Pitcher's Dropping Off Despite Doing Well

Frank Delarocca Jr · 03-31-2007, 03:28 AM

As Battists said to, I'll go ahead and post this in Tech Support.

From what I've seen, pitchers who are doing well and are still decently rated seem to find themself suddenly sent to the pen or released in my historical replays, without injury and without major drops in ratings from season to season.

With recalculate on, these pitchers can win 20 game and post a great ERA, but the next year they move to the bullpen without a single start or get released outright before deciding to retire at age 28, 29, 30, or 31 without ever getting another shot at pitching again, not even in the minors most times.

With recalc off, it effects both hitters and pitchers. Without recalc, guys will be having fine careers, hitting .300/30 hr/800+ ops, whatever, to getting 49 atbats and then released, sort of what happens to pitchers on recalc. I suppose, since recalc is off, it's more of a crapshoot on how guys play out, but with it on, I can't see how pitchers should disappear like they do.

It'd be really great if current performance had more of a hand in how the AI moves players (especially pitchers) around the depth charts. I don't know how easy that would be to adjust, but in it's current state it's hard to watch. It's just not realistic in the way it makes roster adjustments in these cases and gives solo game players a slight edge as they can sign these guys without a problems since the other AI teams never sign them.

I don't think this is such a problem for league play as people wouldn't just toss a pitcher due to a slight loss of ratings, but for the solo player, it's just not a good thing.

-Frank D.

f.montoya · 03-31-2007, 07:34 AM

Frank, while I agree that this should not be occurring with default settings but please try going to the "AI Player Evaluation Options" in the game setup and reduce the weight given to "ratings" and redistribute the percentages among the recent years' stats instead. I think you will have much better results. If you could follow up here after trying the above, at the very least, we could get Markus to adjust the defaults.

RonCo · 03-31-2007, 08:43 AM

I agree with f.montoya's comments. I also wanted to get a grasp on the scope of the problem, so I went and ran a very quick 8-Team league for 80 seasons (three levels of minors, everything essentially default ...scouts/coaches off). I captured the retirment ages of every pitcher who appeared in the majors, an plotted those who left after the age of 30.

At the end of the day I think we're seeing a fairly good curve on the whole. There are a few more 30 and 31s than I would like, but it's not bad. I agree it's a little disconcerting to see some players moving right along and then falling off and not being played. Such ratings fall off does happen in real life, though--especially with pitchers. Usually some team will play them a little longer in real life, though. So adjusting the AI to be more stats driven and less ratings driven would probably help.

That said, you will be ratings-driven. So making that adjustment will probably give you better overall careers, but it will also give you a possible advantage in solo play. As far as online leagues not being influenced that way, I would partially disagree. The better owners in my online league are highly ratings-driven. as soon as they understand the development engine in v2007, they will begin to steer away from early-mid 30 year-olds whose ratings are falling--or at least expect to get them at a discount.

knuckler · 03-31-2007, 04:32 PM

I might be wrong, but Frank isn't really seeing much in the way of ratings fall off. The issue as I understand it is that the players aren't falling off much if at all(ratings or stat wise), and yet the AI is not using them.

In solo play, this is an advantage for controlling a team because it's easy to pick them up, whereas if the AI was using them, it wouldn't be near as easy to pick them up.

In an online league, this wouldn't even really be an issue because owners are going to continue to use guys that are effective and don't fall off.

RonCo · 03-31-2007, 05:37 PM

Quote:

Originally Posted by knuckler

I might be wrong, but Frank isn't really seeing much in the way of ratings fall off. The issue as I understand it is that the players aren't falling off much if at all(ratings or stat wise), and yet the AI is not using them.

In solo play, this is an advantage for controlling a team because it's easy to pick them up, whereas if the AI was using them, it wouldn't be near as easy to pick them up.

In an online league, this wouldn't even really be an issue because owners are going to continue to use guys that are effective and don't fall off.

Probably true. But from the chart above, there's not _that many_ in relation to the whole of the league. I'm guessing that the players he's seeing are hitting a magic number in the ratings category, and the AI is making a decision based on that. I could be wrong, though. That 30-33 yo period is where you can see some pitchers begin to lose both stuff and movement. Going way, way out on a limb here, I predict that those pitchers who are most subject to this behavior are guys who have only moderate control to begin with, hence an erosion of stuff and/or movement hurts the guy's calculated value enough that his mediocer control isn't enough to support relying on him.

That last is a semi-educated guess, anyway.

Dargone · 03-31-2007, 08:59 PM

I noticed this with Noodles Hahn in my historical replay. He won the pitcher of the year award in 1906 and was released in 1907? Seems like there might be something up here? Although I do have AI evaluation set to 40% for ratings but his rating didn't really tank either.

Dargone · 04-01-2007, 03:08 AM

Cy Young just won the pitcher of the year and was released the following year. His ratings are still very good? Seems like something might be going on here...

Dargone · 04-01-2007, 03:22 AM

Well, Young was just signed by Brooklyn so maybe Boston just doesn't want him anymore or can't afford him? However, I figured with free agency turned off that shouldn't be an issue? I'll keep watching out for these types of things as I sim the replay...

RonCo · 04-01-2007, 09:36 AM

I'm not suggesting no issue exists. My data is showing only the system's overall performance.

Frank Delarocca Jr · 04-02-2007, 04:11 AM

Okay, I went ahead and did what F. Montoya said to do. I changed the AI evaluation from the default and lowered the amount weighte for 'ratings' and redistributed it to the others so they were basically even. Rates were now at 10, rest at 30. Recalculate was left on.

I started a new league in 1911. By 1915, I've seen the following:

George Mullin - 1911 - 12-13, 4.04 ERA, 75k/86bb; 1912 - 5-3, 3.10 31 k/28 bb; team just stops using him, got a contract extension, age 32. 1913 - released, doesn't play anywhere, retires the following offseason. His stuff dropped from 5 to a 3 and his control went from 11 to 9 to 10 while movement stayed at 20.

Barney Pelty - 1912 - 16-18, 3.76, 83/137; 1913 - 1 game. signs extensions, age 32. Released 1914 offseason, never pitches again. Stuff drops from 4 to 2, control goes up to 11, movement stays at 19

Joe Lake - 1911 - 21-15, 2.45, 142/103; 1912 - 21-14, 3.57, 132/87; 1913 - 21-13, 3.25, 125/86; 1914 - released, age 32. Stuff goes from 2 to 1 to 2 again, control stays at 15, movement moves from 19 to 18.

Ed Summers - 1911 - 12-21, 4.07, 96/105; 1912 - 17-19, 3.69, 86/104; 1913 - released, sits and retires, age 28. stuff goes from 5 to 3, control 11 to 8, movement is unchanged

Jack Combs - 1911 - 5-2, 1.29, 14 games, 4 starts; 1912 - 4-1, 1.93, 12 games, 3 starts; 1913 - 18-12, 3.78, 118k/112bb - 1914 - released; retires age 31. Stuff goes from 5 to 2 to 5, control 10 to 7 to 10, movement is unchanged

Russ Ford - 1911 - 22-17, 3.67, 178/91; 1912 - 15-19, 3.49, 95/112; 1913 - 2-0, 1.15, 9 games, none started, sent to AAA - 1914, released, retires age 30. Stuff goes from 7 to 6, control 16/14/16/14, movement 19/16/19/16

Dolly Gray - 1911 - 27-8, 2.93, 100/117; 1912 - 2-1, 3.52, 8 games, 2 starts, ends year at AAA; 1913 - released, retires age 34 in 1914. Stuff goes from 7 to 4, control 12/9/13, movement 18/16

Vean Gregg - 1911 - 7-5, 3.03, 57/45; 1912 - 19-13, 3.03, 173/111; 1913 - 18-16, 3.23, 170/146; 1914 - released, age 28, retires age 29. Stuff 13 to a 10, control - 11 to 10, movement stays at 20.

there are another 5-7 that I find as odd as these guys, but it's late and I think I've got enough here for an example of what I'm seeing. The game doesn't bother with performance at all. Ratings do drop a bit in some cases, a bit more than a bit in others, but I don't think they're so drastic that they should be released and retired at such an early age. I do notice that, with the change in evaluation weights that less 28-30 years olds are dropping off the face of the earth, which is good, but I'd just like to see it tweaked so that the game wouldn't kill these guys quite as often as it does. Maybe a tweak on how player's stuff works could help as almost all of these guys stuff ratings drop off before they get discarded, even if they jump back up during their year off in the FA pool.

I don't know what how the recalculate function is affecting this compared to it being off, but I have a feeling these guys might not bounce around as much with it off; then there'd be guys who just dropped off randomly like the old game used to do.

redmarkYankees · 04-02-2007, 08:08 AM

I remember discussing something very similar to this in a thread about OOTP2006, before I gave up on that game, so I'm not sure whether the discussion progressed any further.

My tests with 2006 suggested that, as RonCo says, a minor drop (even a point or two) in one of the 3 pitching categories pushes the player over a 'threshold' in calculated value, at which point he is dropped, released etc. The problem for historical replays, as I said last year, is that most pitchers have virtually no 'stuff' to begin with. They therefore seem very susceptible to a small drop in either movement or control; but to the point where that drop will see them penalised (sometimes even over another pitcher with worse ratings but who hasn't seen them fall).

I haven't looked into this in OOTP2007 as I've been looking at other things, but at a high level it had seemed improved, with less inexplicable pitcher retirements.

redmarkYankees · 04-02-2007, 08:11 AM

dola

I'm currently running a 1901 > sim with annual csv dumps, so I'll try and get some data on this.

knuckler · 04-02-2007, 10:09 AM

Quote:

Originally Posted by RonCo

Probably true. But from the chart above, there's not _that many_ in relation to the whole of the league.

It's not just retirement though from what I understand Frank is saying. Good starters are being sent to the bullpen without ever being given the chance to start the following year, and in some cases being released.

RonCo · 04-02-2007, 10:31 AM

Quote:

Originally Posted by redmarkYankees

The problem for historical replays, as I said last year, is that most pitchers have virtually no 'stuff' to begin with.

This is a "tuning" issue that we've discussed in beta for some time. The new development model in v2007 will peak a pitcher's stuff early (22-24 yo as a rule), and it will fall from there. His control will generally rise throughout his career. The Lahman import process still needs to be improved. Recalc makes this not quite as much of an issue, as that feature recalculated the pitcher's ratings every year based on Lahman stats...so development/aging isn't directly through the development engine.

RonCo · 04-02-2007, 10:36 AM

Quote:

Originally Posted by knuckler

It's not just retirement though from what I understand Frank is saying. Good starters are being sent to the bullpen without ever being given the chance to start the following year, and in some cases being released.

I understand.

The issue that concerns me most in this thread is the apparent limits of pure stats assessment by the AI when that option is selected. The AI decisions (to move a pitcher to the pen or release them when the ratings dip) are defensible, despite being a bit aggravating.

knuckler · 04-02-2007, 04:49 PM

Quote:

Originally Posted by RonCo

I understand.

The issue that concerns me most in this thread is the apparent limits of pure stats assessment by the AI when that option is selected. The AI decisions (to move a pitcher to the pen or release them when the ratings dip) are defensible, despite being a bit aggravating.

My concern is that you're saying it doesn't affect many pitchers based on your chart when the issue isn't limited to retired players.

I think we need to hear more from Frank about what is happening to his players. I don't think all of them are ratings dips, and if he has recalculate ratings on, it would be beneficial to see what ratings changes are occurring to the players that are getting demoted/released. I agree though that if it's still happening after the assessment sliders have been adjusted that that is probably the biggest concern.

RonCo · 04-02-2007, 06:33 PM

Quote:

Originally Posted by knuckler

My concern is that you're saying it doesn't affect many pitchers based on your chart when the issue isn't limited to retired players.

Fair enough. I just note that of the seven examples listed, 100% were guys who stopped playing (released or retired). My chart doesn't differentiate "retired" from "released and no longer play".

Quote:

I think we need to hear more from Frank about what is happening to his players. I don't think all of them are ratings dips.

100% of his examples included ratings drops, but yes, I'm open to hearing more detail.

Quote:

, and if he has recalculate ratings on, it would be beneficial to see what ratings changes are occurring to the players that are getting demoted/released. I agree though that if it's still happening after the assessment sliders have been adjusted that that is probably the biggest concern.

He's already said he's using Recalc, and he's given us seven examples. He's said he's changed the assessment to go more toward stats, and it's improved things but not very much. It's this last that worries me most. If he went to hard-line stats as AI assessment, and guys with good stats are being released, then it's not "only" an issue with ratings sensitivity (which the data he's showing us goes toward), but is an issue with the code that is driving the AI assessment on stats.

I'm not trying to downplay the existence of an issue. I'm an independent tester with no ties to SI. There is obviously at least one issue here ... there could be as many as three:

1) The ratings dip (mostly stuff) hits a level and drops a pitcher either to the pen or in the waste bucket.
2) (related to or an aggravator of #1) The Lahman import is not tuned quite right to the development curves...this is a known issue to me that has been discussed in beta. I believe, but do not know, that Markus will be improving this. Bottom line: Stuff needs to import higher in this version than previous because it is tuned to fall after a pitcher's early 20s.
3) The "stats heavy" AI assessment appears to ignor stats and allow small ratings changes to still provoke an AI action.

My data above show the magnitude of #1 and #2 from a pure release/retire standpoint, and that magnitude is discernable but not massive.

Of all these, I worry the most about #3 because it impacts 100% of the guys who want to play a stats-only game, regardless of whether it is historical, fictional, or some hybrid undiscovered. And for those people the magnitude is large.

Ksyrup · 04-02-2007, 06:39 PM

I've seen the same issue multiple times in my sims. I'm not sure if it's a purely historical sim issue or not, but that's the only sim type I've played so far. Most of the time, rather than outright release, I've seen these guys put up excellent numbers, only to be sent to the minors the next year. I've even seen them get traded, and the new team puts them in the minors. So you'll have a guy win 20 games and not only get traded, but demoted as well.

This issue underscores the problem with the AI's knowledge of ratings and willingness to make moves solely based on ratings and not waiting to see a corresponding drop in performance on the field. IRL, if you have a guy who was a solid SP one year, you wouldn't automatically drop him from your rotation the next year even if you thought he had "lost it" and/or he underperformed during spring training - you'd at least give him a few starts to verify that what you observed in talent drop is manifesting itself on the field. It just isn't realistic to make a preemptive move like that.

I'm not sure what can be done to fix this, but the whole ratings vs. performance evaluation system seems like it isn't functioning as it should.

Frank Delarocca Jr · 04-02-2007, 07:52 PM

I simmed through another year and another few guys just seem to get shuffled off to the FA pile and retire after no one shows interest; I'd post more examples, but I'd just be basically repeating what I already posted.

All of the guys have the same things in common:

Stuff ratings drops a bit before they get released. Sometimes control does too, movement doesn't vary much. I'd say stuff might be the main factor in why they're no longer looked at as viable at this point (educated guess).

A good percentage of the pitchers see their ratings actually move back up a point or two during their season of exile, but that doesn't seem to entice any teams into signing them up.

I haven't been seeing as many guys getting banished to the bullpen in this run, they just get released instead. A few get sent to AAA, but most are outright releases come January-March.

With the weight on ratings/talents being so low, I would hope the stats the players put up would be more what the AI was looking at in evaluation. From what I'm seeing, though, is that it still looks at ratings/talents no matter. I'd say that most, if not all, these guys are borderline guys, and so far in this run, no 'stars' have dropped like flies, which is a good thing, but then, I'm using recalc again, unlike my last run that didn't use it and Walter Johnson turned into silly putty.

I think Ksyrup makes a great point, too. The AI jumps the gun on guys that drop off while, in real life, they'd still be getting a shot until they proved they weren't capable anymore. Then they'd go to the pen/minors and/or released into oblivion. If the game were able to be tweaked so it could emulate that flow, it'd really be damn near perfect.

Of course, that's easier said than done

-Frank D

Frank Delarocca Jr · 04-02-2007, 09:08 PM

Also,

I tried unretiring these guys and bumping their rates up today. Of the 7, 1 was signed and the rest re-retired after sitting for the entire year. The one that signed, Joe Lake, went 20-17 with a 1.55 ERA (I might have bumped him a bit much as his strikeouts went up to over 200

).

-Frank D

03-31-2007, 08:59 PM	#6
Dargone Hall Of Famer Join Date: Feb 2002 Location: Interwebs Posts: 2,862	I noticed this with Noodles Hahn in my historical replay. He won the pitcher of the year award in 1906 and was released in 1907? Seems like there might be something up here? Although I do have AI evaluation set to 40% for ratings but his rating didn't really tank either. __________________ I was never one to patiently pick up broken fragments and glue them together again and tell myself that the mended whole was as good as new. What is broken is broken -- and I'd rather remember it as it was at its best than mend it and see the broken places as long as I lived.-Margaret Mitchell

04-01-2007, 03:08 AM	#7
Dargone Hall Of Famer Join Date: Feb 2002 Location: Interwebs Posts: 2,862	Cy Young just won the pitcher of the year and was released the following year. His ratings are still very good? Seems like something might be going on here... __________________ I was never one to patiently pick up broken fragments and glue them together again and tell myself that the mended whole was as good as new. What is broken is broken -- and I'd rather remember it as it was at its best than mend it and see the broken places as long as I lived.-Margaret Mitchell

04-01-2007, 03:22 AM	#8
Dargone Hall Of Famer Join Date: Feb 2002 Location: Interwebs Posts: 2,862	Well, Young was just signed by Brooklyn so maybe Boston just doesn't want him anymore or can't afford him? However, I figured with free agency turned off that shouldn't be an issue? I'll keep watching out for these types of things as I sim the replay... __________________ I was never one to patiently pick up broken fragments and glue them together again and tell myself that the mended whole was as good as new. What is broken is broken -- and I'd rather remember it as it was at its best than mend it and see the broken places as long as I lived.-Margaret Mitchell

04-02-2007, 08:08 AM	#11
redmarkYankees All Star Starter Join Date: Nov 2003 Posts: 1,283	I remember discussing something very similar to this in a thread about OOTP2006, before I gave up on that game, so I'm not sure whether the discussion progressed any further. My tests with 2006 suggested that, as RonCo says, a minor drop (even a point or two) in one of the 3 pitching categories pushes the player over a 'threshold' in calculated value, at which point he is dropped, released etc. The problem for historical replays, as I said last year, is that most pitchers have virtually no 'stuff' to begin with. They therefore seem very susceptible to a small drop in either movement or control; but to the point where that drop will see them penalised (sometimes even over another pitcher with worse ratings but who hasn't seen them fall). I haven't looked into this in OOTP2007 as I've been looking at other things, but at a high level it had seemed improved, with less inexplicable pitcher retirements. __________________ In times of universal deceit, telling the truth will be a revolutionary act. George Orwell

04-02-2007, 08:11 AM	#12
redmarkYankees All Star Starter Join Date: Nov 2003 Posts: 1,283	dola I'm currently running a 1901 > sim with annual csv dumps, so I'll try and get some data on this. __________________ In times of universal deceit, telling the truth will be a revolutionary act. George Orwell

03-31-2007, 03:28 AM	#1
Frank Delarocca Jr Bat Boy Join Date: Mar 2007 Posts: 13	Pitcher's Dropping Off Despite Doing Well As Battists said to, I'll go ahead and post this in Tech Support. From what I've seen, pitchers who are doing well and are still decently rated seem to find themself suddenly sent to the pen or released in my historical replays, without injury and without major drops in ratings from season to season. With recalculate on, these pitchers can win 20 game and post a great ERA, but the next year they move to the bullpen without a single start or get released outright before deciding to retire at age 28, 29, 30, or 31 without ever getting another shot at pitching again, not even in the minors most times. With recalc off, it effects both hitters and pitchers. Without recalc, guys will be having fine careers, hitting .300/30 hr/800+ ops, whatever, to getting 49 atbats and then released, sort of what happens to pitchers on recalc. I suppose, since recalc is off, it's more of a crapshoot on how guys play out, but with it on, I can't see how pitchers should disappear like they do. It'd be really great if current performance had more of a hand in how the AI moves players (especially pitchers) around the depth charts. I don't know how easy that would be to adjust, but in it's current state it's hard to watch. It's just not realistic in the way it makes roster adjustments in these cases and gives solo game players a slight edge as they can sign these guys without a problems since the other AI teams never sign them. I don't think this is such a problem for league play as people wouldn't just toss a pitcher due to a slight loss of ratings, but for the solo player, it's just not a good thing. -Frank D.

03-31-2007, 07:34 AM	#2
f.montoya Hall Of Famer Join Date: Nov 2004 Posts: 6,069	Frank, while I agree that this should not be occurring with default settings but please try going to the "AI Player Evaluation Options" in the game setup and reduce the weight given to "ratings" and redistribute the percentages among the recent years' stats instead. I think you will have much better results. If you could follow up here after trying the above, at the very least, we could get Markus to adjust the defaults. __________________ Fidel Montoya Asahi2 Baseball ex-Commissioner(Historical League Since 2004) www.allsimbaseball.com (OOTP web hosting - Customized sites for online leagues - Sign up, Connect OOTP and Play!) Share Your Mods - Free, unlimited and easy to upload to share your Mods instantly(free site registration required)

03-31-2007, 08:43 AM	#3
RonCo Hall Of Famer Join Date: Aug 2003 Posts: 10,363	I agree with f.montoya's comments. I also wanted to get a grasp on the scope of the problem, so I went and ran a very quick 8-Team league for 80 seasons (three levels of minors, everything essentially default ...scouts/coaches off). I captured the retirment ages of every pitcher who appeared in the majors, an plotted those who left after the age of 30. At the end of the day I think we're seeing a fairly good curve on the whole. There are a few more 30 and 31s than I would like, but it's not bad. I agree it's a little disconcerting to see some players moving right along and then falling off and not being played. Such ratings fall off does happen in real life, though--especially with pitchers. Usually some team will play them a little longer in real life, though. So adjusting the AI to be more stats driven and less ratings driven would probably help. That said, you will be ratings-driven. So making that adjustment will probably give you better overall careers, but it will also give you a possible advantage in solo play. As far as online leagues not being influenced that way, I would partially disagree. The better owners in my online league are highly ratings-driven. as soon as they understand the development engine in v2007, they will begin to steer away from early-mid 30 year-olds whose ratings are falling--or at least expect to get them at a discount. Attached Images Last edited by RonCo; 03-31-2007 at 08:45 AM.

03-31-2007, 04:32 PM	#4
knuckler Hall Of Famer Join Date: Apr 2003 Posts: 3,347	I might be wrong, but Frank isn't really seeing much in the way of ratings fall off. The issue as I understand it is that the players aren't falling off much if at all(ratings or stat wise), and yet the AI is not using them. In solo play, this is an advantage for controlling a team because it's easy to pick them up, whereas if the AI was using them, it wouldn't be near as easy to pick them up. In an online league, this wouldn't even really be an issue because owners are going to continue to use guys that are effective and don't fall off.

04-01-2007, 09:36 AM	#9
RonCo Hall Of Famer Join Date: Aug 2003 Posts: 10,363	I'm not suggesting no issue exists. My data is showing only the system's overall performance.

04-02-2007, 04:11 AM	#10
Frank Delarocca Jr Bat Boy Join Date: Mar 2007 Posts: 13	Okay, I went ahead and did what F. Montoya said to do. I changed the AI evaluation from the default and lowered the amount weighte for 'ratings' and redistributed it to the others so they were basically even. Rates were now at 10, rest at 30. Recalculate was left on. I started a new league in 1911. By 1915, I've seen the following: George Mullin - 1911 - 12-13, 4.04 ERA, 75k/86bb; 1912 - 5-3, 3.10 31 k/28 bb; team just stops using him, got a contract extension, age 32. 1913 - released, doesn't play anywhere, retires the following offseason. His stuff dropped from 5 to a 3 and his control went from 11 to 9 to 10 while movement stayed at 20. Barney Pelty - 1912 - 16-18, 3.76, 83/137; 1913 - 1 game. signs extensions, age 32. Released 1914 offseason, never pitches again. Stuff drops from 4 to 2, control goes up to 11, movement stays at 19 Joe Lake - 1911 - 21-15, 2.45, 142/103; 1912 - 21-14, 3.57, 132/87; 1913 - 21-13, 3.25, 125/86; 1914 - released, age 32. Stuff goes from 2 to 1 to 2 again, control stays at 15, movement moves from 19 to 18. Ed Summers - 1911 - 12-21, 4.07, 96/105; 1912 - 17-19, 3.69, 86/104; 1913 - released, sits and retires, age 28. stuff goes from 5 to 3, control 11 to 8, movement is unchanged Jack Combs - 1911 - 5-2, 1.29, 14 games, 4 starts; 1912 - 4-1, 1.93, 12 games, 3 starts; 1913 - 18-12, 3.78, 118k/112bb - 1914 - released; retires age 31. Stuff goes from 5 to 2 to 5, control 10 to 7 to 10, movement is unchanged Russ Ford - 1911 - 22-17, 3.67, 178/91; 1912 - 15-19, 3.49, 95/112; 1913 - 2-0, 1.15, 9 games, none started, sent to AAA - 1914, released, retires age 30. Stuff goes from 7 to 6, control 16/14/16/14, movement 19/16/19/16 Dolly Gray - 1911 - 27-8, 2.93, 100/117; 1912 - 2-1, 3.52, 8 games, 2 starts, ends year at AAA; 1913 - released, retires age 34 in 1914. Stuff goes from 7 to 4, control 12/9/13, movement 18/16 Vean Gregg - 1911 - 7-5, 3.03, 57/45; 1912 - 19-13, 3.03, 173/111; 1913 - 18-16, 3.23, 170/146; 1914 - released, age 28, retires age 29. Stuff 13 to a 10, control - 11 to 10, movement stays at 20. there are another 5-7 that I find as odd as these guys, but it's late and I think I've got enough here for an example of what I'm seeing. The game doesn't bother with performance at all. Ratings do drop a bit in some cases, a bit more than a bit in others, but I don't think they're so drastic that they should be released and retired at such an early age. I do notice that, with the change in evaluation weights that less 28-30 years olds are dropping off the face of the earth, which is good, but I'd just like to see it tweaked so that the game wouldn't kill these guys quite as often as it does. Maybe a tweak on how player's stuff works could help as almost all of these guys stuff ratings drop off before they get discarded, even if they jump back up during their year off in the FA pool. I don't know what how the recalculate function is affecting this compared to it being off, but I have a feeling these guys might not bounce around as much with it off; then there'd be guys who just dropped off randomly like the old game used to do.

04-02-2007, 06:39 PM	#18
Ksyrup Major Leagues Join Date: Feb 2002 Posts: 413	I've seen the same issue multiple times in my sims. I'm not sure if it's a purely historical sim issue or not, but that's the only sim type I've played so far. Most of the time, rather than outright release, I've seen these guys put up excellent numbers, only to be sent to the minors the next year. I've even seen them get traded, and the new team puts them in the minors. So you'll have a guy win 20 games and not only get traded, but demoted as well. This issue underscores the problem with the AI's knowledge of ratings and willingness to make moves solely based on ratings and not waiting to see a corresponding drop in performance on the field. IRL, if you have a guy who was a solid SP one year, you wouldn't automatically drop him from your rotation the next year even if you thought he had "lost it" and/or he underperformed during spring training - you'd at least give him a few starts to verify that what you observed in talent drop is manifesting itself on the field. It just isn't realistic to make a preemptive move like that. I'm not sure what can be done to fix this, but the whole ratings vs. performance evaluation system seems like it isn't functioning as it should.

04-02-2007, 07:52 PM	#19
Frank Delarocca Jr Bat Boy Join Date: Mar 2007 Posts: 13	I simmed through another year and another few guys just seem to get shuffled off to the FA pile and retire after no one shows interest; I'd post more examples, but I'd just be basically repeating what I already posted. All of the guys have the same things in common: Stuff ratings drops a bit before they get released. Sometimes control does too, movement doesn't vary much. I'd say stuff might be the main factor in why they're no longer looked at as viable at this point (educated guess). A good percentage of the pitchers see their ratings actually move back up a point or two during their season of exile, but that doesn't seem to entice any teams into signing them up. I haven't been seeing as many guys getting banished to the bullpen in this run, they just get released instead. A few get sent to AAA, but most are outright releases come January-March. With the weight on ratings/talents being so low, I would hope the stats the players put up would be more what the AI was looking at in evaluation. From what I'm seeing, though, is that it still looks at ratings/talents no matter. I'd say that most, if not all, these guys are borderline guys, and so far in this run, no 'stars' have dropped like flies, which is a good thing, but then, I'm using recalc again, unlike my last run that didn't use it and Walter Johnson turned into silly putty. I think Ksyrup makes a great point, too. The AI jumps the gun on guys that drop off while, in real life, they'd still be getting a shot until they proved they weren't capable anymore. Then they'd go to the pen/minors and/or released into oblivion. If the game were able to be tweaked so it could emulate that flow, it'd really be damn near perfect. Of course, that's easier said than done -Frank D

04-02-2007, 09:08 PM	#20
Frank Delarocca Jr Bat Boy Join Date: Mar 2007 Posts: 13	Also, I tried unretiring these guys and bumping their rates up today. Of the 7, 1 was signed and the rest re-retired after sitting for the entire year. The one that signed, Joe Lake, went 20-17 with a 1.55 ERA (I might have bumped him a bit much as his strikeouts went up to over 200 ). -Frank D