This is a sim-mechanics discussion, NOT a post to try and "bring down the establishment" or anything like that. The sim seems whack to many of us - let's figure it out.



Ever notice how the players you draft - especially the ones from the 50s, 60s and early 70s, will not get their RL FTAs but will get more turnovers?

Why is that?

Pace adjustment killed a lot of what the older players were good at, but at the same time, did not fix what they weren't good at. We were given fg%+, but it wasn't implemented correctly. All you have to do is take a look at someone from the 60s... look at their fg%+, then look at their 2pt%#. It helps if you pick someone that doesn't shoot threes (take a Russell or Wilt).

The pace adjustment also does not seem to effect personal fouls. It still seems as though these players are fouling at a rate of per 48 minutes.

So the normalization did the following things: Made older players rebound less, shoot just as bad, foul just as much, turn the ball over just as much if not more, go to the line less, shoot less, score less.

Yet their salaries still seem extremely high compared to what they bring to the table.

Why pay $11-$13 million for Wilt when you can get Moses for $9 million who'll rebound better, shoot better (depending on the wilt, but i'm comparing wilts with 22.1%+ usage), and get to the line just as much.

Why pay $9 million for Bill Russell when you can get more boards, less turnovers, and much more efficient scoring (They'll have about the same amount of shots over a season) with Rodman for a couple hundred thousand less?

9-mill for Baylor vs 9-mill for Lebron? Baylor will grab a few more offensive boards, but Lebron will shoot better, get to the line more, foul less, turn the ball over less and will have a greater positive effect on his teams fg% than Baylor for roughly the same cost.

Three examples off the top of my head. All I can see is that you're paying for totals that you'll never achieve because of the advanced percentages. I think you're paying for a stat that isn't working (fg%+). The only benefit at all for these older guys is the minutes they bring to the table, but you can almost always find a guy who plays the same minutes that does everything better for cheaper. The big difference you'll find is the FTA, but those FTA don't matter because the older players will not get even remotely close to their RL fta.


This is a conversation that went down in a league forum, so read from bottom up:

02/05 15:56 ashamael should be a topic open for discussion as it related to the sim. Unfortunately, I can't post it in its proper place.

02/05 15:37 macphisto sitemail? :)

02/05 15:10 ashamael sorry for spamming this up. Can't exactly hold a forum conversation on the forums with monkee anymore.

02/05 15:08 ashamael Which brings me back to an issue with the sim: The normalization of eras is not reflected in the salary. We're paying salaries for extra stats that our players will never reach. I think we're also paying for guys' fg+, which is not reflected correctly when it gets broken down to 2pt%# and 3pt%# and sure isn't implemented.

02/05 15:05 ashamael This seems pretty consistent with what I see from players from that era, including Wilt, who almost always has fewer turnovers for me, but never gets close to his total possessions even when drafting 1000 to 1500 less possessions than my team ultimately ends up with.

02/05 15:04 ashamael so, I had fewer possessions (not an overdraft issue, either, my team finished with way more possessions than drafted) with the Big O. He turned the ball over about a quarter of a percent less, went to the line 5% less, but shot from the field 5% more.

02/05 15:02 ashamael fga: 1626 vs 1724 fta: 465 vs 813 tov: 318 vs 369. total possessions:2148.6 vs 2450.72. relative percentages: fga% 75.67 vs 70.35; fta%: 9.52% vs 14.6%; tov%: 14.8% vs 15.06%

02/05 14:56 ashamael Yeah. I just finished an OL with 64-65 Oscar. He finished below RL stats in points, fga, fta, tov and reb (both offensive and total). Rebounds are obvious: the normalization and the fact that few rebounders match their RL totals on good sim teams (because you draft so many boards)... Going to do some calculations.

02/05 07:52 monkee if they use TOV it amounts to the same thing - worse maybe - it doesnt take into account the impact of sharing the load and leaves your high usage guys turning the ball over at thier IRL rate while generating all the positive touch stats at a lesser blended rate - that would definitely explain the phenomenon of Wilt does everything a little less than at IRL pace except turn the ball over

02/05 00:55 ashamael monkee, they "said" they use tov% to determine turnovers even though they don't give us that stat to view, search or sort with. Of course, I think you know this sim better than (edited)...

02/04 20:43 monkee in other words Wilt may not get his boards or attempts because those are decided by a distro mechanism that takes aggregate usage into account and distributes the shots and boards accordingly but the turnovers are produced in a per minute non-blended manner (i.e. every possession/round each player is likely to turn it over on his normative IRL per minute TO standard and not on a mitigated 'this guy isnt getting as many touches as he did IRL' standard) my theory

02/04 20:40 monkee btw guys - I have a theory that the problem with sim turnovers is this - they are a per minute function of each player and do not aggregate in the manner that all the other positive 'touch' stats do - so even though everybody on a high usage team tends down in a blending on rpm, apm ppm etc - there is no aggregate blending mechanism and therefore high usage teams equal high turnovers

--------------------------------------------------------------------

I'm not sure what's right here, but I think the sim uses fta% and tov% to determine free throw attempt, turnover or just a field goal attempt. These are numbers we can't see, numbers that we have to calculate ourselves. These are numbers that we can't even look up on a site like basketball reference, because the turnover numbers prior to mid-70s are fabri-created by wis. I feel that numbers that are this vital to the sim should be not only viewable, but searchable and sortable. But that's not the point of this thread right now.

The point is to open discussion with those who play the sim - what experience have you had with sim stats versus RL stats in regards to these possession numbers: fga, fta, tov.

Have you found the fg%# to have any impact, or do you find the increases in fg% come from your typical high assist/up-tempo set-up?

For example, I had 63-64 Bill Russell shoot 47% (43% IRL), but I also had a team with almost 70% ast% on the floor for most of the game that ran up-tempo (shaq shoots 61%, when IRL he shot 57%). Russell's increase is consistent with Shaq's.

Hey let's talk about Shaq real quick:

real numbers on top, sim numbers on bottom

fga: 1665 fta: 824 tov: 223 poss: 2250.56 (ft 362.56)
fga: 1822 fta: 578 tov: 286 poss: 2362.32 (ft 254.32)

breakdown of possessions using advanced stats:

fga%: 73.98% fta%: 16.11% tov%: 9.91%
fga%: 77.13% fta%: 10.77% tov%: 12.11%

Again, if you look at the oscar number I posted in the league forum chat, and compare it to the Shaq numbers... Shaq had 112 more possessions in the sim than IRL. He turned the ball over in 2.2% more in those possessions but went to the line 5.34% less in those possessions.

If real Shaq were consistent with sim-shaq's possessions, his numbers should have been closer to this:

1748 fga
865 fta
234 tov

2% and 5% seem like pretty small variance numbers... but it's the fact that the variance is almost always in the same direction that raises questions.

I think that the questions surrounding this also answer questions about the increased number of assists in this version of the sim:

Because guys are going to the line less, they take more field goals. Because of the increased field goal attempts, and the increased effect on fg% due to up-tempo and ast%, there are more field goals made (duh). Since ast% works in two ways (not only increases fg% but decides the likelihood a shot was assisted, and if so, who got the assist), more buckets leads to more assists.

More attempts, far more successful attempts, far more assists. The more assists drafted, the greater the effect, which is why you see a high efg% like one that airstilt built shoot a normal fg% of 57%!!


I think there's another factor at play here, and it also relates to ast%. It's how wis uses ast%. If you've ever noticed, ast% is different at WIS than other sources. That's because WIS doesn't take out a player's own made field goals from the formula. So, in essences, a high ast% is positively impacting his own fg%. Their reasons for doing this is so that a guy like Chris Paul isn't impacted negatively because the other four guys on the floor's ast% doesn't get past the penalty mark (which we don't know where that is). That makes sense in one respect, but in another, it doesn't.

Regardless of whether it's the way to do it or not, it is the way it's done, and I think it also is responsible.

foul problems and cluster foul problems I think also lead to the issue. Less ftas, less personal fouls. Sort of. There were foul problems with the early versions of this release - and they lasted well up into last year. There used to be a line of code involved that when a player was in foul trouble, his defensive presence was less. They took that out for this sim version, and when they did, fouls went crazy. You had guys fouling out in 10-20 minutes played - guys who averaged 2.5 fouls per 48 minutes!

They adjusted that, to my knowledge, at least twice. It doesn't seem to be an issue anymore, but I think it's partly to blame for the issue I brought up originally in this post. There's a lack of free throw attempts because there's a lack of fouls. Sometimes, anyway. The way owners build their teams, they draft less personal fouls than what typically happens.

In the league with my oscar, russell and shaq example, there was a range of 1200 to 1700 personal fouls per team.

In 2008-2009, there was a range of 1500 to 2000 personal fouls per team.

In 1978-1979, there was a range of 1800 to 2400 personal fouls per team.

That comes out to be quite a difference when you start thinking of the number of teams involved.

I'm NOT saying there ISN'T a problem with foul logic - I shake my head often at box scores - but when you look at it like this, it starts to make sense as to why there are less ftas, which means more fgas, which means more turnovers because of the individual overpossession-penalty.

So, what does this mean?

for reference

I think over the past 30 minutes, through the research, calculating and re-reading the way the sim is supposed to work, I've reached some conclusions. I'm still thinking, though, and will come back later and do some more work. But I definitely feel better about some of the whacky crap I've seen in the sim.

I still really think we need to be able to view a player's tov%, his fouled percentage (what I've been referring to as fta%) and his field goal shooting rate (fga%). Not only view them, but search for them in the draft center, and sort our results ascending or descending with them. Maybe after this discussion I'll submit a ticket for consideration of inclusion in a future release...

What are your experiences with all of this? Do you find your teams and players consistent with the results I'm coming up with? If they're different, how so?

(one more thought real quick)

my 67-68 Wilt ended up with a tov% of about 1% more than RL wilt alleged tov%. His fta% was roughly 10% while IRL it was more than 19%. Note this wilt's usage is 20.5% There is definitely a correlation between lower usage and higher usage when it comes to these percentages.

What about the other teams foul rate? Did you apply those numbers in? IF the other teams are low foul guys which is very possible when drafting SIM teams then RL FTA shouldnt be reached correct? The other team would then have a gripe that their players are fouling higher then RL. IF this was touched on I apologize

Even the sim teams who have "high foul rates" are, in general, lower than modern day and much, much, much lower than previous eras.