Ben E Lou
09-01-2004, 06:37 AM
I'm copying and pasting this from the OOTP Boards. Please comment with your thoughts on this. Either here, or in the original thread, which can be found HERE. (http://www.ootpdevelopments.com/board/showthread.php?t=73132)
--Ben
More On Aging/Development (Everyone who has tested/studied/fretted, please read!)
<hr style="color: rgb(170, 170, 170);" size="1"> <!-- / icon and title --> <!-- message --> Just trying to differentiate what we know (and have tested) from the conjecture. Feel free to contradict anything I put here, and post your observations as well. I want to go to Markus with some concrete stuff to make suggestions for the update.
1. It is clear as a bell that with default settings and fictional rosters, over time the league becomes far too old.
I don't need to know the hard numbers for this one. Person after person has reported that after 10-15 years, their fictional leagues have very few starters that are even 28 or 29 years old--let alone the real MLB peak of 26.
2. Increasing Player Development to +200% is helpful to a degree, and seems to have few/little negative side effects, but isn't enough to solve the problem on its own.
The sentence above pretty much says it all on this topic.
3. Increasing Player Creation Modifiers significantly can lower the average league age significantly, but has some undesirable side effects.
Specifically, increasing the modifiers to the point that is needed to get things in line greatly shifts the ratings scale--to the point where it becomes almost unusable, since ratings over 100 can only be seen as 100. With the creation modifiers in place to do the trick, age-wise, there are literally dozens of players with, for example, a Contact rating of 100. For those who choose to play with ratings on, this is unacceptable.
Potential helpful "easy" fix: Give the option for TRUE ratings to be shown, rather than masking all ratings>100.
<s>4. From reading the various threads on this issue, there seems to be disagreement as to the best use of Player Evaluation AI to help with player age.</s><s>
I've read some posts that indicate a feeling that more emphasis on stats lowers the league age, while others believe that more emphasis on ratings lowers the league age.
Solution: I'm running a test right now to come up with something concrete on this. I'll run a league for 20 seasons with high emphasis on ratings and look at player ages, the restart the same league for another 20 seasons with high emphasis on stats.</s>
EDIT: Testing has now shown that using high ratings evaluation lowers the average age of >350AB hitters by approximately 1/2 year.
5. There doesn't seem to be a consensus on what difference, if any, the number of draft rounds used makes.
There is a seemingly-logical theory out there that the more draft rounds there are, the more likely that a youngster will break out. It would seem to me that this is an area of diminishing returns after a while. I've recently discovered that using a 12-round draft generally keeps all three levels of the minors full, but not overfull.
Suggestion: Someone needs to do a control study on this, to give us a concrete answer as to what difference, if any, the number of draft rounds used makes.
6. According to that 3-page study, with default settings, ability is peaking at age 35. Quote:
<table border="0" cellpadding="6" cellspacing="0" width="100%"> <tbody><tr> <td class="alt2" style="border: 1px inset ;"> For ability rating progression by age, we will look at rating average first. The average ability steadily increases with age with a more or less a uniform speed until age of 35, where the rating average hit the peak. </td> </tr> </tbody></table>
I confess that I didn't read all the charts, graphs and numbers, but the two sentences above really concern me. This seems like a fundamental flaw in the engine.
7. I wonder if the other fundamental player development flaw is coming into play in this as well--10-year minor leaguers who suddenly become good players after age 30.
There seem to be a decent number of these unrealistic guys. For example, this guy (http://www.younglifenorthdekalb.com/ootp603/p3047.html) spent seven years as a mediocre minor leaguer, then got solid production in AAA for four years, then got called up and became a solid big leaguer at age 33. If this happened on the rare occasion, it would be kind of neat, but it happens far too often as it is now. Other examples of guys like this from the same league:
Five years as a crappy Single-A player, 13 years in the minors (http://www.younglifenorthdekalb.com/ootp603/p3092.html)
Ten years as a less-than-mediocre minor leaguer, then becomes a solid big-leaguer (http://www.younglifenorthdekalb.com/ootp603/p3092.html)
Did the AI let this guy hang around for 17 minor league seasons just because it is pre-programmed to be partial to guys named Markus??? (http://www.younglifenorthdekalb.com/ootp603/p2098.html)
I could list many more, but you get the idea. My point here is that this is in a league with +200% player development, and a 12-round draft. These guys should have been cut, moved out, and replaced by younger players. In real life, typically speaking if a guy hits age 24-25 and hasn't progressed above single-A, he will get cut, and no one will pick him up. In OOTP6, too often he sticks around for a decade or more, gets one of those random talent increases, and shows up as a 33-year-old rookie.
8. There seems to be disagreement as to the impact of increasing/decreasing Aging.
The one thing that seems to be a consensus is that increasing or decreasing aging, like increasing or decreasing player development, doesn't have a huge impact.
Suggestion: Another study is needed here. EDIT: Running a test on this now.
9: BOTTOM LINE: We're working WAY out on the margins of the game's ability to be user-tweaked, and still haven't solved this issue. There is clearly a developer tweak needed.
The default settings should be at or very near real life, allowing the user to tweak things to either side of real life.
Comments/thoughts/corrections/additions??? Also, if two people could jump on those other two studies, it would be great. I should have something definitive on ratings vs. stats by the end of the day Wednesday, if not earlier.
--Ben
--Ben
More On Aging/Development (Everyone who has tested/studied/fretted, please read!)
<hr style="color: rgb(170, 170, 170);" size="1"> <!-- / icon and title --> <!-- message --> Just trying to differentiate what we know (and have tested) from the conjecture. Feel free to contradict anything I put here, and post your observations as well. I want to go to Markus with some concrete stuff to make suggestions for the update.
1. It is clear as a bell that with default settings and fictional rosters, over time the league becomes far too old.
I don't need to know the hard numbers for this one. Person after person has reported that after 10-15 years, their fictional leagues have very few starters that are even 28 or 29 years old--let alone the real MLB peak of 26.
2. Increasing Player Development to +200% is helpful to a degree, and seems to have few/little negative side effects, but isn't enough to solve the problem on its own.
The sentence above pretty much says it all on this topic.
3. Increasing Player Creation Modifiers significantly can lower the average league age significantly, but has some undesirable side effects.
Specifically, increasing the modifiers to the point that is needed to get things in line greatly shifts the ratings scale--to the point where it becomes almost unusable, since ratings over 100 can only be seen as 100. With the creation modifiers in place to do the trick, age-wise, there are literally dozens of players with, for example, a Contact rating of 100. For those who choose to play with ratings on, this is unacceptable.
Potential helpful "easy" fix: Give the option for TRUE ratings to be shown, rather than masking all ratings>100.
<s>4. From reading the various threads on this issue, there seems to be disagreement as to the best use of Player Evaluation AI to help with player age.</s><s>
I've read some posts that indicate a feeling that more emphasis on stats lowers the league age, while others believe that more emphasis on ratings lowers the league age.
Solution: I'm running a test right now to come up with something concrete on this. I'll run a league for 20 seasons with high emphasis on ratings and look at player ages, the restart the same league for another 20 seasons with high emphasis on stats.</s>
EDIT: Testing has now shown that using high ratings evaluation lowers the average age of >350AB hitters by approximately 1/2 year.
5. There doesn't seem to be a consensus on what difference, if any, the number of draft rounds used makes.
There is a seemingly-logical theory out there that the more draft rounds there are, the more likely that a youngster will break out. It would seem to me that this is an area of diminishing returns after a while. I've recently discovered that using a 12-round draft generally keeps all three levels of the minors full, but not overfull.
Suggestion: Someone needs to do a control study on this, to give us a concrete answer as to what difference, if any, the number of draft rounds used makes.
6. According to that 3-page study, with default settings, ability is peaking at age 35. Quote:
<table border="0" cellpadding="6" cellspacing="0" width="100%"> <tbody><tr> <td class="alt2" style="border: 1px inset ;"> For ability rating progression by age, we will look at rating average first. The average ability steadily increases with age with a more or less a uniform speed until age of 35, where the rating average hit the peak. </td> </tr> </tbody></table>
I confess that I didn't read all the charts, graphs and numbers, but the two sentences above really concern me. This seems like a fundamental flaw in the engine.
7. I wonder if the other fundamental player development flaw is coming into play in this as well--10-year minor leaguers who suddenly become good players after age 30.
There seem to be a decent number of these unrealistic guys. For example, this guy (http://www.younglifenorthdekalb.com/ootp603/p3047.html) spent seven years as a mediocre minor leaguer, then got solid production in AAA for four years, then got called up and became a solid big leaguer at age 33. If this happened on the rare occasion, it would be kind of neat, but it happens far too often as it is now. Other examples of guys like this from the same league:
Five years as a crappy Single-A player, 13 years in the minors (http://www.younglifenorthdekalb.com/ootp603/p3092.html)
Ten years as a less-than-mediocre minor leaguer, then becomes a solid big-leaguer (http://www.younglifenorthdekalb.com/ootp603/p3092.html)
Did the AI let this guy hang around for 17 minor league seasons just because it is pre-programmed to be partial to guys named Markus??? (http://www.younglifenorthdekalb.com/ootp603/p2098.html)
I could list many more, but you get the idea. My point here is that this is in a league with +200% player development, and a 12-round draft. These guys should have been cut, moved out, and replaced by younger players. In real life, typically speaking if a guy hits age 24-25 and hasn't progressed above single-A, he will get cut, and no one will pick him up. In OOTP6, too often he sticks around for a decade or more, gets one of those random talent increases, and shows up as a 33-year-old rookie.
8. There seems to be disagreement as to the impact of increasing/decreasing Aging.
The one thing that seems to be a consensus is that increasing or decreasing aging, like increasing or decreasing player development, doesn't have a huge impact.
Suggestion: Another study is needed here. EDIT: Running a test on this now.
9: BOTTOM LINE: We're working WAY out on the margins of the game's ability to be user-tweaked, and still haven't solved this issue. There is clearly a developer tweak needed.
The default settings should be at or very near real life, allowing the user to tweak things to either side of real life.
Comments/thoughts/corrections/additions??? Also, if two people could jump on those other two studies, it would be great. I should have something definitive on ratings vs. stats by the end of the day Wednesday, if not earlier.
--Ben