Managertendencies.csv

[MLHFFYP]

This posting refers to the managertendencies.csv file which is in the stats folder. I never really looked in that folder as I am using either random debut databases or games carried forward from OOTP 11 but someone told me that it was there and it contained my research into historical and fictional managers.

This product was done for OOTP 12 and submitted to Markus before beta testing. As I heard nothing from him I did not think that he had bothered with it except for using the overall roster strategies. I had made suggestions on how to implement this project and gave documentation for its support but, again, nothing.

The intent was to address a specific issue: Several people over the last few years have asked about having “real” managers in OOTP. With the encouragement of several members of this forum I made this a project. Having access to a lot of research material, I was able to create profiles for managers so that someone could choose, if they want, to manage against figures ranging from Harry Wright to Tony LaRussa. It took a whole lot of time but I finished and gave a spreadsheet to Markus with the specific in-game strategy settings that would be used by a historical manager. The manager is on the left and the settings go from left to right in integers.

Obviously, the project did not get included in the way that would have made it more accessible for the player. The spreadsheet is in there but without any context on what it is and how to use it. Therefore, you might be able to figure out what it is trying to do but you would have to manually input every integer for every setting and check to make sure that you made no mistakes.

There are also profiles for general managers (reflected in the overall roster strategies area), which GM’s are affiliated with each manager, and other docs to give an idea of what was done and why. For example, I wrote a document with notes on certain managers to let a player know why the ratings were the way they were. In reading this you would see that the Connie Mack profile is for his great teams of the late 20’s and not from the deadball era. These too were not included with the release.

The point is that, for the most part, the result of the project has not been realized. Most people will not want to take the time to manually put in a lot of data.

Good news is that I have done this already. Each of the historical managers had been put into a file where you only have to import that strategy into a team’s strategy settings and you will have that manager. I also have the individual ratings for each associated GM so they can quickly be entered as well. Once the team’s personnel are edited you could have a team managed by Miller Huggins and its GM would be Ed Barrow. Directions for how to do this are available.

If anyone would like these files you can send me a pm and I will give you some more information and how to get them.

I use them in my leagues and they add to the enjoyment of the game and the sense of realism as well.

BTW, I did note that there are historical managers included with OOTP 12. I checked the year 1908 and looked at many of them. For several, the settings sliders are in the middle and the overall roster strategies are not correlated to those settings. Some ratings that are individualized are, in my view, not the most accurate so I don’t use these managers at all.

[akw4572]

Sounds awesome, PM sent.

[MLHFFYP]

All,

PM's have been sent.

For Dreamteams, I sent you a PM but received the following notification:

1. DreamTeams has exceeded their stored private messages quota and cannot accept further messages until they clear some space.

DreamTeams, let me know when some space is cleared and I will answer your PM.

[DreamTeams]

Space cleared. Another PM sent.

[PhillieFever]

I remember reading about this back in beta, and I believe there was an issue with the manager ratings being too "extreme" causing issues. I could be wrong but I don't think so.

[Bluenoser]

Originally Posted by PhillieFever

I remember reading about this back in beta, and I believe there was an issue with the manager ratings being too "extreme" causing issues. I could be wrong but I don't think so.

Care to elaborate?

[MLHFFYP]

Originally Posted by PhillieFever

I remember reading about this back in beta, and I believe there was an issue with the manager ratings being too "extreme" causing issues. I could be wrong but I don't think so.

Being a member of the beta team also I posted updates on the project under historicals in the beta forum. The messages are still there if you want to check them out.

What I believe you are referring to is the posting from Charley Hough who was using some profiles I had sent him.

He wrote the following:

"Thanks for all of your work. I've done some testing with strategy and roster settings based on the ideas you shared in the historical forum, and I've seen some interesting results. Unfortunately, there still seems to be significant randomness involved, so even with extreme slider settings, teams often make trades that don't necessarily match their overall strategy. But that may be due to urgent roster needs or the ability to arrange the right kind of deal with another AI club.

Anyway, as far as suggestions go, here are a few. By the way, I think everyone in the Hall of Fame as a manager should be included. You've covered most of them already. Also, adding some Negro League managers might be an idea worth considering, although many of the most well-known were player/managers.

Billy Martin
Bucky Harris (HOF, player/manager early on, later a full-time manager)
Al Lopez (HOF)
Wilbert Robinson (HOF)
Dick Williams
Clark Griffith (HOF, mostly a player/manager)
Gene Mauch
Ralph Hauk
Chuck Tanner
Joe Torre
Lou Piniella
Danny Murtaugh
Charlie Grimm
Tom Kelly
Bobby Valentine"

As you can see the word extreme was used in reference to the AI making trades that do not match the overall roster strategies that are set. Of course, this is the case with the default overall roster strategies that are included with the game.

The manager ratings are not "too extreme". They are not even ratings. There are in-game strategy settings that are meant to "manage" the game in a certain way. Do you bunt more in the fifth inning as opposed to the ninth? Do you call for the hit and run more in the second inning as opposed to the first? By adjusting these sliders you can effectively create a manager by having the in game settings adjusted to the way he actually ran a game. So you change every setting for every strategy one-by-one and then export the file for later use. This way you can make a manager. It is not that easy to do, obviously. Having settings for John McGraw in the Deadball era will be different for his settings in the years after the Deadball era. His New York Giants won two World Series in both 1921 and 1922 while they won in 1905. The way he ran the game in 1905 and 1921/22 was different and these can be reflected if the settings are adjusted accordingly.

The overall roster strategies mirror a General Manager. If you know that McGraw preferred veteran players during the years that you have his in-game settings set up for then you can adjust the overall roster strategy sliders moved to the prefer veterans to the degree you want. In asking the question in the forum a long while back it was mentioned that the overall roster strategies are used to help forumulate the way drafting and trading are done by the AI.

Therefore, if your overall roster strategies (GM) are set to complement the prefererences for the in-game strategies (the manager you created or are using) then you can say you have a more accurate representation of that team.

There is randomness involved. Not every situation will be handled exactly the same every time. This is how it should be. However, by fashioning the AI preferences a person can assist the AI in making better decisions and make the game more historical to boot.

I have used the manager profiles for thousands of games and they do make a difference. The overall roster strategies (GM) make a difference also but they are more long term and less visible.

If you could, PhillieFever, let me know what specific reference in the beta forum and what issues it caused. Again, the only one that I could find that remotely is similar to what you wrote ("extreme") was not used in that context as I have explained and shown above.

By the way, in one of my Random Debut Leagues Gene Mauch's Phillies with John Quinn as the GM finally won a World Series. As a lifelong Phillies fan I thought it was long overdue and was glad to read the news story about him and the team. He also was named Manager of the Year so it was vindication for him via OOTP.

[PhillieFever]

I was looking earlier today, but I couldn't find the post in which Markus had made a comment that the managerial tendencies "tended" to be extreme, meaning lots of "10's and 0's. In the context of the game, the 10's and 0's don't work out very well, and they were causing problems. A 10 would be the SB slider maxed out a zero would be blanked out if you get what I'm saying. In order for the tendencies to work well in the game, they really should remain between 3 and 7 with really extreme cases occasionally getting a 2 or an 8.

[pstrickert]

http://www.ootpdevelopments.com/boar...p?issueid=2366

[thehef]

I would think that players' stats - and thus their ratings - are somewhat reflective of manager's tendencies, and that manager's ratings are based upon tendencies, which are reflective of player's statistics.

So - in this over-simplified example - if Alston, Howser, or Herzog is rated to favor stolen bases partially because Maury Wills or Willie Wilson or Vince Coleman stole a boat-load of bases under that manager's watch, but Wills, Willson or Colemen are already rated by the game to steal a realistic/accurate number of bases, then by pairing Wills, Wilson, or Coleman with Alston, Howser, or Herzog, won't you end up with combined player & manager ratings that would over-emphasize steals?

I'm probably missing some of the nitty-gritty that 'splains this, but I just thought I'd throw it out there because I was wondering about it... ;-)

[MLHFFYP]

Originally Posted by PhillieFever

I was looking earlier today, but I couldn't find the post in which Markus had made a comment that the managerial tendencies "tended" to be extreme, meaning lots of "10's and 0's. In the context of the game, the 10's and 0's don't work out very well, and they were causing problems. A 10 would be the SB slider maxed out a zero would be blanked out if you get what I'm saying. In order for the tendencies to work well in the game, they really should remain between 3 and 7 with really extreme cases occasionally getting a 2 or an 8.

I read your post with bewilderment. I have looked everywhere on the forum since I saw your first post and have found nothing to confirm what you posted then and I have looked again since this post and, again, found nothing. Unless you could show me the post then I cannot buy into what you are saying.

Perhaps Markus can confirm or deny this assertion as well. It makes no sense to put out a game with tools (sliders) and then have them work so poorly that they cannot be used as intended. From the probability standpoint they make absolute sense. Because there is a random factor thrown in you will never have absolutes. While a manager setting may be on the far left (never) for hit and run in the first inning with 1 out, there will always be a chance that he will call it due to that random factor. To say that they should be between 3 and 7 with the rare exceptions of a 2 and 8 again makes no sense. You are essentially making the idea of using the sliders to make a more competitive AI a moot one. After all, this is OOTP 12 and not OOTP 2 or 3. I simply can't believe that after 12 versions of the game and countless patches that this could be the case.

I have played thousands of games against the computer managers I built and have been very happy with the results. On a rare occasion I will tweak a setting but that is more for accuracy based on what I have researched or because I might have seen the AI do something that I find strange. But again, that is the rare exception.

I started a historical league based on 1915 and checked some of the managers included with the game. Looking at Pat Moran I found that his strategy settings had 5 out of the 14 set to the extreme right. On the strategy settings for the team (the in-game strategy settings) I found six out of the 20 set to the extreme right for the inning 1-3 tied ballgame situation. And this is just for that situation; the many, many others have the same thing as well. Based on what you are saying and what is in the game, anyone playing the 1915 season should have major problems. I am sure this would apply to other seasons as well.

By the way, these are not my settings either. I have a Pat Moran built that is, in my view, more accurate than what is in the game but it shows that the release used the far left and right settings.

So to sum up I would ask you to cite where you are getting your information and, perhaps, relay what testing you have done with the results. What are the problems? How often did they occur? What version had the problems? Did Markus address it or let it go, thereby making the game less than what it should be? It may be that you are right but that would fail to explain why Markus would allow this to occur with his release managers and also why I have not noticed any significant problems with mine.

As someone who is working to make OOTP better I appreciate any input. However, if there is assertions that aren't backed up then I can't make corrections to make the project better.

[MLHFFYP]

Originally Posted by thehef

I would think that players' stats - and thus their ratings - are somewhat reflective of manager's tendencies, and that manager's ratings are based upon tendencies, which are reflective of player's statistics.

So - in this over-simplified example - if Alston, Howser, or Herzog is rated to favor stolen bases partially because Maury Wills or Willie Wilson or Vince Coleman stole a boat-load of bases under that manager's watch, but Wills, Willson or Colemen are already rated by the game to steal a realistic/accurate number of bases, then by pairing Wills, Wilson, or Coleman with Alston, Howser, or Herzog, won't you end up with combined player & manager ratings that would over-emphasize steals?

I'm probably missing some of the nitty-gritty that 'splains this, but I just thought I'd throw it out there because I was wondering about it... ;-)

You have a good point which goes to the heart of every stat based game in existence.

However, I think what the tendencies do is correlate a number of factors that end up with a decision. For example, I believe that in the case of a manager who likes his players to steal bases (Ned Hanlon) his inclination in the game is to send players with the right ratings more often than those who don't. I use Hanlon because in a random league I had going he had Rickey Henderson on his team and Henderson ran a lot. It was frustrating in a way playing against Hanlon's team because I knew if Rickey got on he was going unless the game situation made that too risky.

Conversely, I had Wills on a team managed by Miller Huggins and while he did steal bases he did not have as many steals as he did in real life.

This, I think, is the key to using the sliders. Making the settings such that the manager created will follow his real life tendencies and then see how that correlates with the players on his team. Connie Mack managed differently in 1911 than he did in 1930. This is due to many factors. So a Maury Wills would have loved playing for him more in 1911 than 1930.

So I believe that the tendencies of an Alston would complement Wills and not pile on his proclivity to steal. Alston with a slider set to frequently stealing bases would, I think, send a great base stealer like Wills often.

[PhillieFever]

Originally Posted by MLHFFYP

I read your post with bewilderment. I have looked everywhere on the forum since I saw your first post and have found nothing to confirm what you posted then and I have looked again since this post and, again, found nothing. Unless you could show me the post then I cannot buy into what you are saying.

Perhaps Markus can confirm or deny this assertion as well. It makes no sense to put out a game with tools (sliders) and then have them work so poorly that they cannot be used as intended. From the probability standpoint they make absolute sense. Because there is a random factor thrown in you will never have absolutes. While a manager setting may be on the far left (never) for hit and run in the first inning with 1 out, there will always be a chance that he will call it due to that random factor. To say that they should be between 3 and 7 with the rare exceptions of a 2 and 8 again makes no sense. You are essentially making the idea of using the sliders to make a more competitive AI a moot one. After all, this is OOTP 12 and not OOTP 2 or 3. I simply can't believe that after 12 versions of the game and countless patches that this could be the case.

I have played thousands of games against the computer managers I built and have been very happy with the results. On a rare occasion I will tweak a setting but that is more for accuracy based on what I have researched or because I might have seen the AI do something that I find strange. But again, that is the rare exception.

I started a historical league based on 1915 and checked some of the managers included with the game. Looking at Pat Moran I found that his strategy settings had 5 out of the 14 set to the extreme right. On the strategy settings for the team (the in-game strategy settings) I found six out of the 20 set to the extreme right for the inning 1-3 tied ballgame situation. And this is just for that situation; the many, many others have the same thing as well. Based on what you are saying and what is in the game, anyone playing the 1915 season should have major problems. I am sure this would apply to other seasons as well.

By the way, these are not my settings either. I have a Pat Moran built that is, in my view, more accurate than what is in the game but it shows that the release used the far left and right settings.

So to sum up I would ask you to cite where you are getting your information and, perhaps, relay what testing you have done with the results. What are the problems? How often did they occur? What version had the problems? Did Markus address it or let it go, thereby making the game less than what it should be? It may be that you are right but that would fail to explain why Markus would allow this to occur with his release managers and also why I have not noticed any significant problems with mine.

As someone who is working to make OOTP better I appreciate any input. However, if there is assertions that aren't backed up then I can't make corrections to make the project better.

PStrickert posted a link to the conversation above.

[PhillieFever]

Originally Posted by MLHFFYP

I read your post with bewilderment. I have looked everywhere on the forum since I saw your first post and have found nothing to confirm what you posted then and I have looked again since this post and, again, found nothing. Unless you could show me the post then I cannot buy into what you are saying.

Perhaps Markus can confirm or deny this assertion as well. It makes no sense to put out a game with tools (sliders) and then have them work so poorly that they cannot be used as intended. From the probability standpoint they make absolute sense. Because there is a random factor thrown in you will never have absolutes. While a manager setting may be on the far left (never) for hit and run in the first inning with 1 out, there will always be a chance that he will call it due to that random factor. To say that they should be between 3 and 7 with the rare exceptions of a 2 and 8 again makes no sense. You are essentially making the idea of using the sliders to make a more competitive AI a moot one. After all, this is OOTP 12 and not OOTP 2 or 3. I simply can't believe that after 12 versions of the game and countless patches that this could be the case.

I have played thousands of games against the computer managers I built and have been very happy with the results. On a rare occasion I will tweak a setting but that is more for accuracy based on what I have researched or because I might have seen the AI do something that I find strange. But again, that is the rare exception.

I started a historical league based on 1915 and checked some of the managers included with the game. Looking at Pat Moran I found that his strategy settings had 5 out of the 14 set to the extreme right. On the strategy settings for the team (the in-game strategy settings) I found six out of the 20 set to the extreme right for the inning 1-3 tied ballgame situation. And this is just for that situation; the many, many others have the same thing as well. Based on what you are saying and what is in the game, anyone playing the 1915 season should have major problems. I am sure this would apply to other seasons as well.

By the way, these are not my settings either. I have a Pat Moran built that is, in my view, more accurate than what is in the game but it shows that the release used the far left and right settings.

So to sum up I would ask you to cite where you are getting your information and, perhaps, relay what testing you have done with the results. What are the problems? How often did they occur? What version had the problems? Did Markus address it or let it go, thereby making the game less than what it should be? It may be that you are right but that would fail to explain why Markus would allow this to occur with his release managers and also why I have not noticed any significant problems with mine.

As someone who is working to make OOTP better I appreciate any input. However, if there is assertions that aren't backed up then I can't make corrections to make the project better.

Back in beta for OOTP 10 or 11, forum member SteveP I believe was doing some work with sliders in an attempt to get more accurate league totals. He found that the extreme settings were TOO extreme, my memory can be a little hazy at times, but I'm almost positive about his results. As far as what you're attempting to do, I give you all the credit in the world. Attempting to create accurate managerial tendencies considering the multitude of options is quite daunting, especially when you consider how baseball has changed over the years. I'm sure you've tested this more than anyone, and if you say that you're getting great results than I'm going to defer to you because God knows I don't have the attention span to complete one manager let alone an entire history LOL.

[Bluenoser]

Originally Posted by PhillieFever

PStrickert posted a link to the conversation above.

The link is no good unless you were on the Beta Team. Forum blocked.

[MLHFFYP]

What started out to be a project to help the community by providing more realism for OOTP has turned, in some ways, into a defense of the project itself. This is quite frustrating. So I will again try to set some topics to rest.

1. The thread that is referred to had a response from Markus where he said that if you set a slider all the way to the right or left then that strategy would not happen or go to the opposite extreme. I am running a random debut league where Harry Wright is the manager of the Cleveland Indians. For those who have his profile you will see that his ratings call for leaving the starter and reliever in as long as possible. His ratings are at the extreme end in this regard. His Cleveland Indians have played 151 games and 25 of them have been completed by the starter. If it were true that the extreme right or left means that nothing or everything will happen then all 151 games would have been completed by the starter. I have other leagues where Harry Wright is the manager and not every starter completes every game.

2. The documentation for the profile states that Harry Wright is clear as to what his profile is expected to do. It states:
Harry Wright

Harry Wright can be considered the first manager. He was the manager of the famous Cincinnati Red Stockings who gained fame by winning fifty seven games and tying one during the 1869 season. After leaving Cincinnati he continued managing with the Boston Red Stockings where he also had notable success with his team. After stints with the Providence Grays and Philadelphia Quakers he retired in 1893. Wright was inducted into the Hall of Fame in 1953.

Wright is not a baseball strategist. He is a proponent of strong pitching and defense and heavy hitters. Won’t sacrifice often or use the hit and run. Likes aggressive baserunners. Depends on pitchers going the distance but will use relievers on rare occasions.

In every hook rating in his profile for every situation he has a ten. Again, in my leagues he has not made his starter go the distance every time.

3. The Harry Wright example actually proves, in my opinion, the strength of OOTP. In 1869 he used Asa Brainard as his main starter. Brainard worked 338 innings and Harry Wright was the "change pitcher" working 118 innings and George Wright worked 14 innings. His SABR biography says that Harry did not pitch because Brainard was tired but because he wanted to disrupt opposing batters. Obviously, Harry depended on Brainard as the main pitcher. Now OOTP with the Harry Wright profile does use relievers so it shows that, even with the slowest hook settings, it will evaluate the situation and taking all factors into account, will make decisions. As random debut leagues are tougher to manager, I believe, I think the profile works very well.

4. Again, managers included with OOTP do have the extreme right and left settings. They are they if people want to check. I gave an example so please feel free to look and see that this is true. Please look at Pat Moran in 1915. This is mentioned in a previous post of mine.

5. I am being redundant but I am saying it again; I have played a few thousand games individually against the profiles and have not noticed significant problems.

6. I think that the in-game strategy settings are meant to be a way for a "manager" to run the game with other factors thrown in. This is consistent with what I have seen in other baseball sims. They do make a difference from what I have seen but they do not seem to be iron clad. There seems to be a randomness to them which is a good thing. Think tendencies and not absolutes and I think it makes sense.

7. Regarding the discussion on the beta board I was not aware of it. I posted everything under historical so I expected that any comments would be directed there.

8. Neither Markus nor Andreas ever contacted me with comments on the project including any possible issues with sliders. Because I had already played a large amount of games with most of the profiles I could shown what I had seen, specifically something like what I showed above with Harry Wright. If there were concerns I would have addressed them like, well, I thought we were supposed to do on the Beta Team. In fact, I could have changed everything to a 1 or a 9 even if I did not see any problem myself. My intent was to get something out to the OOTP community so it puzzles me a bit why I did not get a note on this issue.

9. Likewise, no thanks was given for submitting all of the data, which took a whole heckuva lot of time to research, gather, and put in the required format. None. In the game itself under acknowledgements you won't find my name even though the overall roster strategies were included. To be fair, though, I could be one of the ...and many more so that in itself is not a big deal. PStrickert was the person giving me encouragement so thanks go to him from me.

10. I have sent out the profiles and all documentation to a great many people. I have stated that if there are issues or problems I will gladly make changes. All I ask is that it be brought to my attention so that I can make adjustments along with the rationale for changes. So far so good in that I am getting some very positive comments. Glad it is being used and, hopefully, it will add to the enjoyment of OOTP.

There you have it; my top ten. I ask that if anyone wants the profiles pm me and I can make that work. If not then continue to play the game as it is a great product.

[loyalroyal]

Can someone email me a copy of this?

[MLHFFYP]

This is for those of you who have been sent the manager/general manager profiles and information.

I have completed all of the fictional managers so they will be sent soon.

Some may be aware of the issue of whether problems are caused by the profiles. A few things:

1. I have been told that the concern that having 0's and 10's might cause a crash. When I got this information I immediately ran several seasons with all of the managers as presently configured and had no problems. This is in addition to the thousands of manual games that I have played with no problems either.

2. I have not received any negative input from anyone using the profiles so I am fairly confident (knock on wood) that the possible problem is not occurring.

3. However, to be absolutely safe I am redoing every single manager, historical and fictional, to move any slider that is all the way to the left or right one space to the left or right. I will send these out as a group when I am done, which will be fairly soon. Again, I have had no problems but if someone does then let me know.

4. I am reworking the Earl Weaver profile because of an error where all of the sliders for a inning/situation are set in the middle.

5. Have had some requests for specific managers and I will try and work on them as time permits. Two I want to try on the fictional side is a Negro League manager for the 1930's and one that would be representative of the teams of the early 1900's. This would be quite different than the historical Rube Foster profile that I have already made.

Thanks to everyone for the nice comments. They are appreciated.

[pstrickert]

Please add me to your e-mail list, Bob. I'd like to receive your updates. Thanks!

[DreamTeams]

Looking forward to the fictional managers, Bob. If possible could you please shoot me the historical GM profiles as I never received those. Thanks.