New file
Oct 12, 2018 10:30:03 GMT
Post by mgtr81 on Oct 12, 2018 10:30:03 GMT
Finally ... you can download the new file for DDSPB2018. The links to both the file and the html are here:
nah.boards.net/thread/1214/league-file-link
The most important thing, which is players and their ratings, is already done and tested. I still need to add the correct team logos, edit coaches, place the correct draft picks for each team and import all historic data. But at least you can now download the file and take a look at the new ratings and how everything looks. Some important comments:
1. You will see the scoring rating has varied and in some cases greatly. The new engine is much more complex in terms of shot distribution, coupling the scoring rating with the playing style and something else that I'm still missing, that's why I needed so much time to get everything as close as possible to our stats. For some players it has increased, for others has decreased. Still, the outcome should be really close to last season.
2. The blocking rating has gone down quite a bit league wide to adjust to our previous stats.
3. The steal rating has gone down a bit, for the same reason.
4. Rebounding is a very complex story. The engines are also different in that regard. DDS1 put a lot of weight in positional rebounding, that is, a center will get more rebounds than a power forward with the very same rebounding ratings, a power forward more than a small forward and so on. I could have adjusted rebounding ratings to match last season stats, but since players can play multiple positions and not necessarily the one their listed, I decided to leave them as they were.
5. I did my best with the little time I had available, despite the long wait (really busy in real life). So inevitably there will be some players that will perform slightly different to what they did. We will all have to assume that. We will also learn together about this new engine as the season progress.
6. For the time being, there will be no player instructions. The shot distribution seems to be very sensitive to the playing style, so I need to do a bit more testing to make sure which options of the playing style can be changed and which of them should be left as they are.
To finalize this post, here are some statistical comparisons between last season (DDS1) and the new engine (DDSPB18) with current rosters. The first number corresponds to last season and the number between brackets to the current test:
Scoring (team points per game):
League average: 101.3 (104.0)
Maximum: 107.8 (112.0)
Minimum: 92.6 (96.7)
Turnovers (team turnovers per game):
League average: 15.0 (14.7)
Maximum: 17.2 (17.6)
Minimum: 11.4 (12.9)
Shooting (team field goal percentage):
League average: 47.5 (47.2)
Maximum: 52.3 (49.8)
Minimum: 43.9 (44.6)
I wouldn't be too concerned with the FG %'s as the tests were performed with default lineups, where the AI gave more playing time to players with high scoring despite being inefficient (that could also explain the slight increase in scoring).
Please, log in to your teams and post here any issues you have.
nah.boards.net/thread/1214/league-file-link
The most important thing, which is players and their ratings, is already done and tested. I still need to add the correct team logos, edit coaches, place the correct draft picks for each team and import all historic data. But at least you can now download the file and take a look at the new ratings and how everything looks. Some important comments:
1. You will see the scoring rating has varied and in some cases greatly. The new engine is much more complex in terms of shot distribution, coupling the scoring rating with the playing style and something else that I'm still missing, that's why I needed so much time to get everything as close as possible to our stats. For some players it has increased, for others has decreased. Still, the outcome should be really close to last season.
2. The blocking rating has gone down quite a bit league wide to adjust to our previous stats.
3. The steal rating has gone down a bit, for the same reason.
4. Rebounding is a very complex story. The engines are also different in that regard. DDS1 put a lot of weight in positional rebounding, that is, a center will get more rebounds than a power forward with the very same rebounding ratings, a power forward more than a small forward and so on. I could have adjusted rebounding ratings to match last season stats, but since players can play multiple positions and not necessarily the one their listed, I decided to leave them as they were.
5. I did my best with the little time I had available, despite the long wait (really busy in real life). So inevitably there will be some players that will perform slightly different to what they did. We will all have to assume that. We will also learn together about this new engine as the season progress.
6. For the time being, there will be no player instructions. The shot distribution seems to be very sensitive to the playing style, so I need to do a bit more testing to make sure which options of the playing style can be changed and which of them should be left as they are.
To finalize this post, here are some statistical comparisons between last season (DDS1) and the new engine (DDSPB18) with current rosters. The first number corresponds to last season and the number between brackets to the current test:
Scoring (team points per game):
League average: 101.3 (104.0)
Maximum: 107.8 (112.0)
Minimum: 92.6 (96.7)
Turnovers (team turnovers per game):
League average: 15.0 (14.7)
Maximum: 17.2 (17.6)
Minimum: 11.4 (12.9)
Shooting (team field goal percentage):
League average: 47.5 (47.2)
Maximum: 52.3 (49.8)
Minimum: 43.9 (44.6)
I wouldn't be too concerned with the FG %'s as the tests were performed with default lineups, where the AI gave more playing time to players with high scoring despite being inefficient (that could also explain the slight increase in scoring).
Please, log in to your teams and post here any issues you have.