and risky changes that will likely need to be taken.
Quote:Again if you have more questions feel free to let me know. Ill do my best to keep you and anyone else informed.
01-06-2020, 12:51 PMTnlAstatine Wrote: if you look at my recent posts, it literally is like 2 pages of applied and my updates,
i have 0 interest in this site rn its hard to keep updating and be a whore
Quote:Still not quite sure what you mean by all Snipers, PA 40 or PA 90, or something in between? But assuming you kept the rest of the league as it was, a 55% win rate is hardly game breaking and, depending on how big the sample was, could well be within the realm of statistical randomness. Hell, it's worse than a lot of teams have been doing this season even without the exploit. But then again SMJHL is probably is probably quite different than SHL when it comes to this as the TPE-differences both between individual players and between teams are less pronounced.
So I just simply swapped passing and scoring if the players passing was higher than their scoring. Tomen and JY and others just did 40 passing 99 scoring or 70 passing 99 scoring for different teams and saw the outcome. Now I had issues with this in general because it had an inherent assumption that x team should be bad so if this makes them good its broken to some extent like I mentioned with Chicago. 55% over multiple tests relatively consistently (5x500 game seasons) is outside of statistical randomness but not outside of what you might expect in a season. I think in the SHL it might have been higher not positive.
Quote:Yeah that's my main issue, all this panic seems to be based on a select few tests with a limited scope or highly experimental setups whereas the more moderating ones are largely ignored. Hell we had three teams with a win rate above 65% this season and one did much better than the team which is using the exploit, with others on equal footing or not far behind. Hamilton didn't break the league this season, exploit or not. I know it's not particularly attractive to give them the benefit of the doubt at the moment and I sure hope they don't end up winning the cup, but the playoff sample size is way too small to draw any conclusions. Sweeping a Wildcard team with a low TPE goalie and now leading 3-1 in a close series against a team you finished ahead of in the regular season? That's also not gamebreaking, no matter how much we might dislike them at the moment.
Im not touching this one anymore or else Ill piss people off as a Steelhawk myself but I tend to agree with you. Basically a result was released and everyone said that team should be bad and people panicked.
Quote:Interesting as a starting point for sure, but this is exactly the kind of limited and experimental setups that I mean, with an aboslute Superstar line on a single team, but absolutely no one else in the league adjusting to it. It shows that the exploit does indeed exist, but not if and how it would actually play out in a realistic league scenario where teams will adapt and rules can be made to prevent certain extremes. Hell you could even argue that a line with these three players should probably be scoring at such a torrid pace, and that they are undervalued when using their original attributes.
The not trying to adjust was my take from the start. Yes there is an exploit but is there counter play? Now after a season, I still dont know as I have not tried to develop strats to counter it and i doubt gms that have will post results. I dont think there is but I am not positive of this at all. More shots means more chances and goalies either make a save or they dont. With defense being high on literally every player in the league forwards and defensemen have equal opportunity to be in position for the rebound so more shots will lead to more goals on average.
Quote:I actually think that those four concerns are all manageable as well. But i think you missed the one point that was actually the most important one for me at least: A new sim engine would almost definitely be better but even if it was, I'm not sure if the league could survive the switch itself. Plus more than the arguments itself, it often was the dismissive tone in that thread that had me concerned. It didn't seem like constructive criticism was actually wanted, but that the purpose of the thread was to tell everyone who has doubts that they are wrong.
Fair, and this is not something that I have thought of too much tbh. Will numbers go down, ill say probably, but I believe the league will survive since it is so large to begin with. The bottom point again is a fair concern. I left the thread at some point to move on to other things when it got that way or became a "hype" thing. So i dont know what else went on there. It was a good opportunity that some people might have not taken full advantage of.
Quote:I think it would be good to have one central thread with a main post that the league keeps updating, where links to the various test sims and other sources of information are collected, so that people can conduct their own analysis if they chose to, or take a critical look at the data that is already available. So much of this seems to be floating around somewhere, in private conversations, subforums or Discords, but if you don't know the right people or don't have the time to scan everything, you never get the full picture and are either forced to live in the dark and just believe what you are told, or rely on very incomplete information when you wanna participate in the discussion, and risk becoming a part of the problem by spreading misinformation. It's a very unhealthy situation for a league to be in at a time where, as evidenced by the other responses in this thread, frustrations are high and we need to build consensus and elicit as much public support as possible for the significant and risky changes that will likely need to be taken.
Top point I disagree with but I do think a subforum would be a good thing and Ill bring this up come testing time for sure. As far as the top part people in general have an inherent bias (read SHL games or playoff threads) and I think that they will look at the results rather than the data as a whole and what each part means. A sub forum that is actually moderated to get rid of the shit posts and other crap i do believe could be useful however.
01-06-2020, 02:41 PM(This post was last modified: 01-07-2020, 06:10 PM by DeletedAtUserRequest.)
I said this to the head guys a few weeks back that we need like right now to be proactive and begin focusing on the site and keeping interest up while we make the move ...think outside the box and try and find ways to engage the base of members in ways we havnt tried before... cause we’re bleeding members and activity right now... they all got defensive or ignored it... who ever is at the wheel.. it’s time to step up or step aside.
I’ll say it again... time to be creative and make some changes that don’t include FHM (Note: Cory this is not about you, don’t get defensive)
01-06-2020, 02:41 PMMike Izzy Wrote: I said this to the head guys a few weeks back that we need like right now to be proactive and begin focusing on the site and keeping interest up while we make the move ...think outside the box and try and find ways to engage the base of members in ways we havnt tried before... cause we’re bleeding members and activity right now... they all got defensive or ignored it... who ever is at the wheel.. it’s time to step up or step aside.
I’ll say it again... time to be creative and make some changes (Note: Cory this is not about you, don’t get defensive)
We’re currently discussing a requirement of a pass first player on lines 1 2 and 4, as well as somewhere in pairing 1 and 2. A maximum gap of 20.
01-06-2020, 02:41 PMMike Izzy Wrote: I said this to the head guys a few weeks back that we need like right now to be proactive and begin focusing on the site and keeping interest up while we make the move ...think outside the box and try and find ways to engage the base of members in ways we havnt tried before... cause we’re bleeding members and activity right now... they all got defensive or ignored it... who ever is at the wheel.. it’s time to step up or step aside.
I’ll say it again... time to be creative and make some changes (Note: Cory this is not about you, don’t get defensive)
We’re currently discussing a requirement of a pass first player on lines 1 2 and 4, as well as somewhere in pairing 1 and 2. A maximum gap of 20.
01-06-2020, 03:05 PMKeygan Wrote: We’re currently discussing a requirement of a pass first player on lines 1 2 and 4, as well as somewhere in pairing 1 and 2. A maximum gap of 20.
Keygan![[Image: x17WALp.png]](https://i.imgur.com/x17WALp.png)
||![[Image: uNh8ZtE.png]](https://i.imgur.com/uNh8ZtE.png)
![[Image: luff.png]](https://media.discordapp.net/attachments/755517543564837398/1086808363372920953/luff.png)
![[Image: luff2.png]](https://media.discordapp.net/attachments/694020619620188191/1088174131452854313/luff2.png)
![[Image: crtigger.gif]](https://sig.grumpybumpers.com/host/crtigger.gif)
![[Image: murphy.png]](https://media.discordapp.net/attachments/832982603623432192/880210179436716052/murphy.png)
![[Image: OnGNB1G.gif]](https://i.imgur.com/OnGNB1G.gif)
![[Image: cgv4vCv.png]](http://i.imgur.com/cgv4vCv.png)
|![[Image: 95lCCDx.png]](http://i.imgur.com/95lCCDx.png)
|![[Image: KgwtJeY.png]](https://i.imgur.com/KgwtJeY.png)
|
|
![[Image: LB6bY06.gif]](https://i.imgur.com/LB6bY06.gif)
![[Image: tanner.png]](https://media.discordapp.net/attachments/694020619620188191/924086646914289734/tanner.png)
![[Image: sIjpJeQ.png]](https://i.imgur.com/sIjpJeQ.png)
![[Image: sig-e-e-geck-atl.png]](https://cdn.discordapp.com/attachments/467151789825720340/1071860432136065204/sig-e-e-geck-atl.png)
![[Image: sharnaz.png]](https://media.discordapp.net/attachments/831282703767830599/959659352124448778/sharnaz.png)
![[Image: JKortesi81.gif]](http://sig.grumpybumpers.com/host/JKortesi81.gif)
