What factors determine sub. patterns?

[durvasa]

What attributes/settings determine the substitution patterns in the game? Does anyone who understands the inner working of the Live engine know the answer?

In an attempt to improve the Rocket's offensive flow, I increased the pass rating of PG, SG, and SFs by 10 points, and increased their offensive awareness rating by 15 points. To my surprise, this completely changed the rotation being used and the number of minutes they played. T-Mac, for instance, only played 14 out of 24 minutes, and Yao played only 11 out of 24 minutes in the first half. Neither were ever in foul trouble.

Since the game doesn't provide a way to set team's rotations and the minutes various players get (unless, of course, you did manual substitutions), I think its worthwhile to understand what variables the computer uses to make subsitutions. I used to think it was based solely on fatigue and fouls, but I guess there's more to it than that.

[funkydelic]

Re: What factors determine sub. patterns?

Heres an exerpt of part of the engine
hope this helps.
most of it can be interpreted. As you can see it depends on score as well

START_PCT_DOWN_BY_20 //60% starter plays
START_PCT_DOWN_BY_15 //70% starter plays
START_PCT_DOWN_BY_10 //80% starter plays
START_PCT_DOWN_BY_5 //90% starter plays
START_PCT_SCORE_TIED //90% starter plays
START_PCT_UP_BY_5 //90% starter plays
START_PCT_UP_BY_10 //80% starter plays
START_PCT_UP_BY_15 //80% starter plays
START_PCT_UP_BY_20 //70% starter plays

BENCH_DEPTH_DOWN_BY_20{1, 10, 15, 25, 25, 25}
BENCH_DEPTH_DOWN_BY_15{1, 15, 30, 30, 20, 5}
BENCH_DEPTH_DOWN_BY_10{10, 30, 30, 25, 5, 0}
BENCH_DEPTH_DOWN_BY_5 {20, 30, 30, 15, 5, 0}
BENCH_DEPTH_SCORE_TIED{15, 50, 20, 15, 0, 0}
BENCH_DEPTH_UP_BY_5 {20, 30, 30, 15, 5, 0}
BENCH_DEPTH_UP_BY_10 {10, 30, 30, 25, 5, 0}
BENCH_DEPTH_UP_BY_15 {1, 10, 20, 40, 30, 0}
BENCH_DEPTH_UP_BY_20 {1, 1, 15, 15, 20, 50}

PERCENT_PERIOD1_ALTERNATE_LINEUP {2}

PERCENT_PERIOD2_ALTERNATE_LINEUP {4}

PERCENT_FOUL_GETS_1_FREE_THROW {5}

PERCENT_FOUL_GETS_2_FREE_THROW {43}

PERCENT_MISSED_FT_CAN_BE_REBOUND {70}

[funkydelic]

Re: What factors determine sub. patterns?

Heres an exerpt of part of the engine
hope this helps.
most of it can be interpreted. As you can see it depends on score as well

START_PCT_DOWN_BY_20 //60% starter plays
START_PCT_DOWN_BY_15 //70% starter plays
START_PCT_DOWN_BY_10 //80% starter plays
START_PCT_DOWN_BY_5 //90% starter plays
START_PCT_SCORE_TIED //90% starter plays
START_PCT_UP_BY_5 //90% starter plays
START_PCT_UP_BY_10 //80% starter plays
START_PCT_UP_BY_15 //80% starter plays
START_PCT_UP_BY_20 //70% starter plays

BENCH_DEPTH_DOWN_BY_20{1, 10, 15, 25, 25, 25}
BENCH_DEPTH_DOWN_BY_15{1, 15, 30, 30, 20, 5}
BENCH_DEPTH_DOWN_BY_10{10, 30, 30, 25, 5, 0}
BENCH_DEPTH_DOWN_BY_5 {20, 30, 30, 15, 5, 0}
BENCH_DEPTH_SCORE_TIED{15, 50, 20, 15, 0, 0}
BENCH_DEPTH_UP_BY_5 {20, 30, 30, 15, 5, 0}
BENCH_DEPTH_UP_BY_10 {10, 30, 30, 25, 5, 0}
BENCH_DEPTH_UP_BY_15 {1, 10, 20, 40, 30, 0}
BENCH_DEPTH_UP_BY_20 {1, 1, 15, 15, 20, 50}

PERCENT_PERIOD1_ALTERNATE_LINEUP {2}

PERCENT_PERIOD2_ALTERNATE_LINEUP {4}

PERCENT_FOUL_GETS_1_FREE_THROW {5}

PERCENT_FOUL_GETS_2_FREE_THROW {43}

PERCENT_MISSED_FT_CAN_BE_REBOUND {70}

[durvasa]

Re: What factors determine sub. patterns?

Heres an exerpt of part of the engine
hope this helps.
most of it can be interpreted. As you can see it depends on score as well

START_PCT_DOWN_BY_20 //60% starter plays
START_PCT_DOWN_BY_15 //70% starter plays
START_PCT_DOWN_BY_10 //80% starter plays
START_PCT_DOWN_BY_5 //90% starter plays
START_PCT_SCORE_TIED //90% starter plays
START_PCT_UP_BY_5 //90% starter plays
START_PCT_UP_BY_10 //80% starter plays
START_PCT_UP_BY_15 //80% starter plays
START_PCT_UP_BY_20 //70% starter plays

BENCH_DEPTH_DOWN_BY_20{1, 10, 15, 25, 25, 25}
BENCH_DEPTH_DOWN_BY_15{1, 15, 30, 30, 20, 5}
BENCH_DEPTH_DOWN_BY_10{10, 30, 30, 25, 5, 0}
BENCH_DEPTH_DOWN_BY_5 {20, 30, 30, 15, 5, 0}
BENCH_DEPTH_SCORE_TIED{15, 50, 20, 15, 0, 0}
BENCH_DEPTH_UP_BY_5 {20, 30, 30, 15, 5, 0}
BENCH_DEPTH_UP_BY_10 {10, 30, 30, 25, 5, 0}
BENCH_DEPTH_UP_BY_15 {1, 10, 20, 40, 30, 0}
BENCH_DEPTH_UP_BY_20 {1, 1, 15, 15, 20, 50}

PERCENT_PERIOD1_ALTERNATE_LINEUP {2}

PERCENT_PERIOD2_ALTERNATE_LINEUP {4}

PERCENT_FOUL_GETS_1_FREE_THROW {5}

PERCENT_FOUL_GETS_2_FREE_THROW {43}

PERCENT_MISSED_FT_CAN_BE_REBOUND {70}

Thanks! I'd like to understand as much about the game as possible. May I ask where you found this? If there's more, I'd certainly like to take a look at it. I have a programming background, so this stuff interests me.

Edit:

Hmm, I'm trying to make some sense out of this. Ok, so the first part says that the likelihood a starter will stay in the game during a stoppage in play depends on the point differential. That makes sense. So, for instance, after a starter reaches a particular fatigue level, at each stoppage in play there's a 90% chance he'll come out if the score is within 10, 80% if its between 15 and 10, 70% if its between 20 and 15, and 60% if its more than 20.

For the second part dealing with bench depth, I presume the 6 arguments refer to the 6th through 11th man on the roster. Does it mean that if the score is tied, there's a 15% chance the coach will only go 1 deep, 50% chance the coach only will go 2 deep, 20% chance the coach will only go 3 deep, etc? I wonder how that works in conjunction with the first part.

Ok, I'll take a stab at it:

If the starting SF has reached a certain fatigue level and the scores are within 5 points AND the 6th man plays SF, there's a 90% chance he'll stay in the game after the next dead ball. If the 6th man doesn't play SF but the 7th man does, then there's a 100-((100-90)(1-.15)) or 92.5% chance he'll stay in the game. And if only the 8th man plays SF, then there's a 100-((100-90)(1-.65)) or 96.5% chance he'll stay in the game. And so on.

That would explain the logic behind how a starter gets subbed out. But how is it decided when a starter gets subbed back in?

Also, I think that while point differential is an important factor, there are other stategic factors which should factor in. For instance, a coach should be less willing to sub out a starter (and particularly a star player) in the closing minutes of a game. And I'm thinking that it does take into account more than simply fatigue, point differential, and fouls, since alterning certain player attributes of my starters and bench players (pass rating and offensive awareness) changed my rotation.

[durvasa]

Re: What factors determine sub. patterns?

Heres an exerpt of part of the engine
hope this helps.
most of it can be interpreted. As you can see it depends on score as well

START_PCT_DOWN_BY_20 //60% starter plays
START_PCT_DOWN_BY_15 //70% starter plays
START_PCT_DOWN_BY_10 //80% starter plays
START_PCT_DOWN_BY_5 //90% starter plays
START_PCT_SCORE_TIED //90% starter plays
START_PCT_UP_BY_5 //90% starter plays
START_PCT_UP_BY_10 //80% starter plays
START_PCT_UP_BY_15 //80% starter plays
START_PCT_UP_BY_20 //70% starter plays

BENCH_DEPTH_DOWN_BY_20{1, 10, 15, 25, 25, 25}
BENCH_DEPTH_DOWN_BY_15{1, 15, 30, 30, 20, 5}
BENCH_DEPTH_DOWN_BY_10{10, 30, 30, 25, 5, 0}
BENCH_DEPTH_DOWN_BY_5 {20, 30, 30, 15, 5, 0}
BENCH_DEPTH_SCORE_TIED{15, 50, 20, 15, 0, 0}
BENCH_DEPTH_UP_BY_5 {20, 30, 30, 15, 5, 0}
BENCH_DEPTH_UP_BY_10 {10, 30, 30, 25, 5, 0}
BENCH_DEPTH_UP_BY_15 {1, 10, 20, 40, 30, 0}
BENCH_DEPTH_UP_BY_20 {1, 1, 15, 15, 20, 50}

PERCENT_PERIOD1_ALTERNATE_LINEUP {2}

PERCENT_PERIOD2_ALTERNATE_LINEUP {4}

PERCENT_FOUL_GETS_1_FREE_THROW {5}

PERCENT_FOUL_GETS_2_FREE_THROW {43}

PERCENT_MISSED_FT_CAN_BE_REBOUND {70}

Thanks! I'd like to understand as much about the game as possible. May I ask where you found this? If there's more, I'd certainly like to take a look at it. I have a programming background, so this stuff interests me.

Edit:

Hmm, I'm trying to make some sense out of this. Ok, so the first part says that the likelihood a starter will stay in the game during a stoppage in play depends on the point differential. That makes sense. So, for instance, after a starter reaches a particular fatigue level, at each stoppage in play there's a 90% chance he'll come out if the score is within 10, 80% if its between 15 and 10, 70% if its between 20 and 15, and 60% if its more than 20.

For the second part dealing with bench depth, I presume the 6 arguments refer to the 6th through 11th man on the roster. Does it mean that if the score is tied, there's a 15% chance the coach will only go 1 deep, 50% chance the coach only will go 2 deep, 20% chance the coach will only go 3 deep, etc? I wonder how that works in conjunction with the first part.

Ok, I'll take a stab at it:

If the starting SF has reached a certain fatigue level and the scores are within 5 points AND the 6th man plays SF, there's a 90% chance he'll stay in the game after the next dead ball. If the 6th man doesn't play SF but the 7th man does, then there's a 100-((100-90)(1-.15)) or 92.5% chance he'll stay in the game. And if only the 8th man plays SF, then there's a 100-((100-90)(1-.65)) or 96.5% chance he'll stay in the game. And so on.

That would explain the logic behind how a starter gets subbed out. But how is it decided when a starter gets subbed back in?

Also, I think that while point differential is an important factor, there are other stategic factors which should factor in. For instance, a coach should be less willing to sub out a starter (and particularly a star player) in the closing minutes of a game. And I'm thinking that it does take into account more than simply fatigue, point differential, and fouls, since alterning certain player attributes of my starters and bench players (pass rating and offensive awareness) changed my rotation.

[durvasa]

Re: What factors determine sub. patterns?

Ok, I played some more games today, I don't think there has to be more to it than simply point differential. Also, those numbers for bench depth (if I'm interpreting it correctly) can't be right. I have Bob Sura as my 10th man, my only bench player who plays SG, and he subs in for T-Mac everytime in the first quarter, even in close games. But according to those numbers, the 10th man should only sub in 5% of time, which isn't much at all. But still, he seems to get in the game as frequently as any of the others (and plays the same number minutes).

Also, I don't understand how substitution frequency comes into play.

[durvasa]

Re: What factors determine sub. patterns?

Ok, I played some more games today, I don't think there has to be more to it than simply point differential. Also, those numbers for bench depth (if I'm interpreting it correctly) can't be right. I have Bob Sura as my 10th man, my only bench player who plays SG, and he subs in for T-Mac everytime in the first quarter, even in close games. But according to those numbers, the 10th man should only sub in 5% of time, which isn't much at all. But still, he seems to get in the game as frequently as any of the others (and plays the same number minutes).

Also, I don't understand how substitution frequency comes into play.

[SonicMage]

Re: What factors determine sub. patterns?

Ok, I played some more games today, I don't think there has to be more to it than simply point differential. Also, those numbers for bench depth (if I'm interpreting it correctly) can't be right. I have Bob Sura as my 10th man, my only bench player who plays SG, and he subs in for T-Mac everytime in the first quarter, even in close games. But according to those numbers, the 10th man should only sub in 5% of time, which isn't much at all. But still, he seems to get in the game as frequently as any of the others (and plays the same number minutes).

Also, I don't understand how substitution frequency comes into play.

Each player has a fatigue rating also, which determines how many minutes per game that player plays. If you are playing with default rosters and ratings, Sura has the highest fatigue rating for your bench, a 52, which is quite high for a bench player. This essentially makes him the 6th man of the team, even though he may actually be placed in the tenth roster spot. All these factors are combined and probably weighted to determine substitution.

My guess is substitution frequency has to do with how low the stamina meter can get before the AI determines that a player needs to be subbed for.

[SonicMage]

Re: What factors determine sub. patterns?

Ok, I played some more games today, I don't think there has to be more to it than simply point differential. Also, those numbers for bench depth (if I'm interpreting it correctly) can't be right. I have Bob Sura as my 10th man, my only bench player who plays SG, and he subs in for T-Mac everytime in the first quarter, even in close games. But according to those numbers, the 10th man should only sub in 5% of time, which isn't much at all. But still, he seems to get in the game as frequently as any of the others (and plays the same number minutes).

Also, I don't understand how substitution frequency comes into play.

Each player has a fatigue rating also, which determines how many minutes per game that player plays. If you are playing with default rosters and ratings, Sura has the highest fatigue rating for your bench, a 52, which is quite high for a bench player. This essentially makes him the 6th man of the team, even though he may actually be placed in the tenth roster spot. All these factors are combined and probably weighted to determine substitution.

My guess is substitution frequency has to do with how low the stamina meter can get before the AI determines that a player needs to be subbed for.

[durvasa]

Re: What factors determine sub. patterns?

Each player has a fatigue rating also, which determines how many minutes per game that player plays. If you are playing with default rosters and ratings, Sura has the highest fatigue rating for your bench, a 52, which is quite high for a bench player. This essentially makes him the 6th man of the team, even though he may actually be placed in the tenth roster spot. All these factors are combined and probably weighted to determine substitution.

My guess is substitution frequency has to do with how low the stamina meter can get before the AI determines that a player needs to be subbed for.

By fatigue rating, are you talking about stamina? The lower the number, the faster the player gets fatigued, right? That would mean that, at 52, Sura shouldn't get many minutes. Or do I have it backwards?

The annoyance for me is that if EA isn't going to allow you edit your rotations, they can at least make it more clear on how it calculates the substitution patterns.

[durvasa]

Re: What factors determine sub. patterns?

Each player has a fatigue rating also, which determines how many minutes per game that player plays. If you are playing with default rosters and ratings, Sura has the highest fatigue rating for your bench, a 52, which is quite high for a bench player. This essentially makes him the 6th man of the team, even though he may actually be placed in the tenth roster spot. All these factors are combined and probably weighted to determine substitution.

My guess is substitution frequency has to do with how low the stamina meter can get before the AI determines that a player needs to be subbed for.

By fatigue rating, are you talking about stamina? The lower the number, the faster the player gets fatigued, right? That would mean that, at 52, Sura shouldn't get many minutes. Or do I have it backwards?

The annoyance for me is that if EA isn't going to allow you edit your rotations, they can at least make it more clear on how it calculates the substitution patterns.

[SonicMage]

Re: What factors determine sub. patterns?

By fatigue rating, are you talking about stamina? The lower the number, the faster the player gets fatigued, right? That would mean that, at 52, Sura shouldn't get many minutes. Or do I have it backwards?

You have it backwards. Higher ratings mean better stats. Therefore, a higher stamina rating means the player can play longer. Most starters have stamina in the 70's and 80's. Darko has 15, just to compare.

[SonicMage]

Re: What factors determine sub. patterns?

By fatigue rating, are you talking about stamina? The lower the number, the faster the player gets fatigued, right? That would mean that, at 52, Sura shouldn't get many minutes. Or do I have it backwards?

You have it backwards. Higher ratings mean better stats. Therefore, a higher stamina rating means the player can play longer. Most starters have stamina in the 70's and 80's. Darko has 15, just to compare.

[durvasa]

Re: What factors determine sub. patterns?

You have it backwards. Higher ratings mean better stats. Therefore, a higher stamina rating means the player can play longer. Most starters have stamina in the 70's and 80's. Darko has 15, just to compare.

Huh? That's exactly what I said. You said before that Sura had a good "fatigue rating" and it was at 52. Actually, this is a stamina rating, and its not very good at all.

The lower your stamina rating, the you faster you get fatigued (tired). Right?

Edit: Ok, I understand what you mean. But Sura's stamina rating isn't very high compared to others on my bench. I'll try lowering it to the high 30s and see if that helps.

[durvasa]

Re: What factors determine sub. patterns?

You have it backwards. Higher ratings mean better stats. Therefore, a higher stamina rating means the player can play longer. Most starters have stamina in the 70's and 80's. Darko has 15, just to compare.

Huh? That's exactly what I said. You said before that Sura had a good "fatigue rating" and it was at 52. Actually, this is a stamina rating, and its not very good at all.

The lower your stamina rating, the you faster you get fatigued (tired). Right?

Edit: Ok, I understand what you mean. But Sura's stamina rating isn't very high compared to others on my bench. I'll try lowering it to the high 30s and see if that helps.