Relief Pitching Broken Agian in 6.5b - Page 2

LeiterFanatic · 12-01-2005, 09:47 AM

Quote:

Originally Posted by Aordolin

tough go back to 6.12.

This is what I finally did. I agree with what you said. I just ran ANOTHER test league with starter endurance set to fair and reliever usage to high, as suggested and the league leader in the NL was 35 saves and in the AL it was 27 saves. The same league imported from the same Lahman database under 6.12 clocked 50 saves for the League leader in the NL with 9 relievers with 35 or more saves and 44 for the AL with 6 with 30 or more saves. So try and tell me that there is no issue here. At least since this is the final 6.5 I dont have to keep giving it a try after each patch came out.

Derek4jc · 12-02-2005, 03:05 AM

I'm more disappointed to have paid for the upgrade (having saves working fine previously) and then somehow it ended up worse than started. I know it was said to have been the last patch and whatnot, but to not address this 6.5 problem would be (in my opinion) a mistake and a disservice to the paying customers who were certainly caught off guard and disappointed by this setback.

KurtBevacqua · 12-02-2005, 03:45 AM

I don't know that there is a problem, certainly not a bug. I'm seeing a lot of people say in 1, 2, or 3 seasons totals are low. OK, to me that's anecdotal. Come back with 50 seasons, side by side, 6.51 next to a 6.12 version, and lets see what the numbers say.

And if MLB has closers get 86.2% (or whatever) of the saves, and OOTP has 78.9%, not a big deal. OOTP is not Diamond Mind or one of those other games seeking absolute statistical matching. It's not intended to be a replication, but a universe unto itself. There's not official acknowledgement of a problem here because likely the game is doing very close to what Markus intended. If it were a bug where the closer appeared in less save situations than the setup man, then I'd agree there might be a problem with the gameplanning. But as long closers are getting the preponderence of the saves there's really not a lot to bitch about.

LeiterFanatic · 12-02-2005, 06:08 AM

Quote:

Originally Posted by KurtBevacqua

But as long closers are getting the preponderence of the saves there's really not a lot to bitch about.

You are missing the point...save totals worked better before. If they never worked correctly, then I could buy into your arguement. But for some of us anyway, save totals for closers dropped after buying 6.5 and that aspect of the game outweighs having a milestone board and 7 day lineups. So I really dont need to run 50 year comparisons to know that this aspect of the game dispeases me, that it has changed in my experience, and its not as fun. I just dont play it anymore.

Aordolin · 12-02-2005, 09:15 AM

Quote:

Originally Posted by KurtBevacqua

I don't know that there is a problem, certainly not a bug. I'm seeing a lot of people say in 1, 2, or 3 seasons totals are low. OK, to me that's anecdotal. Come back with 50 seasons, side by side, 6.51 next to a 6.12 version, and lets see what the numbers say.

And if MLB has closers get 86.2% (or whatever) of the saves, and OOTP has 78.9%, not a big deal. OOTP is not Diamond Mind or one of those other games seeking absolute statistical matching. It's not intended to be a replication, but a universe unto itself. There's not official acknowledgement of a problem here because likely the game is doing very close to what Markus intended. If it were a bug where the closer appeared in less save situations than the setup man, then I'd agree there might be a problem with the gameplanning. But as long closers are getting the preponderence of the saves there's really not a lot to bitch about.

Its not a bug, its crappy AI. Something was changed in bullpen management with the 6.5 expansion, seems I do remember reading it in the notes. And it made things worse much the way 6.10 made things worse until fixed with 6.12. Yes this isnt a precise replecation of real baseball however that isnt the point really. If your talking about replaying historical seasons, yes your right. If I want to repeat save totals by era exactly your right. The problem IS NOT THE NUMBER OF SAVES OR SAVE CHANCES! I can live with 30 saves ok, but what I dont want to live with is 30 saves from the closer and 25 from the rest of the bullpen. That is wrong. wrong wrong wrong wrong wrong wrong wrong. did i say it was wrong? how can I possibly make this more clear? There is NO point in even having the pitching roles if they mean as much as this. How would you feel if you set your 5 man pitching rotation but have the game just randomly decide to ignore that rotation and start some guy from the bullpen not in the rotation. Because its the same thing. Closer is a position, and like I said before a position that should be filled by the person you put in that role, unless tired which should be rare. you may not think 78% or whatever is not a big deal but your still looking at about 1/4 of all possible starting situations where the player best suited for the position isnt being used. Would you except that from your starting pitcher? or your first baseman? you wouldnt unless thats what you intended.

As far as simming 50 years.. 50 years? are you kidding? How would that be any better than simming 5 or 10? Thats not necessary, its over kill. You can get a solid average over 5 years.

Some posted its better though with this new release, I havent had the chance to test it thoroughly and will look at it this weekend. But I am discouraged that people are still saying its as bad.

LeiterFanatic · 12-02-2005, 09:34 AM

Not just that, but it kills the management aspect of the sim. Why blow 8 million on a closer that the AI wont use? When Gerold Riggan gets as many saves as Armando Benitez on the 2000 Mets, there is something wrong. I see it as the cousin of the "bad player "bug which got everyone outraged. The AI would sit Ruth and play some idiot... well this is the same thing... the AI sits Eckersley and puts Jim Corsi into the save situation. Gah!

Malleus Dei · 12-02-2005, 11:29 PM

I think there may be something awry with your manager or league or era settings, as I am not having this problem with 6.51 in ANY of my leagues. Or maybe I just have mine set to where everything works.

Some people report a bad problem, and most if not many report no problem at all. That is whispering "individual settings" to me.

Derek4jc · 12-03-2005, 12:18 AM

I disagree that it has to do with individual settings. I don't think I am getting less saves, I just know I am getting them spread out in poor ways. Here are some examples from my online league. (Pretty much all of them are like this but these are some)

Date: 7/1

KC Royals:
Mike MacDougal - 8 (setup man) (40 appearences)
Byron Embry - 6 (Closer) (18 appearences)
Jaime Cerda - 1 (setup man) (11 appearences)

Embry has never been hurt and has always been the closer.

Red Sox:
Scott Linebrink - 8 (Closer) (12 Appearences?!?!)
Ryan Wagner - 7 (Setup man) (30 appearences)
Carmen Cali - 2 (Setup man) (20 appearences)

Linebrink has never been hurt and has always been the closer.

On the twins JC Romero never got hurt, and yet he's only been used 11 times. He has pitched 1 times since 5/29. The trouble is he is in a contract year and the guy needs him to have playing time so he'll he able to resign him.

These are the problems. Odd nonuse and the saves being too far spread. As far as I know, there is no setting that I can change that has anything to do with how a bullpen will be used in save ops. If it wasn't an online league, I'd be able to plug the guys in myself, but there is no way on my Red Sox team i'd ever only use my closer 12 times, while totally healthy, in 3 months, and just give out those save ops to my setup guys. it makes no sense. Same thing goes with the KC setup guy having MORE saves then the closer. These results are across the board throughout the league. No a lack of saves, but an improper use of the pen by the AI, mostly in the closer position.

Aordolin · 12-03-2005, 08:14 AM

It was suggested to move endurance to fair. Mine was on normal and things did improve. Im going to run some tests this weekend, and Ill post the results. Something was changed though. Because in 6.12 things worked ok. I changed NOTHING in the settings and after installing 6.5 expansion it went to hell using same settings. After I post my results atleast we can see if they are in line with what MD and others are getting. Atleast then well know if some of us just have different expecations than others.

KurtBevacqua · 12-03-2005, 10:06 AM

.

KurtBevacqua · 12-03-2005, 10:07 AM

Quote:

Originally Posted by Aordolin

As far as simming 50 years.. 50 years? are you kidding? How would that be any better than simming 5 or 10? Thats not necessary, its over kill. You can get a solid average over 5 years.

Because 5 years is not a statistically sound pool of data. It would be like saying because we had 5 hot summers in a row we must have global warming.

And most of the bitching I'm seeing is people saying "I had a couple bad seasons, therefore it's bad." I suggest a larger study with good data would provide ample evidence of a problem. Otherwise it is anecdotal or based on it just not seeming right, when there is no statistical evidence to support your theory.

Find a book on research methodology or research analysis and you'll see what I'm talking about. The arguments here haven't convinced me there is a problem, therefore it's probably doubtful they've convinced Markus either.

Aordolin · 12-03-2005, 05:22 PM

Well here it goes. I intended a 25 season study but I cut it off after 7. Perhaps I will go and sim the rest but the data is so consistant from year to year that is is pointless to continue. Since I am not studying the Earths weather patterns, just baseball, a 7yr sample is more than enough to get a reasonable pattern.

*edit* added 8th, 9th, 10th, and 11th sim seasons. Thats all folks. Unless someone can convince me of a logical reason to continue with this and not get the same results.

My league settings:

OOTP 6.51
20 team 2 leagues, 4 divisions
154 game season
ERA Settings:
Bunting: Often
Hit & Run: Often
Sealing Bases: Normal
Pinch Hitting: Normal
Use Relievers: Very Often
Picther Endurance: Fair
Typical Rotations: 5 man
Leadoff: Favor OBP

Injuries were turned OFF
Team Strategies were all set to default on all teams
Each team had 5 middle relievers and the closer (6 total) with 5 Starting Pitchers
Not sure if it matters, but my league uses the DH
This is a completely fictional league, no roster sets, I built it from the ground up.

CL=Closer
SU1=Setup Man role 1
SU2=Setup Man role 2
MR=Total of all Middle/Mopup Reliever saves.
PCT=Percentage of total saves by closer

Year 1 Total saves: 818 CL: 540 SU1: 132 SU2: 79 MR: 67 PCT: 66.0%

Year 2 Total saves: 784 CL: 526 SU1: 115 SU2: 78 MR: 65 PCT: 67.1%

Year 3 Total saves: 800 CL: 526 SU1: 121 SU2: 76 MR: 77 PCT: 65.7%

Year 4 Total saves: 815 CL: 520 SU1: 118 SU2: 85 MR: 92 PCT: 63.8%

Year 5 Total saves: 835 CL: 562 SU1: 121 SU2: 70 MR: 82 PCT: 67.3%

Year 6 Total saves: 800 CL: 541 SU1: 125 SU2: 59 MR: 75 PCT: 67.6%

Year 7 Total saves: 773 CL: 508 SU1: 116 SU2: 74 MR: 75 PCT: 65.7%

Year 8 Total saves: 832 CL: 578 SU1: 120 SU2: 73 MR: 61 PCT: 69.5%

Year 9 Total saves: 750 CL: 514 SU1: 119 SU2: 55 MR: 62 PCT: 68.5%

Year 10 Total saves: 802 CL: 517 SU1: 129 SU2: 83 MR: 73 PCT: 64.5%

Year 11 Total saves: 785 CL: 524 SU1: 123 SU2: 72 MR: 66 PCT: 66.8%

So looking at those numbers thats about a 66% average. Thats over all the teams. If you look at teams individually, they are better or worse. Some questions people might ask about these numbers.

Where closers with poor percentages ineffective? High ERA, etc?
No. In some cases I had 5.00 ERA closers with 35+ saves and ended up with some of the more favorable % in that season. Some closers where in the 1.50-2.00 range and had a % barely over 50%.

Did the computer manager move pitchers around, going with the hot hand?
No. Everyteam was set to human manager. I set all the rotations before the season. Since injuries were off this removes the possiblity of numbers being skewed because of an injured closer.

Where bad teams worse with their bullpen management than good teams?
No. There was no pattern based on records. Some teams with horrible records had good closer usage while some great teams were among the worst offenders.

I dont think I need to go research MLB save totals. I think its save to say that their closers dont get 66% of a teams saves. Ive researched it in the past trust me it was 90% plus ATLEAST!! I wasnt even counting the number of SAVE OPPORTUNITIES! Would have been way to time consuming to do that, but I think its safe to assume from the ERAs of some of the scrubs who ended up with an abnormal number of saves that they blew enough to make a pretty large effect on their teams win/loss record that season. I conservatively guess as many 8-10 wins. Teams that had really good bullpens were more likely to suffer with poor bullpen usage, but it was not always the case. You would think that SU1 who got 10-15 saves on average a season would be as good as or maybe better than the closer, but sadly even this wasnt the case in every situation. It was in a couple though I must admit that.

Fixes? Its obvious that Im going to have to live with this. The lack of any official response has made that clear. So the only way I can see to get any semblance of statisical reality out of your closer is to place the closer in atleast Setup role 1 and possibly 2 as well. Hell maybe we should just put the closer in every role in the bullpen!

A little poscript to this is my ERA settings. In my Solo league my pitcher endurance was set to normal. Because I like getting complete games about 1/3 of the time. It should be noted that changing the endurance to fair did not change the numbers at all. I dont have the statiscal data on it anymore but I simmed many seasons, atleast 25, since the release of 6.5 and after each of the beta patches to see how it would affect my league before continuing it. My totals in 6.12 using the same era settings (endurance: normal , relievers: very often) I was in the 80% range. WITHOUT using tricks like putting the closer in any other position except CLOSER.

Now I would like to challenge anyone who IS NOT having these problems to post your own statisical data for your own league. I would prefer hearing from those who play like I do with my Solo, basically as a GM/God mod. Basically I play my league like an Online league would be played. So people in Online leagues feel free to post your data as well. How are you setting up your pens? What are your Era settings? I would love to get to the point you are at, If the game isnt broken then tell me what I am doing wrong.

Another interesting observation I noticed when simming those last 4 seasons I hadnt noticed before. The setup1 tends to show up on the leaderboards early in the season but is then replaced as it goes by the closers. I cant confirm this but it seems as though the closer is used better as the season goes on but not in april and may. I cant back this up other than remembering seeing SU1x with 6 saves in april on the leaderboard and remembering at the end of the season he only has 8 for the season. Odd eh?

Derek4jc · 12-03-2005, 06:24 PM

Wow, I guess some people want you to jump though hoops before they're willing to believe there is an issue. Good study Aordolin.

jbmagic · 12-03-2005, 06:43 PM

great study Aordolin

i am surprise this issue with the closer was not fix on 6.51.

Quote:

Fixes? Its obvious that Im going to have to live with this. The lack of any official response has made that clear. So the only way I can see to get any semblance of statisical reality out of your closer is to place the closer in atleast Setup role 1 and possibly 2 as well. Hell maybe we should just put the closer in every role in the bullpen!

this cracks me up, when people say to do this. it might help your team, but it wont help the cpu teams. the cpu closer will still only get 66% of the team total saves.

Aordolin · 12-03-2005, 11:06 PM

Ok as suggested in another thread I have no run some seasons using SkyDogs modifiers. I used the same league as I did when doing my earlier tests. So all the other settings are the same, injuries off, human manager was on all teams, I set the rotations, lineups etc, just see my above post.

I set baseball ERA to modern, recalcuated league totals for accuracy then changed everything as listed below per SkyDogs specs.

http://www.ootpdevelopments.com/boar...67&postcount=1

RATINGS WEIGHT: 67%
CURRENT YEAR STATS WEIGHT: 33%
LAST YEAR STATS WEIGHT: 0%
YEAR BEFORE LAST: 0%

Batter--Contact: 9%
Batter--Gap: 0%
Batter--Power: 9%
Batter--Eye: 22%
Batter--Strikeouts: 41%

Pitcher--Staff: 7%
Pitcher--Control: 25%
Pitcher--Movement: 25%
Pitcher--Endurance: 5%
Player--Speed: 2%
Player--Stealing: 2%

Hitter Development: +65%
Hitter Aging: -54%
Pitcher Development: +65%
Pitcher Aging: -54%

Draft Rounds: 12
Trade Frequency: Low
Trading AI Preference: Favor Prospects
Scouts: Off

ERA SETTINGS
Bunting: Often
Hit & Run: Normal
Stealing: Normal
Pinch Hitting: Often
Use Relievers: Often
Pitcher Endurance: Normal
Rotation: 5 Man
Leadoff Lineup AI: Favor OBP

CL=Closer
SU1=Setup Man role 1
SU2=Setup Man role 2
MR=Total of all Middle/Mopup Reliever saves.
PCT=Percentage of total saves by closer

Year 1 Total saves: 622 CL: 190 SU1: 271 SU2: 134 MR: 27 PCT: 30.5%

Year 2 Total saves: 672 CL: 240 SU1: 254 SU2: 141 MR: 37 PCT: 35.7%

Year 3 Total saves: 649 CL: 214 SU1: 266 SU2: 139 MR: 30 PCT: 32.9%

Year 4 Total saves: 686 CL: 204 SU1: 271 SU2: 178 MR: 33 PCT: 29.7%

Year 5 Total saves: 718 CL: 207 SU1: 283 SU2: 188 MR: 40 PCT: 28.8%

So.. as you can see its alot worse. Its as if the game has now decided to TOTALLY ignore your closer settings and use whoever it wants. These numbers though are not all that different from the ones I posted earlier though... IF.. you substitute SU1 and in some cased SU2 for the closer. Yes, I said it, the Setup men were ALL the league leaders, only twice did an actual closer finish in the top 5 in saves.

You will also notice that the MR saves went down alot, they were actually pretty realistic IMO. Sadly though I do not believe this was do to any settings change, it was mostly because those pitchers under these settings did not pitch nearly as much as they did under my 11 year test run earlier.

I made a mistake earlier, but I found the results kinda interesting. I used Skydogs modifiers however I had not modified my League Totals, just the ERA settings. Also I was running 6.50, the original build released last summer. So I dont know if it was that or my changes in the league Totals. However I simmed 3 years:

Year 1 Total saves: 738 CL: 359 SU1: 210 SU2: 119 MR: 50 PCT: 48.6%

Year 2 Total saves: 745 CL: 380 SU1: 200 SU2: 119 MR: 46 PCT: 51.0%

Year 3 Total saves: 730 CL: 349 SU1: 224 SU2: 106 MR: 51 PCT: 47.8%

It was after season 3 I realized I was not running 6.51 and had not recalcuated the league totals. Unless I misunderstood the insturctions, but since there wasnt anything in the post about league totals being set to a particular thing, I just went with the ERA defaults.

Ok now the last thing I did was I had had the computer auto do the bullpen and rotations. It was still set to human manager I just asked the computer for advice. In some instances it changed the closer I had to one of those favorite setup men it liked to use. So I simmed 5 more seasons with that. LOL it was the same as with my roations. And the most hilarious thing was the closer that they took out of one team and put at setup 1 finally got the saves!! Yes thats right, favorite reliever#1, who the computer moved to closer got just as many saves as the previous closer did! And that old closer got all the saves that reliever got in previous sims! Brillant!

Anyway excuse me for not posting those results. I am to tired and, dont feel like adding them up. But they look pretty much the same as the first set I posted from just looking at them. I would be shocked if they didnt fall in the 30-35% range.

What this means for me is that I am convinced the game is broken, its not me. I just dont see how it can be. But I want to see your test results and I want your league settings and era settings and everything you can send me about your league. Hell email me your league and let me look at it. I just dont see how I can be so far off from everyone else.

Now if you manage your league only and the computer manages everything else, well I think I can see why you dont have a problem. My guess is the computer manager would probably do a better job at it and create alot more realistic results. But Solo players who play like I do and ONLINE leagues especially.. how are you NOT having this same problem? IF you do play the way I do or are in an ONLINE league I want to hear from you! And if Markus reads this, please say something! At this point I guess I can understand that its not going to be fixed. But I dont want to see this repeated in the new version for sure. And I wont be buying it until I feel I have the assurances that statisical accuracy for relief pitchers is going to be a priority. Some people may not think this is a big deal but I dont think you realize how much of a bummer this is for me. It just ruins my league so much that I dont even want to play anymore.

ibdb · 12-04-2005, 01:56 AM

Thank you very much for all your work on this. I'm not crazy (at least about this) after all.

KurtBevacqua · 12-04-2005, 10:03 AM

Quote:

Originally Posted by Aordolin

I dont think I need to go research MLB save totals. I think its save to say that their closers dont get 66% of a teams saves. Ive researched it in the past trust me it was 90% plus ATLEAST!! I wasnt even counting the number of SAVE OPPORTUNITIES! Would have been way to time consuming to do that, but I think its safe to assume from the ERAs of some of the scrubs who ended up with an abnormal number of saves that they blew enough to make a pretty large effect on their teams win/loss record that season. I conservatively guess as many 8-10 wins. Teams that had really good bullpens were more likely to suffer with poor bullpen usage, but it was not always the case. You would think that SU1 who got 10-15 saves on average a season would be as good as or maybe better than the closer, but sadly even this wasnt the case in every situation. It was in a couple though I must admit that.

Here's the data on the National League, 2000-2004, on the total number of saves in the league compared the the number of saves earned by the closers for each team. 90%? Not quite.

League Saves Closer Saves
2000 627 443 70.65%
2001 621 466 75.04%
2002 666 572 85.89%
2003 661 469 70.95%
2004 687 539 78.46%
Total 3262 2489 76.30%

I'll grant the game is off by 10%. Not a deal breaker for me. If it kills the game for some folks then I'm sorry it falls short of their expectations.

KurtBevacqua · 12-04-2005, 10:27 AM

Played with some more numbers, going back a bit further, but still the same current era (1980-current).

1990 AL 637 446 70.02%
1991 AL 618 443 71.68%
1992 AL 619 461 74.47%

So these are even closer to 66%. Then I decided to go back a bit further into the 80's, oh, some of you are not gonna like these numbers.

1984 AL 513 310 60.43%
1985 AL 522 307 58.81%
1986 NL 480 296 61.67%
1987 NL 496 274 55.24%

One could argue that for the 80's the game is actually overly using the closer. In the 21st century it appears the number is around 75%, in the 90's 70-75%, and in the 80's it was 55-65%. I think the game using closers 66% of the time is statistically a very good compromise across eras. Maybe the best thing to do is for Markus to create another era and another setting for relievers. But his current model is not broken, just a compromise.

Aordolin · 12-04-2005, 01:27 PM

agian your missing the point. Yes I realize that the numbers are close but your completely disregarding the reasons why the numbers are low. closers get injured are ineffective.

Reposting this from another thread...

2005 MLB Closer Stats (source ESPN.com)

Baltimore - BJ Ryan 36 Team:38 Pct:94.7%
Arizona - Valverde 15, Lyon 14, Bruney 12 Team:45 Pct:33.3%
Atlanta - Reitsma 15, Kolb 11, Farnsworth 10 Team: 38 Pct:39.4%
Boston - Foulke 15, Timlin 13, Schilling 9 Team: 38 Pct:39.4%
ChiCubs - Dempster 33 Team: 39 Pct:84.6%
ChiSox- Hermanson 34 Team: 54 Pct:62.9%
CinReds - Weathers 15, Graves 10 Team: 31 Pct:48.4%
Cleveland - Wickman 45 Team: 51 Pct:88.2%
Colorado - Fuentes 31 Team 37 Pct:83.8%
Detroit - Urbina 9, Percival 8 Team 38 Pct:23.7%
Florida - Jones 40 Team 42 Pct:95.2%
Houston - Lidge 42 Team 45 Pct:93.3%
KC - MacDougal 21 Team: 25 Pct:84.0%
LAAngels - Rodriguez 45 Team: 54 Pct:83.3%
LADodgers - Brazoban 21, Gagne 8 Team: 40 Pct:52.5%
Milwaukee - Turnbow 39 Team: 46 Pct:84.8%
Minnesota - Nathan 43 Team: 44 Pct:97.7%
NYMets - Looper 28 Team 38 Pct:73.7
NYYanks - Rivera 43 Team 46 Pct:93.5
Oakland - Street 23, Dotel 7 Team: 38 Pct:60.5%
Philadelphia - Wagner 38 Team: 40 Pct:95.0%
Pittsburg - Mesa 27 Team: 35 Pct:77.1%
SanDeigo - Hoffman 43 Team 45 Pct:95.5%
SanFransico- Walker 23, Benitez 19 Team: 46 Pct:50.0%
Seattle - Guardado 36 Team: 39 Pct:92.3
StLouis - Isringhausen 39 Team: 48 Pct:81.3%
TampaBay - Baez 41 Team: 43 Pct:95.3%
Texas - Cordero 37 Team: 46 Pct:80.4%
Toronto - Batista 31 Team: 35 Pct:88.6%
Washington - Cordero 47 Team: 51 Pct:92.2%

Ok time to break it down:

10 teams with 90% or above
19 teams with 80% or above
21 teams with 70% or above

9 teams were under 69%. Ok why?

Dodgers - Gagne was injured almost all year, bullpen by committee or hed by in that 90% or above
Oakland - Dotel injured, Street comes in during the season. I dont have the numbers since he became the closer but I bet he was 90% too.
Atlanta - Kolb is horrible, their bullpen is a disaster since moving Smoltz to rotation. So ofcourse they are going to be in committee situation
Cincinatti - Graves disaster, team is a disaster. You know you got problems with David Weathers is your closer
Boston - Foulke spent a good portion of year on DL.
ChicagoWSox - Hermanson had 34, but I dont know what happened to him, was he injured? Cant explain that one.

San Fransico, Arizona, Detroit - ? I dont know what their deal was besides just being bad

So thats 955 saves from each teams top save man, and 1255 total saves for 76.1%

If I take out the teams with injuries or bullpen issues and just keep the 22 solid bullpens in (I kept White Sox in) the percentage becomes 87.0%

The moral is teams who have a solid closer use it way more than 66% and way more like 85% and above.

Let me break down my league I was posting earlier. Remember I turned off injuries. So injury is not a factor.

Year 1
Team 1: Closer: 28 Team: 46 Pct:60.9%
Team 2: Closer: 20 Team: 30 Pct:66.7%
Team 3: Closer: 32 Team: 46 Pct:69.6%
Team 4: Closer: 29 Team: 48 Pct:60.4%
Team 5: Closer: 26 Team: 35 Pct:74.3%
Team 6: Closer: 23 Team: 35 Pct:65.7%
Team 7: Closer: 30 Team: 43 Pct:69.8%
Team 8: Closer: 39 Team: 53 Pct:73.6%
Team 9: Closer: 24 Team: 38 Pct:63.2%
Team 10: Closer: 30 Team: 38 Pct:78.9%
Team 11: Closer: 32 Team: 52 Pct:61.5%
Team 12: Closer: 32 Team: 42 Pct:76.2%
Team 13: Closer: 27 Team: 42 Pct:64.3%
Team 14: Closer: 8 Team: 28 Pct:28.6%
Team 15: Closer: 19 Team: 40 Pct:47.5%
Team 16: Closer: 20 Team: 34 Pct:58.8%
Team 17: Closer: 28 Team: 37 Pct:75.7%
Team 18: Closer: 27 Team: 33 Pct:81.8%
Team 19: Closer: 37 Team: 48 Pct:77.1%
Team 20: Closer: 29 Team: 50 Pct:58.0%

Ok compare my numbers with MLB. The totals are similar, I have fewer teams but the team save totals are pretty much identical. Remember no injuries are on so thats not a factor as it was in the MLB numbers. I can post all the seasons data if you want but they are all comparable to this. ONE TEAM had a better than 80% usage of its closer, compare that to 19 of 30 teams in 2005.

The issue is the game does not put in the closer. Its not whether the numbers end up in the end to be equal to what that should be in MLB. THEY ARENT SUPPOSED TO BE 66% FOR ALL TEAMS!!!

Quote:

Played with some more numbers, going back a bit further, but still the same current era (1980-current).

1990 AL 637 446 70.02%
1991 AL 618 443 71.68%
1992 AL 619 461 74.47%

So these are even closer to 66%. Then I decided to go back a bit further into the 80's, oh, some of you are not gonna like these numbers.

1984 AL 513 310 60.43%
1985 AL 522 307 58.81%
1986 NL 480 296 61.67%
1987 NL 496 274 55.24%

One could argue that for the 80's the game is actually overly using the closer. In the 21st century it appears the number is around 75%, in the 90's 70-75%, and in the 80's it was 55-65%. I think the game using closers 66% of the time is statistically a very good compromise across eras. Maybe the best thing to do is for Markus to create another era and another setting for relievers. But his current model is not broken, just a compromise.

Well I am talking about modern numbers. I grant you in the 80s the closer was just starting to come into its own. So I dont doubt you at all that those numbers are correct. The game has changed. OK so if someone wants to play in that era fine, but what about those who want to play with modern settings, how are we supposed to do that?

ibdb · 12-04-2005, 06:57 PM

Aordolin -- Have you noticed any difference in NL (non DH) and AL (DH) leagues? I was setting up my own test league for trial and error. After a quick two months sim to see if things seemed to be at least set up correctly, I've noticed distinct differences between non-DH and DH leagues.

In a non-DH league, it's not uncommon for the save distribution to be 12-2-1 (CL-SU1-SU2). In a DH league, it's not uncommon for the save distribution to be more like 9-4-3-3-1 (CL-SU1-SU2-MR1-MR2).

12-02-2005, 03:45 AM	#23
KurtBevacqua Hall Of Famer Join Date: May 2003 Posts: 2,968 Blog Entries: 2	I don't know that there is a problem, certainly not a bug. I'm seeing a lot of people say in 1, 2, or 3 seasons totals are low. OK, to me that's anecdotal. Come back with 50 seasons, side by side, 6.51 next to a 6.12 version, and lets see what the numbers say. And if MLB has closers get 86.2% (or whatever) of the saves, and OOTP has 78.9%, not a big deal. OOTP is not Diamond Mind or one of those other games seeking absolute statistical matching. It's not intended to be a replication, but a universe unto itself. There's not official acknowledgement of a problem here because likely the game is doing very close to what Markus intended. If it were a bug where the closer appeared in less save situations than the setup man, then I'd agree there might be a problem with the gameplanning. But as long closers are getting the preponderence of the saves there's really not a lot to bitch about. __________________ "The type and formula of most schemes of philanthropy or humanitarianism is this: A and B put their heads together to decide what C shall be made to do for D. The radical vice of all these schemes, from a sociological point of view, is that C is not allowed a voice in the matter, and his position, character, and interests, as well as the ultimate effects on society through C's interests, are entirely overlooked. I call C the Forgotten Man" - William Graham Sumner

12-02-2005, 09:34 AM	#26
LeiterFanatic All Star Starter Join Date: Mar 2003 Location: Miami, FL Posts: 1,344	Not just that, but it kills the management aspect of the sim. Why blow 8 million on a closer that the AI wont use? When Gerold Riggan gets as many saves as Armando Benitez on the 2000 Mets, there is something wrong. I see it as the cousin of the "bad player "bug which got everyone outraged. The AI would sit Ruth and play some idiot... well this is the same thing... the AI sits Eckersley and puts Jim Corsi into the save situation. Gah! Last edited by LeiterFanatic; 12-02-2005 at 09:37 AM.

12-03-2005, 10:06 AM	#30
KurtBevacqua Hall Of Famer Join Date: May 2003 Posts: 2,968 Blog Entries: 2	. __________________ "The type and formula of most schemes of philanthropy or humanitarianism is this: A and B put their heads together to decide what C shall be made to do for D. The radical vice of all these schemes, from a sociological point of view, is that C is not allowed a voice in the matter, and his position, character, and interests, as well as the ultimate effects on society through C's interests, are entirely overlooked. I call C the Forgotten Man" - William Graham Sumner Last edited by KurtBevacqua; 12-03-2005 at 10:07 AM.

12-03-2005, 05:22 PM	#32
Aordolin All Star Reserve Join Date: Apr 2004 Location: Houston, TX Posts: 586	Well here it goes. I intended a 25 season study but I cut it off after 7. Perhaps I will go and sim the rest but the data is so consistant from year to year that is is pointless to continue. Since I am not studying the Earths weather patterns, just baseball, a 7yr sample is more than enough to get a reasonable pattern. edit added 8th, 9th, 10th, and 11th sim seasons. Thats all folks. Unless someone can convince me of a logical reason to continue with this and not get the same results. My league settings: OOTP 6.51 20 team 2 leagues, 4 divisions 154 game season ERA Settings: Bunting: Often Hit & Run: Often Sealing Bases: Normal Pinch Hitting: Normal Use Relievers: Very Often Picther Endurance: Fair Typical Rotations: 5 man Leadoff: Favor OBP Injuries were turned OFF Team Strategies were all set to default on all teams Each team had 5 middle relievers and the closer (6 total) with 5 Starting Pitchers Not sure if it matters, but my league uses the DH This is a completely fictional league, no roster sets, I built it from the ground up. CL=Closer SU1=Setup Man role 1 SU2=Setup Man role 2 MR=Total of all Middle/Mopup Reliever saves. PCT=Percentage of total saves by closer Year 1 Total saves: 818 CL: 540 SU1: 132 SU2: 79 MR: 67 PCT: 66.0% Year 2 Total saves: 784 CL: 526 SU1: 115 SU2: 78 MR: 65 PCT: 67.1% Year 3 Total saves: 800 CL: 526 SU1: 121 SU2: 76 MR: 77 PCT: 65.7% Year 4 Total saves: 815 CL: 520 SU1: 118 SU2: 85 MR: 92 PCT: 63.8% Year 5 Total saves: 835 CL: 562 SU1: 121 SU2: 70 MR: 82 PCT: 67.3% Year 6 Total saves: 800 CL: 541 SU1: 125 SU2: 59 MR: 75 PCT: 67.6% Year 7 Total saves: 773 CL: 508 SU1: 116 SU2: 74 MR: 75 PCT: 65.7% Year 8 Total saves: 832 CL: 578 SU1: 120 SU2: 73 MR: 61 PCT: 69.5% Year 9 Total saves: 750 CL: 514 SU1: 119 SU2: 55 MR: 62 PCT: 68.5% Year 10 Total saves: 802 CL: 517 SU1: 129 SU2: 83 MR: 73 PCT: 64.5% Year 11 Total saves: 785 CL: 524 SU1: 123 SU2: 72 MR: 66 PCT: 66.8% So looking at those numbers thats about a 66% average. Thats over all the teams. If you look at teams individually, they are better or worse. Some questions people might ask about these numbers. Where closers with poor percentages ineffective? High ERA, etc? No. In some cases I had 5.00 ERA closers with 35+ saves and ended up with some of the more favorable % in that season. Some closers where in the 1.50-2.00 range and had a % barely over 50%. Did the computer manager move pitchers around, going with the hot hand? No. Everyteam was set to human manager. I set all the rotations before the season. Since injuries were off this removes the possiblity of numbers being skewed because of an injured closer. Where bad teams worse with their bullpen management than good teams? No. There was no pattern based on records. Some teams with horrible records had good closer usage while some great teams were among the worst offenders. I dont think I need to go research MLB save totals. I think its save to say that their closers dont get 66% of a teams saves. Ive researched it in the past trust me it was 90% plus ATLEAST!! I wasnt even counting the number of SAVE OPPORTUNITIES! Would have been way to time consuming to do that, but I think its safe to assume from the ERAs of some of the scrubs who ended up with an abnormal number of saves that they blew enough to make a pretty large effect on their teams win/loss record that season. I conservatively guess as many 8-10 wins. Teams that had really good bullpens were more likely to suffer with poor bullpen usage, but it was not always the case. You would think that SU1 who got 10-15 saves on average a season would be as good as or maybe better than the closer, but sadly even this wasnt the case in every situation. It was in a couple though I must admit that. Fixes? Its obvious that Im going to have to live with this. The lack of any official response has made that clear. So the only way I can see to get any semblance of statisical reality out of your closer is to place the closer in atleast Setup role 1 and possibly 2 as well. Hell maybe we should just put the closer in every role in the bullpen! A little poscript to this is my ERA settings. In my Solo league my pitcher endurance was set to normal. Because I like getting complete games about 1/3 of the time. It should be noted that changing the endurance to fair did not change the numbers at all. I dont have the statiscal data on it anymore but I simmed many seasons, atleast 25, since the release of 6.5 and after each of the beta patches to see how it would affect my league before continuing it. My totals in 6.12 using the same era settings (endurance: normal , relievers: very often) I was in the 80% range. WITHOUT using tricks like putting the closer in any other position except CLOSER. Now I would like to challenge anyone who IS NOT having these problems to post your own statisical data for your own league. I would prefer hearing from those who play like I do with my Solo, basically as a GM/God mod. Basically I play my league like an Online league would be played. So people in Online leagues feel free to post your data as well. How are you setting up your pens? What are your Era settings? I would love to get to the point you are at, If the game isnt broken then tell me what I am doing wrong. Another interesting observation I noticed when simming those last 4 seasons I hadnt noticed before. The setup1 tends to show up on the leaderboards early in the season but is then replaced as it goes by the closers. I cant confirm this but it seems as though the closer is used better as the season goes on but not in april and may. I cant back this up other than remembering seeing SU1x with 6 saves in april on the leaderboard and remembering at the end of the season he only has 8 for the season. Odd eh? Last edited by Aordolin; 12-03-2005 at 07:15 PM.

12-04-2005, 10:27 AM	#38
KurtBevacqua Hall Of Famer Join Date: May 2003 Posts: 2,968 Blog Entries: 2	Played with some more numbers, going back a bit further, but still the same current era (1980-current). 1990 AL 637 446 70.02% 1991 AL 618 443 71.68% 1992 AL 619 461 74.47% So these are even closer to 66%. Then I decided to go back a bit further into the 80's, oh, some of you are not gonna like these numbers. 1984 AL 513 310 60.43% 1985 AL 522 307 58.81% 1986 NL 480 296 61.67% 1987 NL 496 274 55.24% One could argue that for the 80's the game is actually overly using the closer. In the 21st century it appears the number is around 75%, in the 90's 70-75%, and in the 80's it was 55-65%. I think the game using closers 66% of the time is statistically a very good compromise across eras. Maybe the best thing to do is for Markus to create another era and another setting for relievers. But his current model is not broken, just a compromise. __________________ "The type and formula of most schemes of philanthropy or humanitarianism is this: A and B put their heads together to decide what C shall be made to do for D. The radical vice of all these schemes, from a sociological point of view, is that C is not allowed a voice in the matter, and his position, character, and interests, as well as the ultimate effects on society through C's interests, are entirely overlooked. I call C the Forgotten Man" - William Graham Sumner

12-02-2005, 03:05 AM	#22
Derek4jc Minors (Triple A) Join Date: May 2004 Location: Attleboro, MA Posts: 238	I'm more disappointed to have paid for the upgrade (having saves working fine previously) and then somehow it ended up worse than started. I know it was said to have been the last patch and whatnot, but to not address this 6.5 problem would be (in my opinion) a mistake and a disservice to the paying customers who were certainly caught off guard and disappointed by this setback.

12-03-2005, 12:18 AM	#28
Derek4jc Minors (Triple A) Join Date: May 2004 Location: Attleboro, MA Posts: 238	I disagree that it has to do with individual settings. I don't think I am getting less saves, I just know I am getting them spread out in poor ways. Here are some examples from my online league. (Pretty much all of them are like this but these are some) Date: 7/1 KC Royals: Mike MacDougal - 8 (setup man) (40 appearences) Byron Embry - 6 (Closer) (18 appearences) Jaime Cerda - 1 (setup man) (11 appearences) Embry has never been hurt and has always been the closer. Red Sox: Scott Linebrink - 8 (Closer) (12 Appearences?!?!) Ryan Wagner - 7 (Setup man) (30 appearences) Carmen Cali - 2 (Setup man) (20 appearences) Linebrink has never been hurt and has always been the closer. On the twins JC Romero never got hurt, and yet he's only been used 11 times. He has pitched 1 times since 5/29. The trouble is he is in a contract year and the guy needs him to have playing time so he'll he able to resign him. These are the problems. Odd nonuse and the saves being too far spread. As far as I know, there is no setting that I can change that has anything to do with how a bullpen will be used in save ops. If it wasn't an online league, I'd be able to plug the guys in myself, but there is no way on my Red Sox team i'd ever only use my closer 12 times, while totally healthy, in 3 months, and just give out those save ops to my setup guys. it makes no sense. Same thing goes with the KC setup guy having MORE saves then the closer. These results are across the board throughout the league. No a lack of saves, but an improper use of the pen by the AI, mostly in the closer position.

12-03-2005, 08:14 AM	#29
Aordolin All Star Reserve Join Date: Apr 2004 Location: Houston, TX Posts: 586	It was suggested to move endurance to fair. Mine was on normal and things did improve. Im going to run some tests this weekend, and Ill post the results. Something was changed though. Because in 6.12 things worked ok. I changed NOTHING in the settings and after installing 6.5 expansion it went to hell using same settings. After I post my results atleast we can see if they are in line with what MD and others are getting. Atleast then well know if some of us just have different expecations than others.

12-03-2005, 06:24 PM	#33
Derek4jc Minors (Triple A) Join Date: May 2004 Location: Attleboro, MA Posts: 238	Wow, I guess some people want you to jump though hoops before they're willing to believe there is an issue. Good study Aordolin.

12-03-2005, 11:06 PM	#35
Aordolin All Star Reserve Join Date: Apr 2004 Location: Houston, TX Posts: 586	Ok as suggested in another thread I have no run some seasons using SkyDogs modifiers. I used the same league as I did when doing my earlier tests. So all the other settings are the same, injuries off, human manager was on all teams, I set the rotations, lineups etc, just see my above post. I set baseball ERA to modern, recalcuated league totals for accuracy then changed everything as listed below per SkyDogs specs. http://www.ootpdevelopments.com/boar...67&postcount=1 RATINGS WEIGHT: 67% CURRENT YEAR STATS WEIGHT: 33% LAST YEAR STATS WEIGHT: 0% YEAR BEFORE LAST: 0% Batter--Contact: 9% Batter--Gap: 0% Batter--Power: 9% Batter--Eye: 22% Batter--Strikeouts: 41% Pitcher--Staff: 7% Pitcher--Control: 25% Pitcher--Movement: 25% Pitcher--Endurance: 5% Player--Speed: 2% Player--Stealing: 2% Hitter Development: +65% Hitter Aging: -54% Pitcher Development: +65% Pitcher Aging: -54% Draft Rounds: 12 Trade Frequency: Low Trading AI Preference: Favor Prospects Scouts: Off ERA SETTINGS Bunting: Often Hit & Run: Normal Stealing: Normal Pinch Hitting: Often Use Relievers: Often Pitcher Endurance: Normal Rotation: 5 Man Leadoff Lineup AI: Favor OBP CL=Closer SU1=Setup Man role 1 SU2=Setup Man role 2 MR=Total of all Middle/Mopup Reliever saves. PCT=Percentage of total saves by closer Year 1 Total saves: 622 CL: 190 SU1: 271 SU2: 134 MR: 27 PCT: 30.5% Year 2 Total saves: 672 CL: 240 SU1: 254 SU2: 141 MR: 37 PCT: 35.7% Year 3 Total saves: 649 CL: 214 SU1: 266 SU2: 139 MR: 30 PCT: 32.9% Year 4 Total saves: 686 CL: 204 SU1: 271 SU2: 178 MR: 33 PCT: 29.7% Year 5 Total saves: 718 CL: 207 SU1: 283 SU2: 188 MR: 40 PCT: 28.8% So.. as you can see its alot worse. Its as if the game has now decided to TOTALLY ignore your closer settings and use whoever it wants. These numbers though are not all that different from the ones I posted earlier though... IF.. you substitute SU1 and in some cased SU2 for the closer. Yes, I said it, the Setup men were ALL the league leaders, only twice did an actual closer finish in the top 5 in saves. You will also notice that the MR saves went down alot, they were actually pretty realistic IMO. Sadly though I do not believe this was do to any settings change, it was mostly because those pitchers under these settings did not pitch nearly as much as they did under my 11 year test run earlier. I made a mistake earlier, but I found the results kinda interesting. I used Skydogs modifiers however I had not modified my League Totals, just the ERA settings. Also I was running 6.50, the original build released last summer. So I dont know if it was that or my changes in the league Totals. However I simmed 3 years: Year 1 Total saves: 738 CL: 359 SU1: 210 SU2: 119 MR: 50 PCT: 48.6% Year 2 Total saves: 745 CL: 380 SU1: 200 SU2: 119 MR: 46 PCT: 51.0% Year 3 Total saves: 730 CL: 349 SU1: 224 SU2: 106 MR: 51 PCT: 47.8% It was after season 3 I realized I was not running 6.51 and had not recalcuated the league totals. Unless I misunderstood the insturctions, but since there wasnt anything in the post about league totals being set to a particular thing, I just went with the ERA defaults. Ok now the last thing I did was I had had the computer auto do the bullpen and rotations. It was still set to human manager I just asked the computer for advice. In some instances it changed the closer I had to one of those favorite setup men it liked to use. So I simmed 5 more seasons with that. LOL it was the same as with my roations. And the most hilarious thing was the closer that they took out of one team and put at setup 1 finally got the saves!! Yes thats right, favorite reliever#1, who the computer moved to closer got just as many saves as the previous closer did! And that old closer got all the saves that reliever got in previous sims! Brillant! Anyway excuse me for not posting those results. I am to tired and, dont feel like adding them up. But they look pretty much the same as the first set I posted from just looking at them. I would be shocked if they didnt fall in the 30-35% range. What this means for me is that I am convinced the game is broken, its not me. I just dont see how it can be. But I want to see your test results and I want your league settings and era settings and everything you can send me about your league. Hell email me your league and let me look at it. I just dont see how I can be so far off from everyone else. Now if you manage your league only and the computer manages everything else, well I think I can see why you dont have a problem. My guess is the computer manager would probably do a better job at it and create alot more realistic results. But Solo players who play like I do and ONLINE leagues especially.. how are you NOT having this same problem? IF you do play the way I do or are in an ONLINE league I want to hear from you! And if Markus reads this, please say something! At this point I guess I can understand that its not going to be fixed. But I dont want to see this repeated in the new version for sure. And I wont be buying it until I feel I have the assurances that statisical accuracy for relief pitchers is going to be a priority. Some people may not think this is a big deal but I dont think you realize how much of a bummer this is for me. It just ruins my league so much that I dont even want to play anymore.

12-04-2005, 01:56 AM	#36
ibdb Minors (Single A) Join Date: Oct 2005 Posts: 96	Thank you very much for all your work on this. I'm not crazy (at least about this) after all.

12-04-2005, 06:57 PM	#40
ibdb Minors (Single A) Join Date: Oct 2005 Posts: 96	Aordolin -- Have you noticed any difference in NL (non DH) and AL (DH) leagues? I was setting up my own test league for trial and error. After a quick two months sim to see if things seemed to be at least set up correctly, I've noticed distinct differences between non-DH and DH leagues. In a non-DH league, it's not uncommon for the save distribution to be 12-2-1 (CL-SU1-SU2). In a DH league, it's not uncommon for the save distribution to be more like 9-4-3-3-1 (CL-SU1-SU2-MR1-MR2).