Huge Hole in Draft AI Logic

pappyzan · 04-12-2007, 10:32 AM

when performing the yearly draft, the computer AI disregards 4-5 star middle relievers/closers for far too long into the draft. every year it will choose 1.5-2 star potential position players and starting pitchers when there are 4-5 star middle relievers/closers on the board. i've seen some drafts where most of the 4-5 star middle relievers/closers are available within the draft until all of the position players and starting pitchers that are left have a potential rating of only 1 star.

furthermore, some of the Head Scout Recommendations for picking a player in the draft are down right awful, especially in the inaugural draft. i play with scouts and coaches turned off, btw.

more detail on this can be found here:
http://www.ootpdevelopments.com/boar...d.php?t=145151

please provide a TT# whenever possible as this issue seems to be easily recreated.

sixto · 04-12-2007, 07:02 PM

I have to ask this question: in amateur drafts, why does the game generate MRs at all? (Cross-posted; please feel free to delete if it doesn't belong here).

In MLB even today - and this does not account for the bulk of baseball history (say everything before the 90s) - teams draft pitchers for their ability first and endurance a distant second. The notable exception is advanced college closers. Other than that, pitchers become middle relievers only because they fail as starters, and definitely after they are in a minor league system.

At that point, endurance is merely one of several factors, not the most important, in where he ends up. Types of pitches thrown, velocity, "makeup" - whatever that means - all come first, I'd bet.

No discussion of "I saw a guy once" is adequate to the task of determining if I'm right, but for bleeps and giggles, let's start the discussion with flamethrowing reliever Kyle Farnsworth. Farnsworth has amassed nearly 500 relief appearances since 1999. So I imagine it will surprise some of you to note that Farnsworth started 21 games for the Cubs in his rookie year.

You might therefore jump to the conclusion that this proves the point that he was a natural-born MR, as he was not nearly as good as a starter as he has been in middle relief. Only thing is, the Cubs didn't realize this tidbit themselves until a full season and more than 20 starts had passed.

On the 2007 Cardinals alone you have two pitchers, Braden Looper and Adam Wainwright, making a transition to starter from the bullpen. Wainwright was a starter in the minors, but easily transitioned to the pen and threw the last strike in the World Series. Now he's back as a starter, while Looper amassed 572 relief appearances before his first major league start. It's too early to tell how he'll do, but it's fun to note he's got a 2.08 ERA after two starts.

Some people think Giants minor-leaguer Tim Lincecum is too small to start. But for now, he's a minor-league starter. If it turns out that he hasn't got the endurance, he'll end up as a closer in the big leagues. But that just isn't known yet.

Mariano Rivera started 10 games in his rookie season. Lee Smith started 5 games in 1982. Jeff Reardon never started a game.

You can go on and on and on and on. I would never say that some pitchers don't belong in middle relief - they do - but endurance is just one factor. And while many pitchers may lack the endurance to go 9 full innings regularly, almost none has so little endurance that they couldn't go at least 6.

So the question is, what about gameplay? And what about historical replays, wouldn't relievers potentially end up as starters? Well, in terms of gameplay, I think it would be equally fun to make a pitching staff up from top to bottom out of your 10-12 best pitchers available, and not based on your 5-7 best starters and 5-7 best MRs and 1-2 best CLs. Especially in online leagues, it gives different owners the ability to value pitchers differently - one guy may see that flamethrowing lefty as a killer setup man capable of a 2.50 ERA, while another may see him as an innings-eating starter with a 4.50 ERA.

Which is a key point in my desire to eradicate the MR designation for draftees. Don't you think that many pitchers, given the chance to throw one or two innings, would have better stats than if they threw 7 a night? John Smoltz is just one example. (Actually lots of Braves support this notion.)

In terms of historical replays, I have more trouble seeing my way clear of the problem. I think the solution requires a greater mind than mine. But I've already typed too much.

I strongly urge the dev team to reconsider the fundamental importance of endurance in judging pitching prospects. When a pitcher is drafted, he has many potential outcomes in his career. Including a busted elbow. But we can discuss separating injury frequency into pitcher and position player ratings in another thread.

rasnell · 04-12-2007, 07:15 PM

I would like to see spring training and the minors really mean something toward the development of pitchers. For example, if you want to turn an astounding closer into a starter, he would require more than 200 innings of work as a starter in the minors to significantly develop endurance. That might take two seasons and that would be your juggling act as a manager, to hold him back to allow for that kind of development.

I'd like to see the same in developing players into new fielding positions, again with a high standard of more than 100 games in the position before it shows up as a development.

sixto · 04-12-2007, 08:26 PM

Quote:

Originally Posted by rasnell

For example, if you want to turn an astounding closer into a starter, he would require more than 200 innings of work as a starter in the minors to significantly develop endurance. That might take two seasons and that would be your juggling act as a manager, to hold him back to allow for that kind of development.

So you believe that although Adam Wainwright went from relief to starting in one off-season, and the Red Sox were all set to start Jon Papelbon this year, in OOTP you should have to send a pitcher to the minors for two seasons to develop the endurance required to start?

Are you saying that would be good gameplay? Or that it's true to life?

Mr. Capo · 04-13-2007, 12:02 AM

Quote:

Originally Posted by rasnell

For example, if you want to turn an astounding closer into a starter, he would require more than 200 innings of work as a starter in the minors to significantly develop endurance. That might take two seasons and that would be your juggling act as a manager, to hold him back to allow for that kind of development.

Is it even possible for endurance to increase in OOTP? Personally, I've not seen that happen.

Curtis · 04-15-2007, 04:50 AM

Back (more or less) to the topic:

Quote:

Originally Posted by pappyzan

when performing the yearly draft, the computer AI disregards 4-5 star middle relievers/closers for far too long into the draft. every year it will choose 1.5-2 star potential position players and starting pitchers when there are 4-5 star middle relievers/closers on the board. i've seen some drafts where most of the 4-5 star middle relievers/closers are available within the draft until all of the position players and starting pitchers that are left have a potential rating of only 1 star.

I'm wondering whether this issue is really about the draft AI passing by 4/5 star MRs, or if it's really about MRs being grossly overrated. What I'm saying is that my experience is that MRs are generally more highly (often much more highly) rated than they deserve. Maybe the draft AI is doing its job, and the scouts aren't doing theirs.

Just throwing that out there.

akw4572 · 04-15-2007, 10:13 AM

Quote:

Originally Posted by sixto

So you believe that although Adam Wainwright went from relief to starting in one off-season, and the Red Sox were all set to start Jon Papelbon this year, in OOTP you should have to send a pitcher to the minors for two seasons to develop the endurance required to start?

Are you saying that would be good gameplay? Or that it's true to life?

They were both starters up until last year were they not?

Mr. Capo · 04-17-2007, 02:17 AM

Quote:

Originally Posted by sixto

I have to ask this question: in amateur drafts, why does the game generate MRs at all?

Good question! I don't think there should be ANY middle relievers in the rookie draft. 32-year-old free agent MRs, yeah. But middle relief isn't something that a 20-year-old pitcher aspires to. It's something he gets relegated to when they find out he's not good enough to start, or can't throw more than 3 or 4 innings. And it usually takes a long time to find that out.

I think all the rookie pitchers should be aspiring starters, with the occasional oddball relief specialist (a modern phenomenon, anyway). Or better yet, just pitchers. Let the assignments get sorted out as their careers unfold.

SandMan · 04-17-2007, 08:47 AM

I think that all pitchers should just be positioned as pitchers. Let the manager or AI determine what the best role is by the scouting reports if on or the the players expectations, live and learn. Just because a pitcher has low endurance does not mean that he can't go 6 or 7. It just means that he loses his effectiveness prior to that. Also I feel that the endurance rating should be an overall rating and that the effectiveness (role endurance) should be a hidden value or that the ENDURANCE rating is completely thrown out and all the actual endurance ratings remain hidden.

The progression through most of baseball was for a pitcher to come to the bigs, play reliever for a few years to learn the game, then evolve to starter if good enough or a spot opens, and then end his career in the pen again. Today's game is a little different but not by much. I would also like to see more relievers that age and become decent starters, like Charlie Hough, Ken Forsch, etc...

Nutlaw · 05-06-2007, 03:42 PM

We'll let Markus sort out the draft priority issue.

TT 3531

Nutlaw · 05-08-2007, 06:47 PM

Markus seems to indicate that he likes things fine as is. There will be no action taken upon this TT.

CaLíKrAzY · 05-10-2007, 02:21 PM

Wow... That's really ashame.

This huge problem is keeping me from playing the game because unless I have the AI draft for me every year my bullpen depth in just godlike. C'mon, I'm drafting 3 1/2/4 star+ relievers in the 8th and 9th rounds!

Mel Ott · 05-10-2007, 05:02 PM

It seems that the biggest part of this problem is caused by the AI overvaluing pitcher endurance. If Markus isn't interested in addressing the SP/MR/CL quality issue, might he at least consider re-weighting the AI's pitcher preferences to take less notice of the endurance rating?

ulrichsson · 06-29-2007, 08:48 PM

I can only agree that this is a HUGE issue.
I've just completed an inaugural draft with Cubby's new roster and my pen consists of Nathan, Lidge, Gagne, Papebon, K-Rod, Ryan, Street, Shields, Neshek and Rincon.

And no, I did not only draft relievers. Most of these guys were drafted in the 20+ round when all other positions were barren wasteland.

skywalkr2 · 07-03-2007, 03:43 PM

The OP is right, this is a huge hole. This didn't happen in previous versions... in the first few years after expansion you dominate with your MRs. And you can always pick up the good MRs in the draft because the AI completely ignores them. They will be picking guys with POT of 25 and below while MR pitchers with 60+ POT are available.

Rastafreaky · 07-03-2007, 04:16 PM

Couldn't agree more. This issue is rediculous and needs to be dealt with.

koohead · 07-03-2007, 04:30 PM

Quote:

Originally Posted by Nutlaw

Markus seems to indicate that he likes things fine as is. There will be no action taken upon this TT.

I agree with this assessment. MR's tend to be overvalued A 4 star MR is in no way equal to a 4 star position player or SP. Remember that a 4 star MR is rated at 4 stars relative to other MRs...not other positions or other players. He's a great MR...but that 4th OF may be more important to a team.

Curtis · 07-05-2007, 02:37 AM

Quote:

Originally Posted by koohead

I agree with this assessment. MR's tend to be overvalued A 4 star MR is in no way equal to a 4 star position player or SP. Remember that a 4 star MR is rated at 4 stars relative to other MRs...not other positions or other players. He's a great MR...but that 4th OF may be more important to a team.

Or in other words…

Quote:

Originally Posted by Curtis

I'm wondering whether this issue is really about the draft AI passing by 4/5 star MRs, or if it's really about MRs being grossly overrated. What I'm saying is that my experience is that MRs are generally more highly (often much more highly) rated than they deserve. Maybe the draft AI is doing its job, and the scouts aren't doing theirs.

koohead · 07-05-2007, 12:43 PM

yeah that too.

ulrichsson · 07-07-2007, 11:33 AM

Nobody is saying that a four star position players is worth as much as a four star reliever. But they are certainly more valuable than one star position players. How can the AI pass over an All-Star Closer for a player who won't even make the 25 man roster???

When I drafted my team, I did not take a reliever before I had 8 position players and three starters and I still ended up with a bullpen of eight all-star relievers. This is just not right.

04-12-2007, 10:32 AM	#1
pappyzan All Star Reserve Join Date: Mar 2007 Posts: 637	Huge Hole in Draft AI Logic when performing the yearly draft, the computer AI disregards 4-5 star middle relievers/closers for far too long into the draft. every year it will choose 1.5-2 star potential position players and starting pitchers when there are 4-5 star middle relievers/closers on the board. i've seen some drafts where most of the 4-5 star middle relievers/closers are available within the draft until all of the position players and starting pitchers that are left have a potential rating of only 1 star. furthermore, some of the Head Scout Recommendations for picking a player in the draft are down right awful, especially in the inaugural draft. i play with scouts and coaches turned off, btw. more detail on this can be found here: http://www.ootpdevelopments.com/boar...d.php?t=145151 please provide a TT# whenever possible as this issue seems to be easily recreated. __________________ Last edited by pappyzan; 04-12-2007 at 10:37 AM.

04-12-2007, 07:15 PM	#3
rasnell Hall Of Famer Join Date: Jan 2003 Location: Frankfort, Kentucky Posts: 3,746	I would like to see spring training and the minors really mean something toward the development of pitchers. For example, if you want to turn an astounding closer into a starter, he would require more than 200 innings of work as a starter in the minors to significantly develop endurance. That might take two seasons and that would be your juggling act as a manager, to hold him back to allow for that kind of development. I'd like to see the same in developing players into new fielding positions, again with a high standard of more than 100 games in the position before it shows up as a development. __________________ Charlie Root won more games for the Cubs than any pitcher (201), yet was remembered for one pitch to Babe Ruth. Find out more about the 1929 World Series in my book, "Root for the Cubs: Charlie Root and the 1929 Chicago Cubs." See the web site at www.rootforthecubs.com. The book is at http://www.amazon.com/Root-Cubs-Char...t+for+the+cubs. Beta tester, OOTP 2007-2023 and iOOTP 2011-2014.

05-10-2007, 05:02 PM	#13
Mel Ott Minors (Double A) Join Date: Oct 2005 Posts: 139	It seems that the biggest part of this problem is caused by the AI overvaluing pitcher endurance. If Markus isn't interested in addressing the SP/MR/CL quality issue, might he at least consider re-weighting the AI's pitcher preferences to take less notice of the endurance rating? __________________ The National Pastime - fake baseball from 1892!

06-29-2007, 08:48 PM	#14
ulrichsson Minors (Rookie Ball) Join Date: Jan 2006 Posts: 48	I can only agree that this is a HUGE issue. I've just completed an inaugural draft with Cubby's new roster and my pen consists of Nathan, Lidge, Gagne, Papebon, K-Rod, Ryan, Street, Shields, Neshek and Rincon. And no, I did not only draft relievers. Most of these guys were drafted in the 20+ round when all other positions were barren wasteland. __________________

07-05-2007, 12:43 PM	#19
koohead Hall Of Famer Join Date: Dec 2001 Location: Phoenix, AZ Posts: 3,211	yeah that too. __________________ GM - New Jersey Bears of the NPBL;

04-12-2007, 07:02 PM	#2
sixto Hall Of Famer Join Date: Jul 2002 Posts: 2,262	I have to ask this question: in amateur drafts, why does the game generate MRs at all? (Cross-posted; please feel free to delete if it doesn't belong here). In MLB even today - and this does not account for the bulk of baseball history (say everything before the 90s) - teams draft pitchers for their ability first and endurance a distant second. The notable exception is advanced college closers. Other than that, pitchers become middle relievers only because they fail as starters, and definitely after they are in a minor league system. At that point, endurance is merely one of several factors, not the most important, in where he ends up. Types of pitches thrown, velocity, "makeup" - whatever that means - all come first, I'd bet. No discussion of "I saw a guy once" is adequate to the task of determining if I'm right, but for bleeps and giggles, let's start the discussion with flamethrowing reliever Kyle Farnsworth. Farnsworth has amassed nearly 500 relief appearances since 1999. So I imagine it will surprise some of you to note that Farnsworth started 21 games for the Cubs in his rookie year. You might therefore jump to the conclusion that this proves the point that he was a natural-born MR, as he was not nearly as good as a starter as he has been in middle relief. Only thing is, the Cubs didn't realize this tidbit themselves until a full season and more than 20 starts had passed. On the 2007 Cardinals alone you have two pitchers, Braden Looper and Adam Wainwright, making a transition to starter from the bullpen. Wainwright was a starter in the minors, but easily transitioned to the pen and threw the last strike in the World Series. Now he's back as a starter, while Looper amassed 572 relief appearances before his first major league start. It's too early to tell how he'll do, but it's fun to note he's got a 2.08 ERA after two starts. Some people think Giants minor-leaguer Tim Lincecum is too small to start. But for now, he's a minor-league starter. If it turns out that he hasn't got the endurance, he'll end up as a closer in the big leagues. But that just isn't known yet. Mariano Rivera started 10 games in his rookie season. Lee Smith started 5 games in 1982. Jeff Reardon never started a game. You can go on and on and on and on. I would never say that some pitchers don't belong in middle relief - they do - but endurance is just one factor. And while many pitchers may lack the endurance to go 9 full innings regularly, almost none has so little endurance that they couldn't go at least 6. So the question is, what about gameplay? And what about historical replays, wouldn't relievers potentially end up as starters? Well, in terms of gameplay, I think it would be equally fun to make a pitching staff up from top to bottom out of your 10-12 best pitchers available, and not based on your 5-7 best starters and 5-7 best MRs and 1-2 best CLs. Especially in online leagues, it gives different owners the ability to value pitchers differently - one guy may see that flamethrowing lefty as a killer setup man capable of a 2.50 ERA, while another may see him as an innings-eating starter with a 4.50 ERA. Which is a key point in my desire to eradicate the MR designation for draftees. Don't you think that many pitchers, given the chance to throw one or two innings, would have better stats than if they threw 7 a night? John Smoltz is just one example. (Actually lots of Braves support this notion.) In terms of historical replays, I have more trouble seeing my way clear of the problem. I think the solution requires a greater mind than mine. But I've already typed too much. I strongly urge the dev team to reconsider the fundamental importance of endurance in judging pitching prospects. When a pitcher is drafted, he has many potential outcomes in his career. Including a busted elbow. But we can discuss separating injury frequency into pitcher and position player ratings in another thread.

04-17-2007, 08:47 AM	#9
SandMan Hall Of Famer Join Date: Aug 2003 Posts: 2,946	I think that all pitchers should just be positioned as pitchers. Let the manager or AI determine what the best role is by the scouting reports if on or the the players expectations, live and learn. Just because a pitcher has low endurance does not mean that he can't go 6 or 7. It just means that he loses his effectiveness prior to that. Also I feel that the endurance rating should be an overall rating and that the effectiveness (role endurance) should be a hidden value or that the ENDURANCE rating is completely thrown out and all the actual endurance ratings remain hidden. The progression through most of baseball was for a pitcher to come to the bigs, play reliever for a few years to learn the game, then evolve to starter if good enough or a spot opens, and then end his career in the pen again. Today's game is a little different but not by much. I would also like to see more relievers that age and become decent starters, like Charlie Hough, Ken Forsch, etc...

05-06-2007, 03:42 PM	#10
Nutlaw Hall Of Famer Join Date: Apr 2003 Posts: 9,538	We'll let Markus sort out the draft priority issue. TT 3531

05-08-2007, 06:47 PM	#11
Nutlaw Hall Of Famer Join Date: Apr 2003 Posts: 9,538	Markus seems to indicate that he likes things fine as is. There will be no action taken upon this TT.

05-10-2007, 02:21 PM	#12
CaLíKrAzY All Star Reserve Join Date: Dec 2001 Location: Costa Mesa, CaLí Posts: 722	Wow... That's really ashame. This huge problem is keeping me from playing the game because unless I have the AI draft for me every year my bullpen depth in just godlike. C'mon, I'm drafting 3 1/2/4 star+ relievers in the 8th and 9th rounds!

07-03-2007, 03:43 PM	#15
skywalkr2 Major Leagues Join Date: Feb 2004 Location: St Louis, MO Posts: 438	The OP is right, this is a huge hole. This didn't happen in previous versions... in the first few years after expansion you dominate with your MRs. And you can always pick up the good MRs in the draft because the AI completely ignores them. They will be picking guys with POT of 25 and below while MR pitchers with 60+ POT are available.

07-03-2007, 04:16 PM	#16
Rastafreaky Minors (Double A) Join Date: Jun 2002 Location: Seattle, WA Posts: 104	Couldn't agree more. This issue is rediculous and needs to be dealt with.

07-07-2007, 11:33 AM	#20
ulrichsson Minors (Rookie Ball) Join Date: Jan 2006 Posts: 48	Nobody is saying that a four star position players is worth as much as a four star reliever. But they are certainly more valuable than one star position players. How can the AI pass over an All-Star Closer for a player who won't even make the 25 man roster??? When I drafted my team, I did not take a reliever before I had 8 position players and three starters and I still ended up with a bullpen of eight all-star relievers. This is just not right. __________________ Last edited by ulrichsson; 07-07-2007 at 11:35 AM.