Alright Boyz, Here We Go!!!! (OOTP9 First Impressions...)

[Ksyrup]

I understand Ben's thought process - it's like a built in house rule, almost, to make the game more challenging. But that is throwing away realism at the expense of this just being a game, and I'm looking for more than a competitive gaming experience out of this. I'd rather come up with house rules to apply to myself to make the game harder rather than see the AI behave in a way that is devoid of real world logic.

It may be incredibly difficult to do, but I thought (hoped) with the supposed fix to the "zero rating" issue that we might get close to a stat-based AI evaluation process. I don't see it.

[molson]

Bottom line is, what the hell do the AI evaluation settings do? Is there a minimum, hidden, "ratings" %?

Whether someone wants the game to run like Ksyrup or Skydog isn't the point - it's about customization, and what actually can be done.

[Ksyrup]

Originally Posted by molson

It's only "peeking ahead" in terms of probable performance on the field, by using the ratings for the players. A human player can do the same thing, but we'd typically rather actually see a player suck on the field before we pull him, rather than see his ratings drop first and not let him back on the field.

Forget about players of a game, AI versus Human. I'm simply talking about the simulation process in a vacuum. Be it historical or fictional, I want the AI to act not as I would as the human player of this game, but as a real-world manager/Gm would. We need only look at Barry Zito to see the difference here. In OOTP, Zito would have signed that $126M deal and probably started last year in AAA (assuming the game would have shown a ratings drop at the between 2006 and 2007). That's just not realistic at all. Even if they admit their mistake 2-3 years later and dump or demote him, at least they've let his performance dictate their action.

[M GO BLUE!!!]

I built my team by looking at young guys that can play defense, run and have a good eye. Power is secondary to getting on base & causing hell once they get there. Trades and free agents can be hit or miss. I gave the best LF available who looked to be perfect for me $24 mil and he had one good season then slumped. I ended up trading him to lose some salary (cost a few prospects in the deal) Then on the other hand I traded to get a utility man infield depth who was still in AAA and he just took off. He leads the league in stolen bases, is one hell of a player at 2B and is hitting around .330! Two players that I picked up in deals for other players turned out to be steals as well.

[Ksyrup]

Originally Posted by molson

Bottom line is, what the hell do the AI evaluation settings do? Is there a minimum, hidden, "ratings" %?

Whether someone wants the game to run like Ksyrup or Skydog isn't the point - it's about customization, and what actually can be done.

Agreed. Theoretically, when I set the evaluation ratings to 0/50/35/15 (which is what I started with in my first sim), the AI should only have used ratings as a guide where it was completely devoid of stats to evaluate a guy with. And in the Britt Burns example I posted earlier, it's clear that it didn't happen that way. And it was compounded by the decision to continue giving him 1-year extensions, too.

[BigPapi]

Originally Posted by molson

It's only "peeking ahead" in terms of probable performance on the field, by using the ratings for the players. A human player can do the same thing, but we'd typically rather actually see a player suck on the field before we pull him, rather than see his ratings drop first and not let him back on the field.

I can't justify some of the AI decisions I am seeing using this logic. What I am seeing is players lose 3-4% of their talent/ratings- noticeable- but not significant- and certainly not grounds for a trip to the bench or AAA- which is exactly where they end up. This issue is further exacerbated by the AI's decision to replace said player with a clearly inferior one- even using it's own measurement system. I am not saying it's as prevalent as last year- I just can't fathom the algorithm that leads to these screwy decisions. It's almost like the AI cant' stand to leave things alone.

[BigPapi]

Originally Posted by molson

Bottom line is, what the hell do the AI evaluation settings do? Is there a minimum, hidden, "ratings" %?

Whether someone wants the game to run like Ksyrup or Skydog isn't the point - it's about customization, and what actually can be done.

Good point- and I guess that's my gripe; these don't work as advertised, and I am all the more grumpy for it given Markus' assurance that "it's fixed!".

This behavior is not in any way, shape, or form hard to verify if you look for it. It's almost as if there is some other setting that is overridng these settings- if they work at all.

[korme]

Here's a question - in my My Documents folder, I have an OOTP Developments folder with a ton of folders, and also an ootp9 folder that appears to be blank.

Can I delete the second folder?

[INDalltheway]

Originally Posted by korme

Here's a question - in my My Documents folder, I have an OOTP Developments folder with a ton of folders, and also an ootp9 folder that appears to be blank.

Can I delete the second folder?

I deleted it but it came right back after playing again if I recall correctly.

[Ksyrup]

I know this is touched on in the RonCo stuff, but the BB/K issue has to be resolved. It's ridiculous. There are a ton of pitchers in my league with BB/K ratios that are insane - in the bad way. Like 95 BB, 15 K and stuff like that. And worst, I'm seeing guys go from 60 BB / 170 K in Year 1 to 80 BB / 20 K in Year 5. After I sim, I like to go back and look at some of the best and worst-performing teams, and with some of the 100+ loss teams, I'm seeing pitching staffs with not a single pitcher with 100 Ks (one team I looked at topped out at 2 guys at 54 and 52 Ks for over 200 IP each!), and about 75% of them with more BBs than Ks - some of them outrageous numbers (like, oh, a 157 BB / 52 K season!).

[CraigSca]

What kind of league are you doing, Ksyrup? Fictional, historical?

Just hearing this makes me want to vomit. It's so anti-DIPS it's amazing.

Well, not necessarily anti-DIPS, but anti-MLB pitching. You just don't see that.

[Ben E Lou]

Originally Posted by CraigSca

What kind of league are you doing, Ksyrup? Fictional, historical?

Just hearing this makes me want to vomit. It's so anti-DIPS it's amazing.

Well, not necessarily anti-DIPS, but anti-MLB pitching. You just don't see that.

It tends to happen in fictional leagues late in careers, and it happens much more often with RonCo's dev modifiers. It's actually more of a problem of AI teams leaving old guys in there for too many seasons. I've pretty much eliminated it by combining some things I learned from Ron with some adjustments to settings. I hope to be able to post the findings in the next 48 hours. Essentially what's happening is that a guy will have a solid career, but then finish it out with a couple of seasons that look like 5-15, 6.24, 100 BB, 40K. I suspect it's over-favoring fan favorites and/or big-salaried guys when setting the rotation. But like I said, I've been able to pretty much do away with it now.

[Alan T]

Originally Posted by Ben E Lou

It tends to happen in fictional leagues late in careers, and it happens much more often with RonCo's dev modifiers. It's actually more of a problem of AI teams leaving old guys in there for too many seasons. I've pretty much eliminated it by combining some things I learned from Ron with some adjustments to settings. I hope to be able to post the findings in the next 48 hours. Essentially what's happening is that a guy will have a solid career, but then finish it out with a couple of seasons that look like 5-15, 6.24, 100 BB, 40K. I suspect it's over-favoring fan favorites and/or big-salaried guys when setting the rotation. But like I said, I've been able to pretty much do away with it now.

Isn't that exactly the opposite of what people are complaining about regarding a different issue of players that have great seasons and then get dumped into AAA or released afterwards without having a chance to encounter the post-prime of their career? I know my question doesn't have as much to say about the BB/K ratio issue, but it sounds like people are complaining about two opposite issues... 1) Players are released/demoted too soon, and 2) Players are released/demoted too late... How can both be true?

[Ksyrup]

Originally Posted by CraigSca

What kind of league are you doing, Ksyrup? Fictional, historical?

Just hearing this makes me want to vomit. It's so anti-DIPS it's amazing.

Well, not necessarily anti-DIPS, but anti-MLB pitching. You just don't see that.

Historical.

[Ben E Lou]

Originally Posted by Alan T

Isn't that exactly the opposite of what people are complaining about regarding a different issue of players that have great seasons and then get dumped into AAA or released afterwards without having a chance to encounter the post-prime of their career? I know my question doesn't have as much to say about the BB/K ratio issue, but it sounds like people are complaining about two opposite issues... 1) Players are released/demoted too soon, and 2) Players are released/demoted too late... How can both be true?

Two things.

1. That dichotomy is one of the reasons that I speculate that the players that hang on too late has to do with things other than ratings and stats, such as past performance and perhaps fan favorite-ness. The settings that I use to deal with this issue ignore past performance. (I mean, really, who cares what he did two years ago? I don't want the AI handicapped by using past performance. I sure don't use it.)
2. People also like to whine.

[Logan]

Yes, thousands of walks and no Ks is definitely whining.

[Ben E Lou]

"thousands"

[Ben E Lou]

The point is this: when there are complaints at both ends of the spectrum, I tend to suspect that it's not so much systemic as people jumping on their pet issues.

[Alan T]

Originally Posted by Ben E Lou

Two things.

1. That dichotomy is one of the reasons that I speculate that the players that hang on too late has to do with things other than ratings and stats, such as past performance and perhaps fan favorite-ness. The settings that I use to deal with this issue ignore past performance. (I mean, really, who cares what he did two years ago? I don't want the AI handicapped by using past performance. I sure don't use it.)
2. People also like to whine.

I can see if the AI is programed to not want to drop a fan favorite due to a desire for avoiding a Fan interest hit, but I would assume that all of these Cy Young pitchers that then get released would also have been a fan favorite too. At least it doesn't seem in my experience so far that it takes long for a cy young type pitcher to become one of the fan favorites.

As for #2, yeah people do Usually though the whines have some grounding behind them though, or at least some reason for the whine to originate.. It may not be as huge an issue or catastrophic of a problem as they might be leading on, but usually there is at least some issue there.

I'll be curious to see what you comment in 48 hours or whenever you come to some conclusion regarding the issue, it just seemed interesting to me that people have two seemingly opposite scenarios that are issues that I couldn't imagine both to exist.

[Ksyrup]

Originally Posted by Ben E Lou

It tends to happen in fictional leagues late in careers, and it happens much more often with RonCo's dev modifiers. It's actually more of a problem of AI teams leaving old guys in there for too many seasons. I've pretty much eliminated it by combining some things I learned from Ron with some adjustments to settings. I hope to be able to post the findings in the next 48 hours. Essentially what's happening is that a guy will have a solid career, but then finish it out with a couple of seasons that look like 5-15, 6.24, 100 BB, 40K. I suspect it's over-favoring fan favorites and/or big-salaried guys when setting the rotation. But like I said, I've been able to pretty much do away with it now.

Admittedly, the league I'm looking at right now is using RonCo's modifiers, and I know he's acknowledged this as an issue. Still, look at this:

1975 Cincinnati Reds

Ross Grimsley
Age: 25
BB/K Ratio: 157/52 in 214 IP (at age 21, he was at 62/130)

Tom Carroll
Age: 22
BB/K Ratio: 111/54 in 202 IP (he sucked as a rookie the year before, too (30/12) - why was he even on the ML roster?)

Les Cain
Age: 27
BB/K Ratio: 89/55 in 131 IP (he was at 165/122 in 225 IP the year before, but at age 21, 66/118)

Don Gullett
Age: 24
BB/K Ratio: 95/15 in 116 IP (at age 19, he was at 38/75)

Pat Darcy
Age: 25
BB/K Ratio: 77/24 in 100 IP (year earlier, as a rookie, had a 37/36)

Wayne Simpson
Age: 26
BB/K Ratio: 46/18 in 68 IP (at 21, had a 131/184 then a 111/175 in 265-270 innings each year...then went downhill. Fast.)


Even if some of these drop-offs are caused by legit arm injuries, there's no way these numbers should be occurring like this (from a review of their drop-offs and injury history, I can say that Cain and Simpson's drop-offs were tied to injuries). The drop-off either wouldn't be this severe, or they'd never see a AA mound, much less a ML mound, if they had these kinds of ratios.

And these guys aren't old, either. This is the equivalent of having Dontrelle Willis, Rick Ankiel, Kevin Saucier, Steve Blass, and Mark Wohlers on your pitching staff, all at the same time. And letting them pitch the entire year. Hard to believe this team ONLY lost 108 games!

[Ben E Lou]

Originally Posted by Alan T

I can see if the AI is programed to not want to drop a fan favorite due to a desire for avoiding a Fan interest hit, but I would assume that all of these Cy Young pitchers that then get released would also have been a fan favorite too. At least it doesn't seem in my experience so far that it takes long for a cy young type pitcher to become one of the fan favorites.

I'm not seeing Cy Young pitchers getting released though. I'm seeing them take massive ratings dumps due to injury or random dev, (The latter is much more rare in this version than in the past, though...) and then ending up in AAA the next season.

As for #2, yeah people do Usually though the whines have some grounding behind them though, or at least some reason for the whine to originate.. It may not be as huge an issue or catastrophic of a problem as they might be leading on, but usually there is at least some issue there.

I agree. I'm just commenting that some do, and some don't. The biggest issue with OOTP is that there are three completely different ways to play it with completely different needs, and sometimes there is conflict between the three. I hope to keep fictional play working very well and improving, and couldn't possibly care less how historical and MLB play work, so every stitch of feedback I give will be colored by that.

I'll be curious to see what you comment in 48 hours or whenever you come to some conclusion regarding the issue, it just seemed interesting to me that people have two seemingly opposite scenarios that are issues that I couldn't imagine both to exist.

The short version is this: I'm getting very good results with 49/49/1/1 evaluation. (I tend to wonder if 0s screw things up.) Based on the stats in the first decade of this latest test career, I suspect I'll be getting a new QS with RonCo dev and my updated stats modifiers out today or tomorrow. It's going very well so far.

[Ben E Lou]

Originally Posted by Ksyrup

Admittedly, the league I'm looking at right now is using RonCo's modifiers, and I know he's acknowledged this as an issue. Still, look at this:

1975 Cincinnati Reds

Ross Grimsley
Age: 25
BB/K Ratio: 157/52 in 214 IP (at age 21, he was at 62/130)

Tom Carroll
Age: 22
BB/K Ratio: 111/54 in 202 IP (he sucked as a rookie the year before, too (30/12) - why was he even on the ML roster?)

Les Cain
Age: 27
BB/K Ratio: 89/55 in 131 IP (he was at 165/122 in 225 IP the year before, but at age 21, 66/118)

Don Gullett
Age: 24
BB/K Ratio: 95/15 in 116 IP (at age 19, he was at 38/75)

Pat Darcy
Age: 25
BB/K Ratio: 77/24 in 100 IP (year earlier, as a rookie, had a 37/36)

Wayne Simpson
Age: 26
BB/K Ratio: 46/18 in 68 IP (at 21, had a 131/184 then a 111/175 in 265-270 innings each year...then went downhill. Fast.)


Even if some of these drop-offs are caused by legit arm injuries, there's no way these numbers should be occurring like this (from a review of their drop-offs and injury history, I can say that Cain and Simpson's drop-offs were tied to injuries). The drop-off either wouldn't be this severe, or they'd never see a AA mound, much less a ML mound, if they had these kinds of ratios.

And these guys aren't old, either. This is the equivalent of having Dontrelle Willis, Rick Ankiel, Kevin Saucier, Steve Blass, and Mark Wohlers on your pitching staff, all at the same time. And letting them pitch the entire year. Hard to believe this team ONLY lost 108 games!

What year did you start the career? Was it in the 70s? I ask because there's definitely a problem with RonCo's modifiers for the first 30 seasons or so. What I *think* is happening is that the league is generated with an OOTP-normal distribution, and then the combination of much-slower dev and much-faster decline is a double-whammy. There's a pretty big stats shift that happens in those first 30 years, too. (As in league ERA dropping from being over 5.00 for the first 5 years or so, to being under 4.50 every year from year 30 on.)

[Ksyrup]

I started the league in 1901.

This sucks - I just tried to pull up the first sim I did using the default modifiers to see if I could find similar issues, and I got an error saying it was either an old version or corrupt. I did nothing but sim 95 years, close it out, and start a new sim. Hmmm.

[Ben E Lou]

Originally Posted by Ksyrup

I started the league in 1901.

This sucks - I just tried to pull up the first sim I did using the default modifiers to see if I could find similar issues, and I got an error saying it was either an old version or corrupt. I did nothing but sim 95 years, close it out, and start a new sim. Hmmm.

It could also be a case of the AI going into "rebuild" mode? From what I've noticed, it heavily favors prospects in that case, so it's probably looking at future ratings of those guys in making those decisions.


You may also be dealing with the effects of your AI evalutations with RonCo's modifiers. As we both mentioned in that thread, changing the AI evals caused some radical changes to how this sort of thing was handled.

[Anthony]

Originally Posted by Ben E Lou

Two things.

1. That dichotomy is one of the reasons that I speculate that the players that hang on too late has to do with things other than ratings and stats, such as past performance and perhaps fan favorite-ness. The settings that I use to deal with this issue ignore past performance. (I mean, really, who cares what he did two years ago? I don't want the AI handicapped by using past performance. I sure don't use it.)

i have 2 years ago stats set to low, like 10 or 15%. i think i have current stats as the most, followed by last year's stats, then ratings, then 2 years ago.

i also had to take the trading difficulty off "very difficult", the AI simply wasn't interested in doing many trades. i offered, as a test, a 3 star current/5 star potential 20 year old OF for an over 30 year old 4 star reliever and it didn't want to do it. it might've been because i had "stats 2 year ago" a little too high, and perhaps since that OF was a rookie with not much stats to go by maybe it wasn't valuing him enough. i also raised ratings a little. on the second to highest setting the AI is acting more appropriately and is still providing a challenge.

i like how in my career the Rays have won the WS (which is totally possible this year in real life), after going to the WS last season. i'm a little pissed cuz i got all the way to the NLCS and the game crashed on me, i had to resim the start of the playoffs and i got knocked out of the divisional series. i'm not sure if i should have a more hands on role during the playoffs or just let the AI sim the game, i don't know if i give my team too much of an advantage or if i'm a detriment.

[Ben E Lou]

Originally Posted by Hell Atlantic

i have 2 years ago stats set to low, like 10 or 15%.

I'm going to be recommending a big, fat "1" for 2 years ago, fwiw.

[Anthony]

i got no probs with that. like you - i don't even look at stats from 2 years ago when looking for players to aquire. i'm concerned with ratings, current year stats to see if those ratings are accurate and maybe taking a quick glance to see if the current stats are an anomoly or if he was able to do roughly the same thing the previous year. all i care about is what the guy can currently do and what he can be expected to do in the future. the past is the past. i can't win championships based on what a guy did a few years back.

[molson]

How do the AI evaluation settings work in the beginning or middle of a season?

If "current year stats" is 50% or whatever, how the game utilize limited early season-stats? (If someone hits 4 home runs on opening day, but isn't usually a power hitter, is the AI going to put him in the cleanup position the next game?)

In other words, do people change the AI evaluation settings throughout the season to reflect the fact that the current season stats become more and more reliable?

[Ben E Lou]

It has a calculation based on what part of the season you're in.

[Alan T]

From the OOTP boards to help people complaining about the MLB quick start:

Originally Posted by Markus Heinsohn

Unfortunately there were some bad settings in the roster set that was released with the first patch. We have fixed these and the new ML quickstart is now available. To install, you have two choices:

Through the Add-Ons Center
1) Open OOTP, then continue the current game
2) Go to the Game menu, and select "Add-Ons Center"
3) Click "Reload Add-Ons List"
4) A new one called "Updated major league roster set" appears, click on "Download", and after the download on "Install"

Manually
1) Download the following zip file: www.ootpdevfiles.com/ootp/rosters/majors.zip
2) Unzip the file to the major league quickstart folder, for example "My Documents\Out of the Park Developments\OOTP Baseball 9\quickstart_games\Majors.quick"

That's it, now when you start a new 2008 ML game the new rosters are being used.

Enjoy!
Markus

:flowers:

[DaddyTorgo]

that's cool. interested to see how SD's 49/49/1/1 type thing plays out. May try that tonight.

SD - did you also say that you'd have some back-and-forth with RonCo regarding player development modifiers too you thought?

I'd like to see a bit less of the phenomenon of "massive ratings hit for no reason" - things like a pitcher going from 26-3 with an era in the 2's one year to 14-17 with an era in the high 4's/5's the next year. Especially when the guy is 30 with no injury history.

Is there a setting that I just missed somewhere (or editing of the config files) to reduce the random ratings drops? Basically I want a league with fewer "one season wonders."

[Ksyrup]

Originally Posted by DaddyTorgo

I'd like to see a bit less of the phenomenon of "massive ratings hit for no reason" - things like a pitcher going from 26-3 with an era in the 2's one year to 14-17 with an era in the high 4's/5's the next year. Especially when the guy is 30 with no injury history.

Is there a setting that I just missed somewhere (or editing of the config files) to reduce the random ratings drops? Basically I want a league with fewer "one season wonders."

Yep, I'm seeing this a lot, too. If it's tied to injuries, OK, but many of these aren't.

[BigPapi]

Originally Posted by Ben E Lou

It tends to happen in fictional leagues late in careers, and it happens much more often with RonCo's dev modifiers. It's actually more of a problem of AI teams leaving old guys in there for too many seasons. I've pretty much eliminated it by combining some things I learned from Ron with some adjustments to settings. I hope to be able to post the findings in the next 48 hours. Essentially what's happening is that a guy will have a solid career, but then finish it out with a couple of seasons that look like 5-15, 6.24, 100 BB, 40K. I suspect it's over-favoring fan favorites and/or big-salaried guys when setting the rotation. But like I said, I've been able to pretty much do away with it now.

I posted this over at the main board- but aren't some of these results a logical results of the exaggerated settings RonCo is using?

It makes sense to me that if you have modifiers set to accelerate the deterioration process of your older players- but you have purposely delayed the growth of your prospects, the AI will have no choice but to leave some of these older, washed up pitchers at the MLB level for lack of a better alternative. I happen to think his modifiers are needlessly extreme. I also think that if the modifers are changed, they need to be changed in tandem, e.g., if you slow the development cycle, you need to likewise slow the aging cycle- or vice-versa. I am seeing very good results with both aging and development modifiers at 0.8. This is with a modified version of the included MLB league (I tend to agree with those that think these are a mess at anything more than a cursory glance) with a 128 team, 12 conference College feeder and 48 team HS feeder leagues. But these settings hold up as well long after the MLB players have vacated 50 years into the sim. For example, in the year 2068 I don't have a rash of 40+ year old players dominating the league- but I do have 11 guys with starting jobs between the ages of 37 and 41. Development peaks at age 26-27 rather than 27-29- but I can live with that- and I still have the occasional 21 or 22 year old who is good enough to qualify for a starting gig at the MLB level.

Perhaps part of this difference is because I have a 30 team league- but I am not really seeing some of the same things as RonCo that would neccesitate the extreme aging/development modifiers.

[Ksyrup]

OK, so I found my first sim, and the one that was corrupt wasn't it (incidentally, I think there's a problem with the game not erasing deleted saves from the list of games that can be loaded - I deleted this one, but it's still there and only goes away when I physically go to the files and delete the entire file). This one was simmed with default modifiers and evaluation settings of 0/50/35/15 (zero for ratings).

I found one of the most egregious examples of this EVER. I give you...Ed Halicki:

1974: 12-5, 2.51 ERA, 143.1 IP, 40 BB, 95 K
Age 23

1977: 2-17, 9.62 ERA, 102 IP, 109 BB, 2 K
Age 26

And the funny thing is, he was released by the Giants on June 13th and picked up by the Phillies on June 15th - and he was awful for both, although worse for the Phillies.

Now, this one looks like it might be injury-related (since he was imported in February of 1974, I can't see his original ratings, but he had 2 injuries in 1974 that sidelined him for 10 weeks and his ratings likely dropped on 1/1/75 - although his ratings REALLY dropped over the next couple of years without any significant injuries), but still...how can this guy be given 22 starts by 2 teams?

[Ben E Lou]

BigPapi:

Are you not seeing the peak years at age 23/24? I'm definitely seeing that with the standard settings. Ron's are much better for causing peaks to happen in the 26-29 range. Yet even with the 200 aging, I still see some guys putting up solid numbers after age 40.

[Ksyrup]

Originally Posted by BigPapi

I posted this over at the main board- but aren't some of these results a logical results of the exaggerated settings RonCo is using?

It makes sense to me that if you having modifiers set to accelerate the deterioration process of your older players- but you have purposely delayed the growth of your prospects, the AI will have no choice but to leave some of these older, washed up pitchers at the MLB level for lack of a better alternative. I happen to think his modifiers are needlessly extreme. I also think that if the midifers are changed, they need to be changed in tandem, e.g., if you slow the development cycly, you need to likewise slow the aging cycle- or vice-versa. I am seeing very good results with both aging and development modifiers at 0.8. This is with a modified version of the included MLB league (I tend to agree with those that think these are a mess at anything more involved thatn a cursory glance) with a 128 team, 12 conference College feeder and 48 tema HS feeder leagues. But these settings hold up as well long after the MLB players have vacated 50 years into the sim. For example, in the year 2068 I don't have a rash of 40+ year old players dominating the league- but I do have 11 guys with starting jobs between the ages of 37 and 41.

Perhaps part of this difference is because I have a 30 team league- but I am not really seeing some of the same things as RonCo that would neccesitate the extreme aging/development modifiers.

But how do you explain my 1975 Reds example, where the oldest guy putting up these ridiculous numbers is 27?

[Ben E Lou]

Originally Posted by Ksyrup

This one was simmed with default modifiers and evaluation settings of 0/50/35/15 (zero for ratings).

I have some real concerns regarding using a "0." I suspect that it screws up some of the calcs. That's why I'm using the 49-49-1-1 to ignore previous stats as much as possible, rather than 50-50-0-0.

Also, for your guy, how did he do the two seasons before the horrid one? I ask because I believe that the "stats" evaluations are scaled as to where you are in the season. It may be that for the first couple of months, it's ignoring his horrible ratings, and it's also largely ignoring his 9+ ERA as well. At what point in the season did he sit?

[BigPapi]

Originally Posted by Ben E Lou

BigPapi:

Are you not seeing the peak years at age 23/24? I'm definitely seeing that with the standard settings. Ron's are much better for causing peaks to happen in the 26-29 range. Yet even with the 200 aging, I still see some guys putting up solid numbers after age 40.

Well , I am using .8 for both pitcher and hitter development as I said- so I suspect my peak ages will be better, but not quite as good as with the 0.5 that RonCo was using. I see a few players peaking a bit early- but looking at my top 100 players in the league 80% or better are between 26-28. On the other hand- I didn't like seeing several players with peak ages of 29 and 30 with the setting at 0.5 either- so I suppose this is give and take as well as personal preference.

You say you are seeing "some guys" putting up solid numbers after age 40- but how many? Five of Fifty? Pitchers or hitters- or both? I am comfortable with a Glvaine, or Randy Johnson or Barry Bonds here or there- but I would agree that if it's more than a half dozen total it's getting out of hand. I am just not seeing this- though I suspect that the average age of my 4 star and up players is a little higher than normal....But that's ok too, as RonCo pointed out- OOTP is a little low for players average career length anyway.

[BigPapi]

Originally Posted by Ksyrup

But how do you explain my 1975 Reds example, where the oldest guy putting up these ridiculous numbers is 27?

I agree- those numbers are nuts...That would depress me to see that many examples in the whole league- much less the same team...That's got to be a 120 loss team there (not 108)...

I don't really play fictional- so I am not sure what's going on.....This is with Ron's settings? If so, did you see this with default modifiers? I have never seen anything like that....In fact, I do sometimes see the precipitous drop in stuff from ages 37-39- but guys don't seem to be able to hold to their jobs after this for the most part. I also see some guys 37 and 38 averaging 6-7 K's per nine as well. On the whole- I think this a problem- but I am not seeing it to the degree that you are.

[Ksyrup]

Originally Posted by BigPapi

I agree- those numbers are nuts...That would depress me to see that many examples in the whole league- much less the same team...That's got to be a 120 loss team there.

I don't really play fictional- so I am not sure what's going on.....This is with Ron's settings? If so, did you see this with default modifiers? I have never seen anything like that....In fact, I do sometimes see the precipitous drop in stuff from ages 37-39- but guys don't seem to be able to hold to their jobs after this for the most part. I also see some guys 37 and 38 averaging 6-7 K's per nine as well. On the whole- I think this a problem- but I am not seeing it to the degree that you are.

The Ed Halicki example is from an historical sim using the default modifiers. The Reds examples were from the second sim, using RonCo's modifiers.

Maybe part of the problem has to do with historical sims? I'm seeing a number of guys do just what Halicki did - come in immediately and dominate during their rookie season, and within 5 years, they suck.

[Ksyrup]

Originally Posted by Ben E Lou

I have some real concerns regarding using a "0." I suspect that it screws up some of the calcs. That's why I'm using the 49-49-1-1 to ignore previous stats as much as possible, rather than 50-50-0-0.

Also, for your guy, how did he do the two seasons before the horrid one? I ask because I believe that the "stats" evaluations are scaled as to where you are in the season. It may be that for the first couple of months, it's ignoring his horrible ratings, and it's also largely ignoring his 9+ ERA as well. At what point in the season did he sit?

Here are Halicki's entire ML stats:

1974: 12-5, 2.51 ERA, 143.1 IP, 119 HA, 40 BB, 95 K
1975: 14-12, 4.18 ERA, 217.2 IP, 222 HA, 133 BB, 108 K
1976: 4-7, 7.31 ERA, 88.2 IP, 114 HA, 58 BB, 11 K
1977: 2-17, 9.62 ERA, 102 IP, 109 BB, 2 K

In 1976, he only threw 17 IP in the minors, so I think he was on the ML roster for most of that year, but just not used as much. It seems clear that his 1974 injuries are to blame for his decline, though.

And yeah, I understand that I'm arguing extremes here - on the one hand, this guy should have been dropped from any ML roster at some point in 1976, at least until he proved in the minors that he belongs, and on the other hand, I've pointed out guys like Britt Burns who are getting prematurally (IMO) yanked from an ML roster before they've proven they suck. But I think both of these are issues.

[Cork]

Markus has posted that a couple of new features will be included in patch #2. One is adjustable scouting system accuracy on a scale from 1 to 5. The other will be some sort of player development report.

Good stuff.

-Cork

[Ksyrup]

OK, I'm going to create and sim about 75-90 years of a 2008 fictional league (2 leagues, 2 divisions, 24 teams total) with the default modifiers and evaluation settings of 1/50/48/1 and see what happens.

[DaddyTorgo]

Originally Posted by Ksyrup

Yep, I'm seeing this a lot, too. If it's tied to injuries, OK, but many of these aren't.

right. in this example he was totally uninjured. i understand that at some point his ratings have to drop, but i just don't like seeing my #1 starter one year become my #5 starter the next year. It makes no flipping sense, it's entirely unrealistic, and really handicaps you when you've given the guy a $20mil contract on the basis of the fact that the game ought to be able to correctly "age" him and he should be able to earn that contract.

and he's just one example...it ends up happening to pretty much every pitcher i've ever had (and some batters too).

I just want to see if a guy gets to his "peak" that he stays there for 1-2 big contracts. too many guys have a career year, sign a huge contract and then blow chunks.

[RainMaker]

Originally Posted by Ksyrup

OK, so I found my first sim, and the one that was corrupt wasn't it (incidentally, I think there's a problem with the game not erasing deleted saves from the list of games that can be loaded - I deleted this one, but it's still there and only goes away when I physically go to the files and delete the entire file). This one was simmed with default modifiers and evaluation settings of 0/50/35/15 (zero for ratings).

I found one of the most egregious examples of this EVER. I give you...Ed Halicki:

1974: 12-5, 2.51 ERA, 143.1 IP, 40 BB, 95 K
Age 23

1977: 2-17, 9.62 ERA, 102 IP, 109 BB, 2 K
Age 26

And the funny thing is, he was released by the Giants on June 13th and picked up by the Phillies on June 15th - and he was awful for both, although worse for the Phillies.

Now, this one looks like it might be injury-related (since he was imported in February of 1974, I can't see his original ratings, but he had 2 injuries in 1974 that sidelined him for 10 weeks and his ratings likely dropped on 1/1/75 - although his ratings REALLY dropped over the next couple of years without any significant injuries), but still...how can this guy be given 22 starts by 2 teams?

Do we happen to know how the "Current Year Stats Weight" works? I mean, when does the computer take it into effect? If my backup C with low ratings has a hot first two weeks to the season, is he given the starting job? Or does the game wait a month or two before compiling the "Current Year Stats"?

I just wonder if the computer doesn't take into account the current year stats till maybe the mid-point in the season. So lets say the AI decides to use the current year stats on June 15th. At which point he is released. While released, there is something missing in the code that doesn't factor in his current year stats while he's a free agent (or only counts current year stats for the team he is with). So the other computer teams aren't basing their decision off this year's stats, and pick him up and start him till he builds up enough bad stats with his current team to warrant not starting any longer.

Hope that makes sense. I just wonder if this is an issue with the current year stats evaluator and whether stats are factoring in for other teams they were on that year, and at what point the AI uses those values.

[CraigSca]

I would hope it's weighted by atbats and batters-faced, too. But you never know.

[Ksyrup]

Originally Posted by RainMaker

Do we happen to know how the "Current Year Stats Weight" works? I mean, when does the computer take it into effect? If my backup C with low ratings has a hot first two weeks to the season, is he given the starting job? Or does the game wait a month or two before compiling the "Current Year Stats"?

I just wonder if the computer doesn't take into account the current year stats till maybe the mid-point in the season. So lets say the AI decides to use the current year stats on June 15th. At which point he is released. While released, there is something missing in the code that doesn't factor in his current year stats while he's a free agent (or only counts current year stats for the team he is with). So the other computer teams aren't basing their decision off this year's stats, and pick him up and start him till he builds up enough bad stats with his current team to warrant not starting any longer.

Hope that makes sense. I just wonder if this is an issue with the current year stats evaluator and whether stats are factoring in for other teams they were on that year, and at what point the AI uses those values.

In one of the pre-release threads, I raised this issue about what I'd seen in OOTP7 and Marcus said he found a nasty bug in how the evaluation settings worked and that he'd fixed it. For in-season ratings, there was some sort of calculation done. I'm not sure of the specifics, but Marcus said it was "fixed."

[lighthousekeeper]

Originally Posted by RainMaker

Do we happen to know how the "Current Year Stats Weight" works? I mean, when does the computer take it into effect? If my backup C with low ratings has a hot first two weeks to the season, is he given the starting job? Or does the game wait a month or two before compiling the "Current Year Stats"?

I just wonder if the computer doesn't take into account the current year stats till maybe the mid-point in the season. So lets say the AI decides to use the current year stats on June 15th. At which point he is released. While released, there is something missing in the code that doesn't factor in his current year stats while he's a free agent (or only counts current year stats for the team he is with). So the other computer teams aren't basing their decision off this year's stats, and pick him up and start him till he builds up enough bad stats with his current team to warrant not starting any longer.

Hope that makes sense. I just wonder if this is an issue with the current year stats evaluator and whether stats are factoring in for other teams they were on that year, and at what point the AI uses those values.

it is continuously weighted based on the number of at bats (or perhaps plate appearances) in the current season. so if you have it set to 100% on Current Year stats, and your catcher has 100 PA this season, it uses something like:
Current Year stats weighting: 100/550*100% ~= 18%
Last Year stats weighting: 82%

This method avoids the problem you described because at any point in the season, the game is theoretically goin back to use the last 550 PA's as a substitute for "Current Year Stats".

[lighthousekeeper]

...and if the previous season doesn't have enough AB's to reach 550, then it goes back to the year before. and then if there's still not enough at bats, then i think the remainder defaults to ratings.

[RainMaker]

Originally Posted by lighthousekeeper

it is continuously weighted based on the number of at bats (or perhaps plate appearances) in the current season. so if you have it set to 100% on Current Year stats, and your catcher has 100 PA this season, it uses something like:
Current Year stats weighting: 100/550*100% ~= 18%
Last Year stats weighting: 82%

This method avoids the problem you described because at any point in the season, the game is theoretically goin back to use the last 550 PA's as a substitute for "Current Year Stats".

Would his 35% setting for last season still count toward last season?

In effect, his pitcher would only have this season weighted 40% based on the day he was released. Thus making at June 15, his current season being only 20% of what the computer uses to make the decision (40% of 50%). It would seem that the AI's logic isn't flawed if he had a good season the previous year (which would account for 65% of the evaluation on June 15th). A solution for this would seem to really put the focus heavily on Current Year Stats since it will also use last years up to a point. I wonder if the evaluator was set 0/80/10/10, if he would have been released much earlier and not picked up by the computer.