Fictional Leagues and Created Player Stats

Hambil · 11-11-2018, 03:07 PM

I've been playing a lot of Fictional Leagues lately and have noticed that the generated players are underwhelming. To get and maintain a major league quality team (1 all star and at least two 4 star players is the benchmark I'm using) is nearly impossible. The best solution I've found is to created a historic feeder league where I can draft some higher quality historic players but them it's really just a historic league. I've tried messing with the player generation numbers but they effect all players. There does not seem to be a way tell it you just want a few more star players. Anyone have a way to craft a good fictional league they can share?

krownroyal83 · 11-12-2018, 12:14 PM

I ran into this problem in my 14 team fictional league. Draft classes seemed to be getting worse after 7 or 8 years at least for hitters so ended up making my own draft classes each year with a simple formula that i came up with to keep the league balanced among pitchers and batters.

Eugene Church · 11-12-2018, 01:12 PM

I have found the same problem... the AI doesn't create enough rookies that can play well in my league... I noticed it when I had to choose a Rookie of the Year... in my 64-team league, there would only be about 5-10 contributing rookies with good numbers... most of the rookies would just sit on the reserve roster and age, never playing in the league... Each season I would have to purge the over 30s that never played or ever had decent seasons.

As a workaround, I created my own free agents and gave them ratings so they could compete and do well in the league... I would vary the talent ratings so that most of the rookies would be just average or a little below average... but there would some good ones... and occasionally a standout.

This has worked out well for me, but it is a lot of work.

I wish the AI would handle it all and do it well.

I hope someone has a solution to this big problem... one would think the AI would create enough talent to keep the league from declining over time... I would hope Markus designed OOTPB that way.

NoOne · 11-12-2018, 03:10 PM

possible fixes -- change Player Creation Modifiers. League settings->players or whatever. Or, change LTM (modifiers) until you get production you want and adjust eye to the ratings that achieve that production.

i know '19 looks different, but that doesn't mean you cna differentiate the same ways as you always have with a little adaptation.

i'd look at what you want to change, first. adjust the PCM or LTM accordingly. you may need to do both, if you adjust PCMs. that will be a rising tide if you increase, say power PCM. whatever the 'average' draft player potential power was before, it is now higher. the random ebb and flow around it is shifted too (larger range of varied results in most contexts of an increase likely).

-------------
https://www.baseball-reference.com/l...-rookies.shtml

rookies usually aren't so good. the top 1-2 might be, but no guarantee. i didn't look at the pitching. in this link, the top guys are 3-4war. one of them should be excluded at age 28 and PA over 3 years, but still technically a rookie. he was the 'best' by war, i didn't look too closely at this, though.

*and, i'm not a fan of war if you've read a few of my posts, lol..

looking at one real year is dangerous. this may or may not be an 'average' year, and even so doesn't tell you much. most rookies are junk, though.

also, 1 top-shelf guy and 3or4 4-star guys would be a hefty 'average'. there aren't 30 amazing batters in the mlb. if you spread it out, you'd probably get 2-3 decent players and the rest ~middling and below. very few at the top. sometimes the 2-3 wouldn't even be that stunningly good. (ootp "80/80" represents more than the 'top' players.)

perception is skewed by emotion. i think we've all experienced looking up some stats from a decade or more ago and realizing our memories were a bit rose-colored and not quite realistic about that particular team in history. unless you are looking at the '27 yankees. then, likely your perceptions are very near reality most likely, lol.

we like players and think of them as better than what they are. therefore, we think there are more better players than there are. (grammatically correct, read it slowly,

)

it's no surprise the league BA is astonishingly low. most players aren't so good, because average isn't good. that average is skewed high compared to median -- it's a 'false' level of production to expect from a truely average player. an average player's expected production is lower than 'league average' due to distribution curve of talent. further reason why most players aren't so good, relative to what 'good' is statistically.

while any average player is capable of an amazingly good year, it won't happen often. hence, one-year wonders every year crop up. it is inevitable. this is why it takes a larger sample to be confident about a player's ability. 2-3 years, and exceptions for drastic re-working of mechanics/mental approach, which as bystanders we'll never know for sure. think JD Martinez... he changes swing and approach and you can see the black and white difference of his historical sample before and after that point -- a real causal change was made.

Hambil · 11-12-2018, 03:34 PM

I tried that. I think I saw it suggested in another post. It didn't seem to have much effect but more testing is required.

NoOne · 11-13-2018, 04:48 AM

PCM affect new draft classes... so after any change it will take a few years just to see a few of those players from a previous draft class and 20-30 years before the whole league is turned over.

probably have to move the dial 10-20 pts to notice anythign with the eye too.

turn on 100% accuracy and go sort by whichver ratings you beefed up... then zoom out 20 years and compare what you see. how many >80, >70, >60 etc.. use a larger scale even if temporarily while testing.

Hrycaj · 11-13-2018, 06:19 AM

Just going off the original post I would offer this. How do the numbers look? What I'm getting at is that stars are deceiving depending on how you have them set. Are they league wide or position specific? Are 2 star players preforming well statistically? How old is your fictional league? The initial player creation is going to give you a pretty balanced league and your draft classes will seem underwhelming because the game is programmed to keep that balance over time.

You could easily test this. Start a game and turn off stars and ratings for that matter. Let it sim and just look at the numbers. If the league looks legit in your eyes statistically, go on and turn the light back on and you may be surprised at the true ratings of guys, and how they still produce.

If it isn't to your liking then you may want to play with pcm stuff. But you may just have a case of visually falling for a look. Good luck.

phenom · 11-13-2018, 08:04 AM

How many rounds is your draft and how many rounds worth of players does your game create?

Hambil · 11-13-2018, 10:09 AM

Quote:

Originally Posted by phenom

How many rounds is your draft and how many rounds worth of players does your game create?

Default. I hadn't thought about changing that. If I say the draft is 5 rounds but have players for 10 am I twice as likely to get a generated player with good stats? It seems like I would be. While I'm in there I noticed there is a randomness setting for player development. I wonder if moving it from 100 (default) to 200 (maximum) would give me a broader ranger of player talent.

NoOne · 11-14-2018, 02:41 AM

depth of draft is almost entirely based on # of teams, i believe... if you only have ~1 round, that could skew it, of course. so much lost TCR and such plus not enough to fill league.

with a real-world setup or league, the MLB minors require the default or more rounds for sure... do not reduce unless you also change structure of MiL... like reducing # of rookie leagues for some teams.

some teams have 4? rookie teams, 5? .. even with ~4 you need a ton of rookies each year to fill them up along with leftovers from recent drafts not good enough for promotion, of course.

TCR is the randomness setting. this will cause an accurately rated player to change for random reasons... increasing to 200 will make it happen more often. 0 will mean that hte potential they have is all they ever can get... no guarantees it develops, of course.

i don't know if it gives a larger range or just makes it more maleable.

i've been turning it down for a few years now.. i still see >max players and such.

side note to what hycaj mentions above -- improved accuracy makes drafts look thinner too.. if you recently changed that setting or improved scout, it can make draft classes seem worse, but you're just not seeing as many false-positives.

if you hit on 1/2 your first rounders, you'd be a baseball god in real life... expect more failure than winning with draft picks.

11-11-2018, 03:07 PM	#1
Hambil Minors (Rookie Ball) Join Date: Nov 2018 Posts: 25 Thanks: 0 Thanked 4x in 4 posts	Fictional Leagues and Created Player Stats I've been playing a lot of Fictional Leagues lately and have noticed that the generated players are underwhelming. To get and maintain a major league quality team (1 all star and at least two 4 star players is the benchmark I'm using) is nearly impossible. The best solution I've found is to created a historic feeder league where I can draft some higher quality historic players but them it's really just a historic league. I've tried messing with the player generation numbers but they effect all players. There does not seem to be a way tell it you just want a few more star players. Anyone have a way to craft a good fictional league they can share?

11-12-2018, 01:12 PM	#3
Eugene Church Hall Of Famer Join Date: Aug 2002 Posts: 23,474 Thanks: 2,500 Thanked 2,186x in 1,417 posts	I have found the same problem... the AI doesn't create enough rookies that can play well in my league... I noticed it when I had to choose a Rookie of the Year... in my 64-team league, there would only be about 5-10 contributing rookies with good numbers... most of the rookies would just sit on the reserve roster and age, never playing in the league... Each season I would have to purge the over 30s that never played or ever had decent seasons. As a workaround, I created my own free agents and gave them ratings so they could compete and do well in the league... I would vary the talent ratings so that most of the rookies would be just average or a little below average... but there would some good ones... and occasionally a standout. This has worked out well for me, but it is a lot of work. I wish the AI would handle it all and do it well. I hope someone has a solution to this big problem... one would think the AI would create enough talent to keep the league from declining over time... I would hope Markus designed OOTPB that way. Last edited by Eugene Church; 11-12-2018 at 01:18 PM.

11-12-2018, 03:10 PM	#4
NoOne Hall Of Famer Join Date: Apr 2015 Posts: 6,006 Thanks: 605 Thanked 1,615x in 1,309 posts	possible fixes -- change Player Creation Modifiers. League settings->players or whatever. Or, change LTM (modifiers) until you get production you want and adjust eye to the ratings that achieve that production. i know '19 looks different, but that doesn't mean you cna differentiate the same ways as you always have with a little adaptation. i'd look at what you want to change, first. adjust the PCM or LTM accordingly. you may need to do both, if you adjust PCMs. that will be a rising tide if you increase, say power PCM. whatever the 'average' draft player potential power was before, it is now higher. the random ebb and flow around it is shifted too (larger range of varied results in most contexts of an increase likely). ------------- https://www.baseball-reference.com/l...-rookies.shtml rookies usually aren't so good. the top 1-2 might be, but no guarantee. i didn't look at the pitching. in this link, the top guys are 3-4war. one of them should be excluded at age 28 and PA over 3 years, but still technically a rookie. he was the 'best' by war, i didn't look too closely at this, though. and, i'm not a fan of war if you've read a few of my posts, lol.. looking at one real year is dangerous. this may or may not be an 'average' year, and even so doesn't tell you much. most rookies are junk, though. also, 1 top-shelf guy and 3or4 4-star guys would be a hefty 'average'. there aren't 30 amazing batters in the mlb. if you spread it out, you'd probably get 2-3 decent players and the rest ~middling and below. very few at the top. sometimes the 2-3 wouldn't even be that stunningly good. (ootp "80/80" represents more than the 'top' players.) perception is skewed by emotion. i think we've all experienced looking up some stats from a decade or more ago and realizing our memories were a bit rose-colored and not quite realistic about that particular team in history. unless you are looking at the '27 yankees. then, likely your perceptions are very near reality most likely, lol. we like players and think of them as better than what they are. therefore, we think there are more better players than there are. (grammatically correct, read it slowly, ) it's no surprise the league BA is astonishingly low. most players aren't so good, because average isn't good. that average is skewed high compared to median -- it's a 'false' level of production to expect from a truely average player. an average player's expected production is lower than 'league average' due to distribution curve of talent. further reason why most players aren't so good, relative to what 'good' is statistically. while any average player is capable of an amazingly good year, it won't happen often. hence, one-year wonders every year crop up. it is inevitable. this is why it takes a larger sample to be confident about a player's ability. 2-3 years, and exceptions for drastic re-working of mechanics/mental approach, which as bystanders we'll never know for sure. think JD Martinez... he changes swing and approach and you can see the black and white difference of his historical sample before and after that point -- a real causal change was made. Last edited by NoOne; 11-12-2018 at 03:25 PM.*

11-13-2018, 04:48 AM	#6
NoOne Hall Of Famer Join Date: Apr 2015 Posts: 6,006 Thanks: 605 Thanked 1,615x in 1,309 posts	PCM affect new draft classes... so after any change it will take a few years just to see a few of those players from a previous draft class and 20-30 years before the whole league is turned over. probably have to move the dial 10-20 pts to notice anythign with the eye too. turn on 100% accuracy and go sort by whichver ratings you beefed up... then zoom out 20 years and compare what you see. how many >80, >70, >60 etc.. use a larger scale even if temporarily while testing. Last edited by NoOne; 11-13-2018 at 04:51 AM.

11-13-2018, 06:19 AM	#7
Hrycaj All Star Reserve Join Date: Oct 2010 Posts: 570 Thanks: 195 Thanked 151x in 103 posts	Just going off the original post I would offer this. How do the numbers look? What I'm getting at is that stars are deceiving depending on how you have them set. Are they league wide or position specific? Are 2 star players preforming well statistically? How old is your fictional league? The initial player creation is going to give you a pretty balanced league and your draft classes will seem underwhelming because the game is programmed to keep that balance over time. You could easily test this. Start a game and turn off stars and ratings for that matter. Let it sim and just look at the numbers. If the league looks legit in your eyes statistically, go on and turn the light back on and you may be surprised at the true ratings of guys, and how they still produce. If it isn't to your liking then you may want to play with pcm stuff. But you may just have a case of visually falling for a look. Good luck. __________________ Click on my signature to read "A Baseball Story". A fictional dynasty report about the great game of baseball.

11-12-2018, 12:14 PM	#2
krownroyal83 All Star Reserve Join Date: Feb 2015 Posts: 566 Thanks: 171 Thanked 187x in 115 posts	I ran into this problem in my 14 team fictional league. Draft classes seemed to be getting worse after 7 or 8 years at least for hitters so ended up making my own draft classes each year with a simple formula that i came up with to keep the league balanced among pitchers and batters.

11-12-2018, 03:34 PM	#5
Hambil Minors (Rookie Ball) Join Date: Nov 2018 Posts: 25 Thanks: 0 Thanked 4x in 4 posts	I tried that. I think I saw it suggested in another post. It didn't seem to have much effect but more testing is required.

11-13-2018, 08:04 AM	#8
phenom Hall Of Famer Join Date: Mar 2003 Location: Near the Great Wall. On the GOOD side. Posts: 3,520 Thanks: 34 Thanked 190x in 99 posts	How many rounds is your draft and how many rounds worth of players does your game create? __________________ reported

11-14-2018, 02:41 AM	#10
NoOne Hall Of Famer Join Date: Apr 2015 Posts: 6,006 Thanks: 605 Thanked 1,615x in 1,309 posts	depth of draft is almost entirely based on # of teams, i believe... if you only have ~1 round, that could skew it, of course. so much lost TCR and such plus not enough to fill league. with a real-world setup or league, the MLB minors require the default or more rounds for sure... do not reduce unless you also change structure of MiL... like reducing # of rookie leagues for some teams. some teams have 4? rookie teams, 5? .. even with ~4 you need a ton of rookies each year to fill them up along with leftovers from recent drafts not good enough for promotion, of course. TCR is the randomness setting. this will cause an accurately rated player to change for random reasons... increasing to 200 will make it happen more often. 0 will mean that hte potential they have is all they ever can get... no guarantees it develops, of course. i don't know if it gives a larger range or just makes it more maleable. i've been turning it down for a few years now.. i still see >max players and such. side note to what hycaj mentions above -- improved accuracy makes drafts look thinner too.. if you recently changed that setting or improved scout, it can make draft classes seem worse, but you're just not seeing as many false-positives. if you hit on 1/2 your first rounders, you'd be a baseball god in real life... expect more failure than winning with draft picks.