Kevin Hollis
Former Thunder GM for 7 years
All Star
Posts: 2,838 
|
Post by Kevin Hollis on Jan 28, 2017 17:55:55 GMT
I guess at the end of the day, I haven't heard the argument of x only gets such stats in the sim so y can't get such and such of a rating before now. I am sure you know ten time more than me when it comes to the rating system here, but it's just my point of view. I think that is because we are trying to say that the sim is fucked up. If you think about it, the only time the sim league player should be averaging the exact same as he is in real life in the sim is if: 1. First he has to be rated completely accurately on that given day to what he is like on that same game in real life. 2. The team he is on needs to be the same in both RL and in game, and they too need to be rated accurately. 3. The team has to disperse their minutes in the same fashion as both RL and in-game 4. Their opponent must also follow rules 2 and 3. 5. The player needs to have the same stat opportunities in the game as in RL, i.e. both teams need to play exactly the same way, in both style, scheme, minutes, everything. Everything needs to be the exact same in both the sim and RL in order for a player to produce identical numbers in the sim that aligns with the real life game. Then, and only then, can a fair assessment be made as to the player's rating and attribute distribution. Now those 5 things can all be correct, and the sim could still fuck it all up. That being said, think about how many of those things are different in the sim than in real life. I think the Warriors are the only team that have the same two teammates that are rated above an 82 as in game and real life. And I don't just mean the Warriors, I mean, say if the Rockets had KAT and Wiggins, then it'd be the same. But the other 10 guys likely aren't. So you have all these different factors that we just plain and simple can't control, which is why a player's sim stats will never be identical for a game as in real life, let alone a single season. Plus, we don't play the same schedule in the sim as we do RL, and there's another thing. Which makes the argument trying to compare how a player plays in RL to how they play in the sim, and then basing an increase on that, completely invalid. Instead, what we are left with is essentially a tier system. If player X falls into the top 25-30% of the stat, they get Y rating, and so on and so forth. However, assists are so fucked up. The sim is from the world of the ISO META, so it plays that way which is why assist numbers are so low. I don't think you are trying to make that argument, Kevin, but I am not really sure what you are trying to argue. Wasn't trying to argue, but just say that before now, I have never heard that argument that Walt posted. |
|
|
|
Post by Charles Barkley on Jan 28, 2017 18:13:23 GMT
Furthermore, everyone rating Plumlee in the high 70's really low 80s just doesn't understand the way the Blazers play basketball.
A lot of people are quick to look at the guys one the Blazers and say McCollum and Lillard are what their offense hinges on, and its not true. McCollum and Lillard are having good years, like they were last year. McCollum, by all accounts, is having a better year this year than last year, and Lillard is only down in %'s. Which says a lot about Lillard, his numbers are up and his %'s are down, which means he is both taking more shots and missing more shots. And efficiency is a big thing in the NBA last time I checked. If a team's best player is having a worse year efficiency wise, their record will be worse, and it is usually a lot worse. But, as I said, the Blazers offense, and defense for that matter, does not hinge on Lillard; it hinges on Plumlee.
Mason Plumlee is having a career year, his stats are up in every category, besides FT%, and his minutes are only up 3 per game. Believe it or not, and I know it is hard to, Joakim Noah was actually a good player at one point in his career. That point really started the 2011-2012 season.
In 2011-2012, Joakim Noah put up this stat line in 30.4 Minutes per game:
3.9-7.7 FGM-FGA for 50.8% FG, 74.8 % FT, 3.8 ORB, 6.1 DRB for 9.9 RPG, 2.5 AST, 1.4 BLK, 0.6 STL, 10.2 PPG, and 1.4 TOV
Mason's stat line this season in 28 MPG is:
4.4-8.3 FGM-FGA for 53.1% FG, 55.7 FT%, 2.1 ORB, 5.8 DRB for 7.9 RPG, 4.2 AST, 1.2 BLK, 0.9 STL, 11.1 PPG, 1.6 TOV.
Adjusting those numbers for PER 36, with Noah on top, we have:
4.6-9.1 for 50.8 FG%, 74.8 FT%, 4.5 ORB, 7.2 DRB, for 11.7 RPG, 3.0 AST, 0.8 STL, 1.7 BLK, 12.1 PPG
5.7-10.7 for 53.1% FG, 55.7% FT, 2.7 ORB, 7.4 DRB for 10.1 RPG, 5.4 AST, 1.2 STL, 1.7 BLK, 12.1 PPG
As you can see, Plumlee is much more efficient from the field while being far worse in FT, better in DRB, but worse in ORB, much better in AST, better in STL, push in BLK, push in points, but Plumlee's all come from the field.
Plumlee's FT% in the sim is currently 48 and the first thing we will do is adjust it to 56, so that's 8 points right there. His O-Aware is 66, but averages just only 1.3 AST per game less than their star point guard who plays 8 more minutes per game than Plumlee and has an O-Aware in the sim of 91. So that needs a lot of love. As does his pass, as illustrated in the posts above. Plumlee's STL and BLK are both 54's, and his BLK definitely needs to be increased. His Off-Reb and D-Reb are both 80's, and in all honesty, his Off-Reb could probably drop 10 and his D-Reb increase 10. He probably also needs a stamina boost, as 74 is kinda low.
So half a point right away goes to FT, his FG in game is appropriate but his Inside could use some points. His O-Aware and Pass are where the points really need to go, as that is his best area of his game and it is no where close to being appropriately rated. BLK could easily take 20 points, so that 1.25 OVR.
Looking at it all, 4 points isn't enough. 4 Points is probably what needs to go on his passing alone. But of course, he won't get the rating he needs, so it will likely suffer.
All in all, think he needs about 7 rating points.
83
|
|
|
|
Post by Charles Barkley on Jan 28, 2017 18:19:01 GMT
I think that is because we are trying to say that the sim is fucked up. If you think about it, the only time the sim league player should be averaging the exact same as he is in real life in the sim is if: 1. First he has to be rated completely accurately on that given day to what he is like on that same game in real life. 2. The team he is on needs to be the same in both RL and in game, and they too need to be rated accurately. 3. The team has to disperse their minutes in the same fashion as both RL and in-game 4. Their opponent must also follow rules 2 and 3. 5. The player needs to have the same stat opportunities in the game as in RL, i.e. both teams need to play exactly the same way, in both style, scheme, minutes, everything. Everything needs to be the exact same in both the sim and RL in order for a player to produce identical numbers in the sim that aligns with the real life game. Then, and only then, can a fair assessment be made as to the player's rating and attribute distribution. Now those 5 things can all be correct, and the sim could still fuck it all up. That being said, think about how many of those things are different in the sim than in real life. I think the Warriors are the only team that have the same two teammates that are rated above an 82 as in game and real life. And I don't just mean the Warriors, I mean, say if the Rockets had KAT and Wiggins, then it'd be the same. But the other 10 guys likely aren't. So you have all these different factors that we just plain and simple can't control, which is why a player's sim stats will never be identical for a game as in real life, let alone a single season. Plus, we don't play the same schedule in the sim as we do RL, and there's another thing. Which makes the argument trying to compare how a player plays in RL to how they play in the sim, and then basing an increase on that, completely invalid. Instead, what we are left with is essentially a tier system. If player X falls into the top 25-30% of the stat, they get Y rating, and so on and so forth. However, assists are so fucked up. The sim is from the world of the ISO META, so it plays that way which is why assist numbers are so low. I don't think you are trying to make that argument, Kevin, but I am not really sure what you are trying to argue. Wasn't trying to argue, but just say that before now, I have never heard that argument that Walt posted. Ahh ok, was just kind of confused as to what you were trying to say what Walt was saying. To make the argument that LeBron gets 9 RPG in RL, and only like 7 in the sim, but Durant gets only 8 IRL, so he needs to have a lot less than LeBron because he gets 9 in game, kind of goes along with what I was saying though. You can't compare sim stats to RL stats, or to sim players for that matter, because you're comparing them on based on their RL numbers. The players need to be rated in their tier according to how they perform in their position within their category, i.e. Plumlee should have a high pass rating versus other bigs, but versus point guards it should be relatively average. Which is what an 85 in pass rating would do. For example, if a point guard were getting 1.2 blocks per game, we'd be saying that's nuts, he's the best blocking PG in the game. And we'd rate him somewhere in the 80's, which is about par for the course for the big men. And that's what we need to do with Plumlee and his pass rating. |
|
Kevin Hollis
Former Thunder GM for 7 years
All Star
Posts: 2,838
|
Post by Kevin Hollis on Jan 28, 2017 18:32:08 GMT
Hey, I agree. People are greatly underrating him just because he isn't a big name.
|
|
|
|
Post by Walt Frazier on Jan 29, 2017 1:34:27 GMT
Ahh ok, was just kind of confused as to what you were trying to say what Walt was saying. To make the argument that LeBron gets 9 RPG in RL, and only like 7 in the sim, but Durant gets only 8 IRL, so he needs to have a lot less than LeBron because he gets 9 in game, kind of goes along with what I was saying though. You can't compare sim stats to RL stats, or to sim players for that matter, because you're comparing them on based on their RL numbers. The players need to be rated in their tier according to how they perform in their position within their category, i.e. Plumlee should have a high pass rating versus other bigs, but versus point guards it should be relatively average. Which is what an 85 in pass rating would do. For example, if a point guard were getting 1.2 blocks per game, we'd be saying that's nuts, he's the best blocking PG in the game. And we'd rate him somewhere in the 80's, which is about par for the course for the big men. And that's what we need to do with Plumlee and his pass rating. That's actually not what I normally do. 1.2 blocks is 1.2 blocks. I don't trust the game to figure out we're saying he's good for a guard. I think a guard at an 80-something block rating will average 2+ blocks in our game, just like a Center with an 80-something block rating. That would make sense but I really don't think that's how this game works. Jump may play into it but then guards all have higher jump ratings than bigs do so that would cancel out. |
|
|
|
Post by Mike Krzyzewski on Jan 30, 2017 2:32:15 GMT
Please change my vote from 78 to 79.
|
|
|
|
Post by Brandon Roy on Jan 31, 2017 10:10:15 GMT
80 for me.. He is a great passer for a bigman though he lack post moves but still a skilled big man.
|
|
|
|
Post by Jeremiah Hill on Jan 31, 2017 23:58:31 GMT
What happens if we take the roster as currently rated. Reorganized it to actual real rosters sim a season and see what poops out?
|
|
|
|
Post by James Kay on Feb 1, 2017 0:20:35 GMT
What happens if we take the roster as currently rated. Reorganized it to actual real rosters sim a season and see what poops out? i second this motion |
|
|
|
Post by Walt Frazier on Feb 1, 2017 0:45:15 GMT
What happens if we take the roster as currently rated. Reorganized it to actual real rosters sim a season and see what poops out? 1) Sim engine has its own issues which we've said multiple times. 2) We can't control the style of play the sim uses for each team. 3) We'd want to sim many many many seasons and then take the average of those seasons. The sim also seems to try to simulate the randomness of results. Otherwise the better rated team would win every single time, which they don't. I guess it would be fun but I don't know if it would actually be that illuminative. |
|
|
|
Post by Jeremiah Hill on Feb 1, 2017 4:26:32 GMT
What happens if we take the roster as currently rated. Reorganized it to actual real rosters sim a season and see what poops out? 1) Sim engine has its own issues which we've said multiple times. 2) We can't control the style of play the sim uses for each team. 3) We'd want to sim many many many seasons and then take the average of those seasons. The sim also seems to try to simulate the randomness of results. Otherwise the better rated team would win every single time, which they don't. I guess it would be fun but I don't know if it would actually be that illuminative. Well a sim would basically weight each thing. Scoring vs Defense, your Rebounding/jump/strength vs theirs. etc Do we know for a fact if the coaching of each team is different in NBA Live? Or if it automatically changes based on roster and whos on it. |
|
|
|
Post by Walt Frazier on Feb 1, 2017 14:13:20 GMT
1) Sim engine has its own issues which we've said multiple times. 2) We can't control the style of play the sim uses for each team. 3) We'd want to sim many many many seasons and then take the average of those seasons. The sim also seems to try to simulate the randomness of results. Otherwise the better rated team would win every single time, which they don't. I guess it would be fun but I don't know if it would actually be that illuminative. Well a sim would basically weight each thing. Scoring vs Defense, your Rebounding/jump/strength vs theirs. etc Do we know for a fact if the coaching of each team is different in NBA Live? Or if it automatically changes based on roster and whos on it. No, just my experience with this Sim engine being used in multiple sim leagues over the past 10 years or so. Ian would likely echo the general sentiment I'm talking about as well. |
|
|
|
Post by Charles Barkley on Feb 1, 2017 15:11:16 GMT
1) Sim engine has its own issues which we've said multiple times. 2) We can't control the style of play the sim uses for each team. 3) We'd want to sim many many many seasons and then take the average of those seasons. The sim also seems to try to simulate the randomness of results. Otherwise the better rated team would win every single time, which they don't. I guess it would be fun but I don't know if it would actually be that illuminative. Well a sim would basically weight each thing. Scoring vs Defense, your Rebounding/jump/strength vs theirs. etc Do we know for a fact if the coaching of each team is different in NBA Live? Or if it automatically changes based on roster and whos on it. We pretty much do. Basketball in 2006-2007 was a lot different than today. Therefore, the game is designed to interpret attributes and play into that style. This is evidenced by the low assist numbers, the game doesn't believe in passing the ball the way they do now. We can't adjust for this stylistic revolution the same way we can with the 3-PT shot. If you want to sim games like that, I am down for it. However, theres a lot more factors at play than just the teams not being properly set up. The schedules aren't the same. The players may be improperly rated. One team may utilize its bench more than others. If we do this, I think we should also do a sim like this in the Real-GM engine and compare results to that of real life. Personally, I like the Real-GM engine more, but it honestly seems like that engine, while representing RL basketball styles and attributes ALOT better, has way more randomness, i.e. a team gets hot and wins a bunch in a row or loses a bunch in a row, the wins and losses come in bunches. |
|
Deleted
Posts: 0
Nov 27, 2024 16:43:38 GMT
|
Post by Deleted on Feb 1, 2017 17:24:36 GMT
If we do this, I think we should also do a sim like this in the Real-GM engine and compare results to that of real life. Personally, I like the Real-GM engine more, but it honestly seems like that engine, while representing RL basketball styles and attributes ALOT better, has way more randomness, i.e. a team gets hot and wins a bunch in a row or loses a bunch in a row, the wins and losses come in bunches. I've been looking into it, and there is no part of the game that carries over from game to game. It's the way the schedule is, you tend to play the same divisions in bunches, though at the end of the season it's a real NBA schedule, you'll play like 10 games against the same shitty division in a row. |
|
Minnesota Timberwolves
LA Clippers
Charlotte Hornets
