Depth chart wonkiness

swat-spas · 08-16-2019, 11:35 AM

My impression from the manual and threads is an intended setting of "play every 4th game" on a platoon will result in a player getting roughly every 4th start. That's a pretty intuitive setting, if it works.

Testing is not showing this to be the case for me (injuries off, in commish mode, coaches off, average player fatigue, etc). If I'm expecting and get roughly 44-48 games against lefties in a season, most of the "every four" guys get very few starts, more like 0-3, instead of the intended 10-12. These aren't catchers, which is the one position sort of working as intended (usually).

It's not working predictably the other way either. If the intention is something like "every 8", to get in most of the fatigue required starts in a favorable split way, they're often starting 12-15 times against lefties, up to ~1/3 of the time, or every 10th start against RHP, gets a couple but not 10-15 games as I'd expect. Full time platoons and "when starter tired" settings generally work as intended, and if its, say, every third game against both it works. Everything else in between is producing messy and unpredictable numbers of starts.

- related issue with depth charts. If a platoon player is to be used at two different positions with the intention of playing every day against LHP, (e.g. usually starts at 1b but plays rest of time at DH), then they don't move even if a starter is assigned to replace them the appropriate number of games. If they are given some days off, this mostly works, but if I say "every 5th game you get a day off from 1b and are the DH" on the depth chart, that seems like it ought to work too. It works okay usually to make them DH the 6th day. But not the "full time vs L" mode I wanted to set. They just stay at full time 1b.

- I find this is all fine, if tedious and with its own minor glitchiness, to use 7DL to work around it to play out a season, and indeed that works very well for pitching rotations with spot starts and other test factors I wish to look at, and to adjust on the fly if someone is having an amazing or terrible year. But I also do like being able to bang out say 10 test seasons in an hour instead of just one in that time. I am having to adjust the intended playing time often to absurd degrees (1/4 becomes 1/2, 1/8 becomes 1/15, etc) to get someone to do some rough amount of set spot starts in the field rather than anything like the more intuitive figures I had intended to use.

Not sure if I'm doing something wrong in settings or strategy that interferes, is overriding me, or if the depth chart AI is just hopelessly confused by what I want it to do when simming.

- pinch hitting tends to use a lower bench player (who is sometimes not even assigned as one) too much. This is fine to get them an extra few PA for testing but I'd rather the AI use who I listed to max out the platoon splits if a PH is needed when I'm actually simming out a year. Those players are generally much better hitters when they are available, and there is always at least one or two on the list who should be available to PH. They do get used, but not nearly as often as a scrub hitter like a backup INF. Pinch running seems to work more easily to bring in only the faster players, and defensive subbing is doing what I'd want/expect. This though is also acting weird.

Argonaut · 08-16-2019, 11:46 AM

Check out my thread on this issue here: https://www.ootpdevelopments.com/boa...d.php?t=305094

I agree with you that the depth chart/sub system needs a lot of work. It's often hard to get it to do what you want.

swat-spas · 08-16-2019, 12:24 PM

Yeah. The impression from testing I was getting it is it uses the "every 4th" thing for looking back at every day rather than the platoon effect, and if they're playing every day against RHP, say, then they've played every 4th day and it only does it if they get a run of lefties or the other starter is somehow tired. Basically makes spot starting position players practically useless in simming and that you can only use full platoons or "if tired" as the setting. Particularly for LHP since it's rarer to get a run of those that anything over "every 2" or "every 3" is a waste of time. That makes it far, far less intuitive for part-time platoons to figure out what you want to do. I can get something like the right number of games, but it's a mess.

I also noticed that issue for players that you want to back up at 2 or more positions to get something approximating the intended amount of starts at each position there is a mess.

swat-spas · 08-17-2019, 05:43 PM

Workaround, of sorts, for those interested:

I am pretty confident the depth chart AI tracks more based on the number of games played at that position overall, and applies the depth chart rules based on a lookback to check if they've played there recently or not and not based on something like "this player should start ~25% of the time at this position vs LHP".

To get a full-time vs RHP player to start ~1/4 games vs LHP in my league:

Instead of setting them to start "every 4th game" as would be the intuitive setting. I set them as the starter vs LHP, with the backup starting "every 2nd game" and that gave me about 25% starts vs LHP for the intended players. There are more lefties in my league than an average league I believe, but mathematically that worked as I intended.

Similarly to get a player playing full-time vs LHP to "every 3rd game" vs RHP, they were set to start every other game vs RHP (as the backup).

08-16-2019, 11:35 AM	#1
swat-spas Minors (Single A) Join Date: Aug 2002 Posts: 53	Depth chart wonkiness My impression from the manual and threads is an intended setting of "play every 4th game" on a platoon will result in a player getting roughly every 4th start. That's a pretty intuitive setting, if it works. Testing is not showing this to be the case for me (injuries off, in commish mode, coaches off, average player fatigue, etc). If I'm expecting and get roughly 44-48 games against lefties in a season, most of the "every four" guys get very few starts, more like 0-3, instead of the intended 10-12. These aren't catchers, which is the one position sort of working as intended (usually). It's not working predictably the other way either. If the intention is something like "every 8", to get in most of the fatigue required starts in a favorable split way, they're often starting 12-15 times against lefties, up to ~1/3 of the time, or every 10th start against RHP, gets a couple but not 10-15 games as I'd expect. Full time platoons and "when starter tired" settings generally work as intended, and if its, say, every third game against both it works. Everything else in between is producing messy and unpredictable numbers of starts. - related issue with depth charts. If a platoon player is to be used at two different positions with the intention of playing every day against LHP, (e.g. usually starts at 1b but plays rest of time at DH), then they don't move even if a starter is assigned to replace them the appropriate number of games. If they are given some days off, this mostly works, but if I say "every 5th game you get a day off from 1b and are the DH" on the depth chart, that seems like it ought to work too. It works okay usually to make them DH the 6th day. But not the "full time vs L" mode I wanted to set. They just stay at full time 1b. - I find this is all fine, if tedious and with its own minor glitchiness, to use 7DL to work around it to play out a season, and indeed that works very well for pitching rotations with spot starts and other test factors I wish to look at, and to adjust on the fly if someone is having an amazing or terrible year. But I also do like being able to bang out say 10 test seasons in an hour instead of just one in that time. I am having to adjust the intended playing time often to absurd degrees (1/4 becomes 1/2, 1/8 becomes 1/15, etc) to get someone to do some rough amount of set spot starts in the field rather than anything like the more intuitive figures I had intended to use. Not sure if I'm doing something wrong in settings or strategy that interferes, is overriding me, or if the depth chart AI is just hopelessly confused by what I want it to do when simming. - pinch hitting tends to use a lower bench player (who is sometimes not even assigned as one) too much. This is fine to get them an extra few PA for testing but I'd rather the AI use who I listed to max out the platoon splits if a PH is needed when I'm actually simming out a year. Those players are generally much better hitters when they are available, and there is always at least one or two on the list who should be available to PH. They do get used, but not nearly as often as a scrub hitter like a backup INF. Pinch running seems to work more easily to bring in only the faster players, and defensive subbing is doing what I'd want/expect. This though is also acting weird.

08-16-2019, 11:46 AM	#2
Argonaut All Star Starter Join Date: Apr 2019 Posts: 1,085	Check out my thread on this issue here: https://www.ootpdevelopments.com/boa...d.php?t=305094 I agree with you that the depth chart/sub system needs a lot of work. It's often hard to get it to do what you want. __________________ Texas Rangers Dynasty (Est. 1972) Hartford Whalers Dynasty: Brass Bonanza

08-16-2019, 12:24 PM	#3
swat-spas Minors (Single A) Join Date: Aug 2002 Posts: 53	Yeah. The impression from testing I was getting it is it uses the "every 4th" thing for looking back at every day rather than the platoon effect, and if they're playing every day against RHP, say, then they've played every 4th day and it only does it if they get a run of lefties or the other starter is somehow tired. Basically makes spot starting position players practically useless in simming and that you can only use full platoons or "if tired" as the setting. Particularly for LHP since it's rarer to get a run of those that anything over "every 2" or "every 3" is a waste of time. That makes it far, far less intuitive for part-time platoons to figure out what you want to do. I can get something like the right number of games, but it's a mess. I also noticed that issue for players that you want to back up at 2 or more positions to get something approximating the intended amount of starts at each position there is a mess.

08-17-2019, 05:43 PM	#4
swat-spas Minors (Single A) Join Date: Aug 2002 Posts: 53	Workaround, of sorts, for those interested: I am pretty confident the depth chart AI tracks more based on the number of games played at that position overall, and applies the depth chart rules based on a lookback to check if they've played there recently or not and not based on something like "this player should start ~25% of the time at this position vs LHP". To get a full-time vs RHP player to start ~1/4 games vs LHP in my league: Instead of setting them to start "every 4th game" as would be the intuitive setting. I set them as the starter vs LHP, with the backup starting "every 2nd game" and that gave me about 25% starts vs LHP for the intended players. There are more lefties in my league than an average league I believe, but mathematically that worked as I intended. Similarly to get a player playing full-time vs LHP to "every 3rd game" vs RHP, they were set to start every other game vs RHP (as the backup).