Feedback from long running game

[pdandy]

These are observations from speed running a league up to year 2062+. Gary, you have the same file up to 2050 or so.

Good:
- Stability: I've hit only one crash CAllStar::Intialize(). I tried to sim up to a day before and lost the repro by doing so. Anyways, the fact that I'm up to year 2062 and have only hit 1 crash is AMAZING. Stress test passed with flying colors.
- Game still runs smoothly despite 38+ years of data.
- Apron constraints is a lot of fun. I thought this would be hard for the AI to plan around. I thought the AI would end up renouncing players too aggressively and there would be a lot of player movement as a consequence but AI is doing well. Extremely difficult to achieve/maintain a dynasty while keeping your job at the same time.
- Love the breaking news popup. Love the All Star weekend updates.
- The player personality and interests system makes the game very challenging. Makes the team feel more like a team rather than a bunch of numbers. I am excited to see where Gary takes this system to in the future.
- Definitely not an exhaustive list of all the great things in this game. I'm generally having a blast on this watcher run and my main Waves playthrough.

Would like to see improvement:
Play signings:
- As I reported in the FirstAccess thread, there are players in real life who fumble the bag but in the game I think it is too prevalent and magnified. There are young high rated 80-90+ players who end up waiting until the final 4 or so days of free agency and end up signing long term MLE or Bi-annual exceptions with no player option. I expect such high quality players to sign shorter term exception contracts so they can try again sooner. This ability to sign great players for long term at such a discount rate gives teams an unfair advantage and is an exploit for the human player. You can reproduce this by observing the last few days of free agency. In my most recent season I see
82 rated player, 29 years old, full MLE for 4 years with player option
83 rated player, 25 years old, full MLE for 4 years with player option
85 rated player, 30 years old, full MLE for 4 years with player option
83 rated player, 25 years old, full MLE for 4 years with player option
Earlier in free agency, these types of players were getting 22M+ per year deals. I've also seen more egregious examples with the bi-annual exception.
- Extensions should be a bit more common just like today's league especially for the topmost players and star rookies. In the NBA, the elite players rarely hit free agency. Right now it seems it's mostly the players around high 70s - low 80s rating signing extensions. I know this is a slippery slope as I do enjoy how the game creates a more exciting offseason by letting so many top players enter free agency. Secondly I would love to see the transaction page show for how long and how much the extension was. Right now it just says a player was extended.


Player stats:
I'm only reporting the ones I FEEL stood out like sore thumbs. There could still be other things that bother others more.
- Throughout this long run, I've seen Centers who score 40+ ppg. I reported this issue already. I think it is related to the buff C/PFs got in FirstAccess. I think if you look at the FG%, big men got buffed too much. I would tune that back down a tick. I think the C/PF Scoring rating logic should be tuned down a tick as well since it seems like they are taking a ton of shots. Also big men are getting more FTs in the game than in real life. Currently I see Trevor Hadfield (C) with 43ppg with only 30 mpg, before him it was Emilio Escobar (C) with 45+ ppg, before him it was Luciano Giannini (C) with 44-46ppg. In current year, #2 and #3 scorers are another two centers with 35+ppg. These guys score up to 50ppg in the playoffs. I think once the default roster was retired from the game, it has always been Centers who have been scoring champ. And very timely, the rookie of the year for this season is a center who plays 24.9mpg and scores 27 ppg; he got All-PBL 3rd team honors.
- Steals: in the game there are 26 players with over 2 steals per game. In the 2023-2024 season IRL, there was only 2 players. In 2022-2023 season IRL, there were no players with 2+ spg. This will address the high-ish turnovers as well.
- Triple double total is much lower than IRL. Game has ~6-8 per year. NBA 2023-2024 season ~125 total. But even going back 15 years NBA with the slower game, the triple double count was ~30.

Misc.:
- In the Milestones page, Matthias Tellenbach is listed as "rookie rookie"
- Would love to see sign and trade orchestrated by AI. Huge part of the NBA offseason.

[Gary Gorski]

So the new update addresses alot of stuff mentioned here. I think the MLE usage is better now and definitely a change in the bigs in the sim engine so hopefully the latest update brings these items more in line

[ashantewarrier]

Thanks Gary!

[Scoman]

pdandy,have you done any long term testing with the newest version?

[binballin]

I just did a quick 11-year sim and only one guard averaged more than 25 ppg(https://imgur.com/a/EJvipQ7). Compared to the NBA last year 12 guards surpassed 25ppg.

Luka Dončić (Dallas Mavericks) - 33.9 PPG
Shai Gilgeous-Alexander (Oklahoma City Thunder) - 30.1 PPG
Jalen Brunson (New York Knicks) - 28.7 PPG
Devin Booker (Phoenix Suns) - 27.1 PPG
Donovan Mitchell (Cleveland Cavaliers) - 26.6
De'Aaron Fox (Sacramento Kings) - 26.6 PPG
Stephen Curry (Golden State Warriors) - 26.4 PPG
Tyrese Maxey (Philadelphia 76ers) - 25.9 PPG
Anthony Edwards (Minnesota Timberwolves) - 25.9 PPG
Kyrie Irving (Dallas Mavericks) - 25.6 PPG
Trae Young (Atlanta Hawks) - 25.5 PPG

[Gary Gorski]

If you would like to submit your game files through our ticket system I would like to examine the files to look at the guards and see if they perhaps need a slight development boost

[binballin]

I have submitted the files.

[John1708]

Triple doubles need a boost too

[pdandy]

Did a new speed run using v*.1 patch.

I think the big men nerf is working well. In year 2059 top 10 of scoring is:
1) SG
2) SF
3) PF
4) C
5) SF
6) C
7) SF
8) PF
9) PF
10) PF
11) PG
12) SG
13) PF
14) PF
15) SF

Good distribution compared to First Access where it was C/PF dominant. Despite #1 being an SG, I do buy into the assertion that guards (default db and generated players) can use a small buff.

I wonder if this relates to the high-ish turnovers issue too. If guards keep hold of the ball 1 or 2 more possessions per game, it might bump them up in scoring a few ticks needed.

[pdandy]

Forgot to mention that player ratings do hold up in this game very well. Say what you want about this or that stat should be higher/lower, the general numbers are consistent year over year. No crazy growth or outliers anymore.