Stats comparison

[mgtr81]

Here are some stats comparison between last season in DDS3 and the three first pre-season sims. While the sample size may be small for individual players, it is already representative for league stats. The following are the four key stats that best represent ratings outcome (for example, rebounds will be a direct consequence of FG % so it's not a primary stat). The first value is DDS3, the second value is DDS19.

Steals per game per team: 6.90 vs 9.61
Blocks per game per team: 5.73 vs 8.02
Turnovers per game per team: 13.54 vs 17.01
Field goal percentage: 45.35 vs 44.93

Numbers speak for themselves. Shooting ratings are ok, but blocks and steals need to be lowered.

Update after preseason is over:

Steals per game per team: 6.90 vs 9.43
Blocks per game per team: 5.73 vs 8.16
Turnovers per game per team: 13.54 vs 16.23
Field goal percentage: 45.35 vs 44.98

[Dennis]

FG% we dropped from #1 to #25. It’s no good !

[mgtr81]

FG% we dropped from #1 to #25. It’s no good !

You alongside New Orleans and Philadelphia are the teams who have suffered a biggest drop on FG %. The increase in blocks also mean lower FG % for the opponent. You may have faced opponents with big increase in blocks. The teams with a larger increase in blocks are Lakers, Magic and Raptors. You have played all three so far.

[TheTwoWallaces]

Word yea those steal and block figures don’t look like flukes

[Myles]

In every test that Jon and I did, we looked at everything, especially looking at team stats. We looked at team scoring, team shooting %, team pace, team rebounding, team steals, team blocks, team turnovers, team points per possession, and all that stuff for what they are allowing the other team to do as well. We didn’t release the file to you all until we had run multiple tests that came back with every single result within the normal ranges of what we saw in the CSL. If I remember correctly, most teams were between 11 and 17 turnovers per game. Like the best team in the league was usually between 10.5 and 11.5 turnovers, and the worst team was between 17 and 19 turnovers per game.

I think part of it is we ran pretty vanilla rotations and schemes for our test. No one played more than 28 mpg, I think we had every team play balanced on both sides of the ball. Now that everyone is throwing everything at the wall, in terms of rotations, offensive strats, and defensive strats, it’s changing how the game works.

The other part of it is a lot of these things go hand in hand. Steals are up, which means turnovers are up. Now the question is, is the issue a real thing that will need to be adjusted, or is it a two-week sample size and if we adjust it then we are going to run into the opposite issue soon? Where steals and turnovers are too low, because we overreacted to what essentially amounted to a fluke?

This is just SO far outside the bounds of what we found when we tested, where we very thoroughly tested and made sure that everything fit in the ranges of what we saw in the CSL in DDS3. So there's clearly more testing to be done, and maybe we need to run tests utilizing the depth charts/strategies you've sent in, because I don't want to make adjustments based on just a two-week sample size.

[mgtr81]

In every test that Jon and I did, we looked at everything, especially looking at team stats. We looked at team scoring, team shooting %, team pace, team rebounding, team steals, team blocks, team turnovers, team points per possession, and all that stuff for what they are allowing the other team to do as well. We didn’t release the file to you all until we had run multiple tests that came back with every single result within the normal ranges of what we saw in the CSL. If I remember correctly, most teams were between 11 and 17 turnovers per game. Like the best team in the league was usually between 10.5 and 11.5 turnovers, and the worst team was between 17 and 19 turnovers per game.

I think part of it is we ran pretty vanilla rotations and schemes for our test. No one played more than 28 mpg, I think we had every team play balanced on both sides of the ball. Now that everyone is throwing everything at the wall, in terms of rotations, offensive strats, and defensive strats, it’s changing how the game works.

The other part of it is a lot of these things go hand in hand. Steals are up, which means turnovers are up. Now the question is, is the issue a real thing that will need to be adjusted, or is it a two-week sample size and if we adjust it then we are going to run into the opposite issue soon? Where steals and turnovers are too low, because we overreacted to what essentially amounted to a fluke?

This is just SO far outside the bounds of what we found when we tested, where we very thoroughly tested and made sure that everything fit in the ranges of what we saw in the CSL in DDS3. So there's clearly more testing to be done, and maybe we need to run tests utilizing the depth charts/strategies you've sent in, because I don't want to make adjustments based on just a two-week sample size.

I didn't mean that no testing was done, I was just posting some info that I think can be worth looking at. I know you both did a great work, but I think the answer is in your post. If you were running the standard sets and default rotations, then that's it. Also, despite being just a two-week sample, the amount of total games is already of some significance, I think we're well above 200 games.

[Stockton12]

I think more sims are needed to see what happens with the numbers

[mgtr81]

Update after preseason is over:

Steals per game per team: 6.90 vs 9.43
Blocks per game per team: 5.73 vs 8.16
Turnovers per game per team: 13.54 vs 16.23
Field goal percentage: 45.35 vs 44.98

The landscape has not changed much since the first three sims. It makes sense, as 200 games was already a representative sample for global stats. Steals, blocks and field goal percentage is almost the same that after the first three sims, with turnovers having decreased 0.78 per game.

Edit: four examples from my team:

Ja Morant - 84 handles, 58 passing, 0.69 PAS/HAN ratio, 11 % drive/pass, 9 % drive shot, 72 % pass -2.5 turnovers per 36 minutes
Joe Jackson - 83 handles, 56 passing, 0.67 PAS/HAN ratio, 3 % drive/pass, 4 % drive shot, 80 % pass - 1.6 turnovers per 36 minutes
Troy Brown - 76 handles, 58 passing, 0.76 PAS/HAN ratio, 7 % drive/pass, 10 % drive shot, 73 % pass - 3.3 turnovers per 36 minutes
Frank Mason -74 handles, 54 passing, 0.73 PAS/HAN ratio, 7 % drive/pass, 10 % drive shot, 73 % pass - 2.5 turnovers per 36 minutes

I don't see why such a big difference between Brown and Mason as they are almost identical in the relevant areas for turnovers. In fact, all four of them are well below a PAS/HAN ratio of 1, which should translate into very few turnovers (at least it was like that in previous versions of the game).

I hope these stats can help fine tune and improve the already excellent work done by the admin team (I know how difficult it is to transfer from one engine to the other).

[bt]

I'm just guessing here but it could be that pace, press and defensive intensity are different in this engine. In addition, 20 points spread versus what here? Do we know what teams were using?

I'd say to truly try compare stats, that would need to be taken into account first.