File Transition
- emplep7
- Commissioner & CSL Champ 2018
- Posts: 6694
- Joined: Fri Jul 11, 2014 5:06 pm
- Contact:
File Transition
Hey,
So just to provide an update on where we are with the final released file, see some additional notes below:
1) THE PROBLEM: In regards to the FG%s that are being observed in a good chunk of players throughout the league, this is 100% correlated to the Shot Location and Preferences for each player. The bottom line is that DDS19 was a poor game (which we all knew) that GMs figured out was easily manipulated if you just sent everyone's shot location to be primarily in the post even if they had great shooting ratings everywhere else. It resulted in inflated FG%s from DDS3 numbers and some unrealistic game play overall where strong defenders had less of an impact.
In DDS22, that isn't going to work anymore - in fact it has proven to hurt players. This game seems to be built more around the modern era of basketball where you will need to spread your location and preference points around the players strengths in order to get them back to where you expect them to be. This has been proven through our testing by shifting these preferences around for players that have struggled in the preseason.
Spacing between teams isn't that impactful here, but spacing within a players skillset is REALLY important. Now if you have a player that sucks from everywhere, but is a really strong post scorer - send the poor guy to the paint 100% and they will still shoot efficiently. But if you have a 6'3 guard who can score well from mid range and is solid from 3PT range and you send him to the paint primarily he is going to get shut down and shoot low percentages.
This also brings back the value of those players who are offensively versatile which DDS19 completely wiped out their value.
THE SOLUTION: We understand this is going to take time to revert players back to where they should be based on several seasons of catering to the DDS19 engine. So on top of the 20 instructions that we are allotted each season, we are going to allow GMs to select up to 3 players to use a chunk number of instruction points against to set their players how they want ahead of the season start (more details to come soon). These will not count against the 20 that are provided each year, but will allow everyone to at a minimum focus on their top players to try and correct the wrongs from the previous engine.
Additionally, Andre is close to unveiling a program that will be available to all GMs that will provide player potentials (with some deviations similar to scouting) which will include shooting ratings provided as a letter grade (again with some deviation similar to scouting). This can be used as a tool to set preferences around each players strengths.
2. THE PROBLEM: So yeah, about those blocks. We are going to need to lower them a bit on the higher end of the spectrum since the top guys are blocking shots at a much higher clip than we saw during testing. We want blocking to be elevated league-wide in general still, but we dont want players averaging 5 BPG since that is just unrealistic. I will say that this has minmum to negligible impact to the FG% issues we have seen at least, but was an interesting by product regardless.
THE SOLUTION: As stated above, the top 25% of the league is going to be restored back to their previous ratings while the low/mid pack will get a tiered bump less than what they are seeing now. This will still elevate blocks league wide, but remove the chance of DeAndre Ayton getting a triple double with blocks every other game (sorry Adam).
3. THE PROBLEM: Personality ratings were adjusted by the game following the initial roster file load. This was an odd one as well, but is something that is being adjusted back now.
THE SOLUTION: This is the easy one and we are in the process of fixing it now.
Additional Notes
- The 40 COR max is going to be removed. GMs will have the ability to spread their 100 points around the locations however they please. The maxes in place for the preferences are still going to remain in effect for the time being, but we might adjust this starting NEXT season.
- The HTML should be working now with certain players who were leading them to the wrong player. If you find anything else wrong with those, please let us know.
- The last is more of a question - do you all want three more pre-season sims once you have a chance to adjust your player preferences? We are going back and forth on this, but wanted to get your input. Let us know.
So just to provide an update on where we are with the final released file, see some additional notes below:
1) THE PROBLEM: In regards to the FG%s that are being observed in a good chunk of players throughout the league, this is 100% correlated to the Shot Location and Preferences for each player. The bottom line is that DDS19 was a poor game (which we all knew) that GMs figured out was easily manipulated if you just sent everyone's shot location to be primarily in the post even if they had great shooting ratings everywhere else. It resulted in inflated FG%s from DDS3 numbers and some unrealistic game play overall where strong defenders had less of an impact.
In DDS22, that isn't going to work anymore - in fact it has proven to hurt players. This game seems to be built more around the modern era of basketball where you will need to spread your location and preference points around the players strengths in order to get them back to where you expect them to be. This has been proven through our testing by shifting these preferences around for players that have struggled in the preseason.
Spacing between teams isn't that impactful here, but spacing within a players skillset is REALLY important. Now if you have a player that sucks from everywhere, but is a really strong post scorer - send the poor guy to the paint 100% and they will still shoot efficiently. But if you have a 6'3 guard who can score well from mid range and is solid from 3PT range and you send him to the paint primarily he is going to get shut down and shoot low percentages.
This also brings back the value of those players who are offensively versatile which DDS19 completely wiped out their value.
THE SOLUTION: We understand this is going to take time to revert players back to where they should be based on several seasons of catering to the DDS19 engine. So on top of the 20 instructions that we are allotted each season, we are going to allow GMs to select up to 3 players to use a chunk number of instruction points against to set their players how they want ahead of the season start (more details to come soon). These will not count against the 20 that are provided each year, but will allow everyone to at a minimum focus on their top players to try and correct the wrongs from the previous engine.
Additionally, Andre is close to unveiling a program that will be available to all GMs that will provide player potentials (with some deviations similar to scouting) which will include shooting ratings provided as a letter grade (again with some deviation similar to scouting). This can be used as a tool to set preferences around each players strengths.
2. THE PROBLEM: So yeah, about those blocks. We are going to need to lower them a bit on the higher end of the spectrum since the top guys are blocking shots at a much higher clip than we saw during testing. We want blocking to be elevated league-wide in general still, but we dont want players averaging 5 BPG since that is just unrealistic. I will say that this has minmum to negligible impact to the FG% issues we have seen at least, but was an interesting by product regardless.
THE SOLUTION: As stated above, the top 25% of the league is going to be restored back to their previous ratings while the low/mid pack will get a tiered bump less than what they are seeing now. This will still elevate blocks league wide, but remove the chance of DeAndre Ayton getting a triple double with blocks every other game (sorry Adam).
3. THE PROBLEM: Personality ratings were adjusted by the game following the initial roster file load. This was an odd one as well, but is something that is being adjusted back now.
THE SOLUTION: This is the easy one and we are in the process of fixing it now.
Additional Notes
- The 40 COR max is going to be removed. GMs will have the ability to spread their 100 points around the locations however they please. The maxes in place for the preferences are still going to remain in effect for the time being, but we might adjust this starting NEXT season.
- The HTML should be working now with certain players who were leading them to the wrong player. If you find anything else wrong with those, please let us know.
- The last is more of a question - do you all want three more pre-season sims once you have a chance to adjust your player preferences? We are going back and forth on this, but wanted to get your input. Let us know.
- Dylan
- Senior Vice President
- Posts: 1909
- Joined: Fri Jul 11, 2014 12:44 pm
- Contact:
Re: File Transition
Nice, all of that makes sense. I don't know if anyone will be able to abuse the chunk instructions but I hope it levels the playing field a little more.
Regarding more preseason sims, I think one or two more would be more than enough. If we know what the issues are and how to combat them, I don't know why we need to do a ton more preseason.
Regarding more preseason sims, I think one or two more would be more than enough. If we know what the issues are and how to combat them, I don't know why we need to do a ton more preseason.
-
- Chairman of the Board
- Posts: 4661
- Joined: Fri Jul 11, 2014 5:55 am
- Contact:
Re: File Transition
Yes on more preseason sims
- thecoolestjedi
- General Manager
- Posts: 489
- Joined: Sat Jan 01, 2022 1:18 am
- Contact:
Re: File Transition
Definitely yes on more pre season sims
Phoenix Suns GM
Season recap: 2030-31: 36-46
"Fall Seven Times, Get Up Eight"
Season recap: 2030-31: 36-46
"Fall Seven Times, Get Up Eight"
- thecoolestjedi
- General Manager
- Posts: 489
- Joined: Sat Jan 01, 2022 1:18 am
- Contact:
Re: File Transition
Also appreciate the work of the admin team on figuring out the issues and getting some great solutions to them.
Phoenix Suns GM
Season recap: 2030-31: 36-46
"Fall Seven Times, Get Up Eight"
Season recap: 2030-31: 36-46
"Fall Seven Times, Get Up Eight"
- wms02a
- CEO
- Posts: 2299
- Joined: Mon May 23, 2016 5:18 pm
- Location: Arkasnas
- Contact:
Re: File Transition
AWesome Thank yoU!
Yes on more preseason
Yes on more preseason
Dallas Maverick GM
2024: San Antonio Spurs - Southwest Division Champion, GM of the Year, Coach of the Year
2025: San Antonio Spurs - Western Conference Champions
2025: San Antonio Spurs - Western Conference Champions
- Marcos_Beck
- CSL Champ 2019
- Posts: 7758
- Joined: Fri Jul 11, 2014 5:15 am
- Location: Brazil
- Contact:
Re: File Transition
Yes, more preseason please!!
Good work guys
Good work guys
18-19 Chicago Bulls: CSL Champions
#1 D.Rose #11 J.Holiday #21 J.Butler #42 A.Horford #13 J.Noah
29-30 Philadelphia 76ers: CSL Champions
#1 L.Ball #15 M.Beasley #23 B.Bowen #21 M.Wagner #51 K.Towns
#1 D.Rose #11 J.Holiday #21 J.Butler #42 A.Horford #13 J.Noah
29-30 Philadelphia 76ers: CSL Champions
#1 L.Ball #15 M.Beasley #23 B.Bowen #21 M.Wagner #51 K.Towns
- KW
- CSL Champ 2017
- Posts: 10688
- Joined: Fri Jul 11, 2014 5:18 am
- Location: CO
- Contact:
Re: File Transition
Fantastic work guys! Agree on all fronts, and want more preseason sims.
I still believe RA Rate / Dunk Rate has always been too low and believe that to be a bigger issue than the rim protection... the amount of 3-10 foot shots is simply way too high in this league and has been since 19. Was hoping to see that adjusted.
I still believe RA Rate / Dunk Rate has always been too low and believe that to be a bigger issue than the rim protection... the amount of 3-10 foot shots is simply way too high in this league and has been since 19. Was hoping to see that adjusted.
- Marcos_Beck
- CSL Champ 2019
- Posts: 7758
- Joined: Fri Jul 11, 2014 5:15 am
- Location: Brazil
- Contact:
Re: File Transition
With players finally moving away from the post which was really an abuse of how bad DDS19 was, those in-between shots will be lowered a lot.. Three pointers should be very helpful in the new engine, finally. I fully believe knockdown shooters’ value will be increased now.KW wrote: ↑Sat Jul 30, 2022 3:55 am Fantastic work guys! Agree on all fronts, and want more preseason sims.
I still believe RA Rate / Dunk Rate has always been too low and believe that to be a bigger issue than the rim protection... the amount of 3-10 foot shots is simply way too high in this league and has been since 19. Was hoping to see that adjusted.
Defense matters again, finally, as perimeter players will have to initiate offense away from the post, there will be way more drives faced (and stopped), and less free post ups.
18-19 Chicago Bulls: CSL Champions
#1 D.Rose #11 J.Holiday #21 J.Butler #42 A.Horford #13 J.Noah
29-30 Philadelphia 76ers: CSL Champions
#1 L.Ball #15 M.Beasley #23 B.Bowen #21 M.Wagner #51 K.Towns
#1 D.Rose #11 J.Holiday #21 J.Butler #42 A.Horford #13 J.Noah
29-30 Philadelphia 76ers: CSL Champions
#1 L.Ball #15 M.Beasley #23 B.Bowen #21 M.Wagner #51 K.Towns
- Dennis
- CSL Champ 2028
- Posts: 10360
- Joined: Fri Jul 11, 2014 10:19 am
- Location: Düsseldorf
- Contact:
Re: File Transition
That sounds good that it is “only” the prefs of players and the comvination of floor position instead of completely new usage of the shoting ratings.
Also cool to see we are allowed to rework players. Fast and good decisions by the armin team. Seems some offline testing very valuable.
Given we have guys without the game to run tests, is it possible you post 2-3 examples of how you remodelled some players impacted during preseason and what impact it had on the players FG%? Like doumbia, taylor, fox, etc come to mind
Also cool to see we are allowed to rework players. Fast and good decisions by the armin team. Seems some offline testing very valuable.
Given we have guys without the game to run tests, is it possible you post 2-3 examples of how you remodelled some players impacted during preseason and what impact it had on the players FG%? Like doumbia, taylor, fox, etc come to mind
Boston Celtics Hall of Fame
PG G. Dragic [7] - Jersey Retired <> 2013 - 2015
PG R. Westbrook [0] - Jersey Retired <> 2016 - 2019
Fox - Johnson - Bridges - Hlinason - Whiteside
CSL Champions 2028
PG G. Dragic [7] - Jersey Retired <> 2013 - 2015
PG R. Westbrook [0] - Jersey Retired <> 2016 - 2019
Fox - Johnson - Bridges - Hlinason - Whiteside
CSL Champions 2028
- OneNole
- Chairman of the Board
- Posts: 4576
- Joined: Fri Feb 27, 2015 7:36 pm
- Contact:
Re: File Transition
I need to go back and look at old scouts and see what wings are good inside and how they are performing.
Andre has certainly won the switch lotto. I said it during that draft that Culver was the beat player of that draft and now with the switch he went from all star to superstar. Thankfully he didnt get boosted like his other high picks in doum and bruno
Andre has certainly won the switch lotto. I said it during that draft that Culver was the beat player of that draft and now with the switch he went from all star to superstar. Thankfully he didnt get boosted like his other high picks in doum and bruno
FINALLY a logical HC Rule our work is done
- Dennis
- CSL Champ 2028
- Posts: 10360
- Joined: Fri Jul 11, 2014 10:19 am
- Location: Düsseldorf
- Contact:
Re: File Transition
Btw, that also kills the Harden at PF thing without having a kyrus rule ..
Boston Celtics Hall of Fame
PG G. Dragic [7] - Jersey Retired <> 2013 - 2015
PG R. Westbrook [0] - Jersey Retired <> 2016 - 2019
Fox - Johnson - Bridges - Hlinason - Whiteside
CSL Champions 2028
PG G. Dragic [7] - Jersey Retired <> 2013 - 2015
PG R. Westbrook [0] - Jersey Retired <> 2016 - 2019
Fox - Johnson - Bridges - Hlinason - Whiteside
CSL Champions 2028
- mgtr81
- Chairman of the Board
- Posts: 3623
- Joined: Wed Jul 06, 2016 6:14 am
- Contact:
Re: File Transition
Thanks for the info and for the work.
I would be in favour of running more pre-season tests after the new ratings / preferences are updated and before choosing the three players to tweak.
Edit: Regarding guards in the post, those with size should not be penalized, I think. I'm talking 6'6" and above.
I would be in favour of running more pre-season tests after the new ratings / preferences are updated and before choosing the three players to tweak.
Edit: Regarding guards in the post, those with size should not be penalized, I think. I'm talking 6'6" and above.
- bt
- CSL Champ 2020, 2023 & 2027
- Posts: 5815
- Joined: Fri Jul 11, 2014 6:12 am
- Contact:
Re: File Transition
What Kyrus said is 100% the issue and it relates to blocking also obviously.
Now he would know if the RA rates are too high as the draft guys set a ton of them but if DDS19 was too strong in the post with what seems like poor inside ratings, then it really needs to be fixed, especially for big men because if you compare the default DDS19 and DDS22 roster files, the RA rate is pretty high in both games yet has tanked in DDS22.
Have any admins compared the inside rates we have to the default files? I assume this should have been picked up in testing though.
Now he would know if the RA rates are too high as the draft guys set a ton of them but if DDS19 was too strong in the post with what seems like poor inside ratings, then it really needs to be fixed, especially for big men because if you compare the default DDS19 and DDS22 roster files, the RA rate is pretty high in both games yet has tanked in DDS22.
Have any admins compared the inside rates we have to the default files? I assume this should have been picked up in testing though.
sacramento kings
tremont waters | skylar mays | nigel johnson
matisse thybulle | justin james
robert franks | darius bazely | john butler
donta hall | xavier tillman | patrick williams
jon collins | jonathan isaac | norvel pelle | mfiondu kabengele
CSL CHAMPION 2020, 2023, 2027
tremont waters | skylar mays | nigel johnson
matisse thybulle | justin james
robert franks | darius bazely | john butler
donta hall | xavier tillman | patrick williams
jon collins | jonathan isaac | norvel pelle | mfiondu kabengele
CSL CHAMPION 2020, 2023, 2027
- Flaming Homer
- Commissioner
- Posts: 5762
- Joined: Tue Feb 20, 2018 6:31 pm
- Contact:
Re: File Transition
At the very beginning we compared the FGRA and DUNK ratings of the the CSL to the default file.
They are indeed quite a bit lower. We ran test's with increased FGRA thru out the league but it didn't show the results we wanted to see back then.
They are indeed quite a bit lower. We ran test's with increased FGRA thru out the league but it didn't show the results we wanted to see back then.
- Dennis
- CSL Champ 2028
- Posts: 10360
- Joined: Fri Jul 11, 2014 10:19 am
- Location: Düsseldorf
- Contact:
Re: File Transition
Porter interesting case though. Anyone an idea why he’s that low in preseason? I mean he’s basically a big man with a SF Skillset and a fairly good inside game
Boston Celtics Hall of Fame
PG G. Dragic [7] - Jersey Retired <> 2013 - 2015
PG R. Westbrook [0] - Jersey Retired <> 2016 - 2019
Fox - Johnson - Bridges - Hlinason - Whiteside
CSL Champions 2028
PG G. Dragic [7] - Jersey Retired <> 2013 - 2015
PG R. Westbrook [0] - Jersey Retired <> 2016 - 2019
Fox - Johnson - Bridges - Hlinason - Whiteside
CSL Champions 2028
-
- General Manager
- Posts: 320
- Joined: Thu Jun 16, 2022 9:04 pm
- Contact:
Re: File Transition
Great work on all points.
I'm ambivalent on more preseason.
I'm ambivalent on more preseason.
- KW
- CSL Champ 2017
- Posts: 10688
- Joined: Fri Jul 11, 2014 5:18 am
- Location: CO
- Contact:
Re: File Transition
When testing 19, we'd seen everyone shooting like 70-90% from the restricted area, so admins and draft team alike dropped all interior ratings and tendencies. An unintended consequence was too few 0-3, too much 3-10, but as the draft team my job is to keep ratings level with what we've got in the league rather than blazing my own trail and making years of prospects that are much higher or lower at a certain skill based on how I feel they should be. This way it's easy for a global change should it be necessary.bt wrote: ↑Sat Jul 30, 2022 9:47 am What Kyrus said is 100% the issue and it relates to blocking also obviously.
Now he would know if the RA rates are too high as the draft guys set a ton of them but if DDS19 was too strong in the post with what seems like poor inside ratings, then it really needs to be fixed, especially for big men because if you compare the default DDS19 and DDS22 roster files, the RA rate is pretty high in both games yet has tanked in DDS22.
Have any admins compared the inside rates we have to the default files? I assume this should have been picked up in testing though.
- KW
- CSL Champ 2017
- Posts: 10688
- Joined: Fri Jul 11, 2014 5:18 am
- Location: CO
- Contact:
Re: File Transition
Check where he's shooting from. Large amount of 3-10 being chucked up at garbage efficiency last tim I checked
- emplep7
- Commissioner & CSL Champ 2018
- Posts: 6694
- Joined: Fri Jul 11, 2014 5:06 pm
- Contact:
Re: File Transition
Yeah, the CSL Dunk Rate is actually higher than that of the Default files. The RA Rate is a bit lower which is where one of the potential changes is coming into play through testing.
Who is online
Users browsing this forum: No registered users and 7 guests