AI Minor Leage Madness

[frosticle]

May I make a suggestion? How about an added category of some sort that would revolve around readiness for the majors.

Take for example the 1999 draft, where the top three picks were Josh Hamilton, Josh Beckett, and Eric Munson.

The story behind these three picks were that they were all in different stages. Hamilton would need a few minor league seasons before being ready to go pro (forget about the suspensions :P), Beckett would need some minor league time and would end up with probably less talent overall than Hamilton, while Munson would be the least talented of the three, but was thought to be ready for the majors right off.

I think it would be interesting if you could get the program to factor in how ready a young player is for the big show.Too many times on this game you see a young pitcher just come up and blow everyone away while he is 18, which is too rare in real life. You could have him be really talented, but make it so that he isn't ready for the majors yet, so if he comes up he might struggle and you could even see his ratings drop.

On the other hand perhaps you find someone in the draft with slightly less talent, but he is already major league ready.

Lets say two pitchers. Both starters. One is 81/97 and the other is 82/83. You could have them want you to keep that 81/97 at AAA for a year or something, while the 82/83 guy is ready to step into the spotlight.

In a way you could tie it in with your system, like have players of a certain readiness who are a certain overall be auto sorted to AAA, while guys who aren't ready, even if they have the same rating may go even lower and guys who are more ready may be promoted.

You could then have the readiness be random. For example you could call up that 81/97 guy and have him pitch well sometimes and all of a sudden you watch his readiness go up, while at the same time you might see him struggle and you can watch his potential and overall ratings fall off because they used him too soon.

Anyway, I don't know.. Just a thought.

[SFSteveG]

Interesting. Sounds like a lot of work but it might be something I'd like to see if if it works well. I'd just hate to see a really good guy that has done well never be MLB ready.

[dolfanar]

I think it's a great concept in theory, but to get BBM to start making "judgment calls" on players you'd have to completely remake the player rating system. As it is there are concrete ways to know if a player is "ready", mainly having to do with overall ratings and career peaks. There isn't any instinct involved, it's pure math (like most of BBM). If rookie A has a rating > veteran B, then rookie A is ready. Now these ratings can be off if your scouting isn't good, so maybe that's where instinct comes in, but the AI basically has to deal with the numbers it sees.

The issues with BBM playing players too early or at higher or lower levels than appropriate all come down to interpretation of the numbers. As shown by examples above, the game is making strange calls as to where to send players. I am still seing 24 year olds in Rookie ball for instance, and players are still getting called up from Rokkie and A-ball for lengthy major league stints. Solve those problems and the rest will very quickly fall into place.

[frosticle]

Originally posted by dolfanar
I think it's a great concept in theory, but to get BBM to start making "judgment calls" on players you'd have to completely remake the player rating system. As it is there are concrete ways to know if a player is "ready", mainly having to do with overall ratings and career peaks. There isn't any instinct involved, it's pure math (like most of BBM). If rookie A has a rating > veteran B, then rookie A is ready. Now these ratings can be off if your scouting isn't good, so maybe that's where instinct comes in, but the AI basically has to deal with the numbers it sees.

The issues with BBM playing players too early or at higher or lower levels than appropriate all come down to interpretation of the numbers. As shown by examples above, the game is making strange calls as to where to send players. I am still seing 24 year olds in Rookie ball for instance, and players are still getting called up from Rokkie and A-ball for lengthy major league stints. Solve those problems and the rest will very quickly fall into place.

I see your example and agree that there needs to be fixes, unfortunately your example does not take into account phenoms or anyone who may never make it out of the lower leagues.

The thing is that you can have all the talent in the world, but may not be able to handle the major leagues.

Now Steve, I know that it would hurt to draft someone with a ton of potential and watch them be a career minor leaguer at best, but isn't that what like half of the first round of like every draft is anyway, at least. Let alone the rest of the draft has the potential to be good or bad in real life.

I would love to see some guy on the game taken in the 6th round, 60 overall, readiness of 100. All of a sudden you need someone for the bench, bring him up and he gets up to a 70, even when his potential didn't show it.

Dolfanar, simply put, I don't believe that sticking guys around simply by number and age is enough. Your minor league system hurts the major leagues since you don't not account for anything really. All you have are numbers and where they should be placed, when that simply isn't realistic without accounting for what I said.

Some guys will be 19 and ready for the majors, while some guys will have to sit in the minors until they are 26 before they should ever get the call.

I honestly love what you are trying to say, but I think the game needs to account for young players possibly not being ready to play, as well as what you are saying.

You did say right now it's pure math, which is what your minor league idea was. Combine it with something variable like my idea and it would work better, at least I think.

It's variable in two ways. I mean like I said there would be the initial rating, but the second way would be that they could play good or bad despite that. I mean how many times have teams had to rely on players before they wanted to and had them up end being stars.

I don't think it will all fall into place if we just do what you recommended, simply because I do that anyway in about 20 seconds of sorting things manually.

In short. I am a fan of variablity because it's real. you never know how good your guys will be year to year in real life, so why should you on baseball mogul. The best you can hope for is playing them at the level they are ready for. Some great players, like Josh Hamilton would have to play many years in the minor leagues to ever become a true star, while some players like Aundrew Jones rise to the occasion very young.

[SFSteveG]

I'm only going to address the point responce directed at me because i think others will be able to more properly get into the meat and bones of those issues.

The facts of the game are as stands; there are 6 draft rounds as opposed to 50 IRL. In BBM 9 out of 10 1st round picks should eventually make the majors and at least be competition for the bench/back of the pen.

I'm not sure about how many players per draft, per team make the majors past a reasonable ammount of time (5 years maybe) but I'm sure that at least 9 out of 10 times at least one of the first five picks sees time.

[robinhoodnik]

I usually just look at stats. Avg. and doubles/ home runs for hitters and if the fielding % is respectable I'll bring 'em up. With pitchers I'll look for an era of around 4.50 or under at AAA level, and strikeouts vs. innings pitched, and bases on balls if the era is up a bit . I try not to rely on the ratings too much as a player will do well at a higher level frequently even if he is struggling a bit at a lower level.

[dolfanar]

Originally posted by frosticle
snip

Frosticle, as SFsteve mentioned the 6 draft rounds, this is CRITICAL to understanding the mogul concept. Quick and dirty. 6 rounds, and near every player is a potential major leaguer... you need only play historical scenarios to see the concentration of true major leaguers you can stuff into 6 rounds.

Again the whole touchy feely "is he ready mentally" aspect of the game is redundant in Mogul. A player isn't rated based on "talent", he is rated based on "what he can do". That inherantly involves mental and physical aspects. When a pitcher is predicted to have 30 saves, that isn't a physical characteristic, that's simply what he is capable of mentally and physically. That's one of the reasons at the major league level Mogul does so well generating statistics for the most part. It doesn't pretend to be High Heat and rate pitchers fastballs, it just sticks to the stats. The only margin for error is how good your scouting staff is at judging how much a player is capable of.

Right now the minor leagues in BBM, isn't even really a minor league system with teams, it's a holding bucket for players. There aren't "games" being played, only random statistics being generated. Trying to determine a players "mental readiness" or any such elements when games aren't even being played is throwing another wildcard into a system that is really only there to create the illusion of a minor league system. It's completely unnecessary and would only serve to undermine the statistical predictiveness of the engine.

All you need in Mogul is a simple setup that allows the AI to bring people along slowly and build up logical statistics that look pretty. Nothing fancy.

Understand that this thread isn't about straightjacketing players on how THEY run their teams. You want to call up all your 18 year olds? Great, go for it. But the AI has to atleast appear to be a competent Major League GM, and the average debut ages and time spent at various levels has to look right, or the game loses credibility. The system I described wasn't something I pulled out of a hat. It was obtained by a close study of BBM's rating system, and would allow for a more realistic average debut age (and eliminate players regularly jumping AA and AAA), while still allowing for the very rare 18 or 19 year older to play in the majors.

[seanpatin]

I think a nice solution might be to make separate lineups for each minor league team and be able to auto-sort them so there isn't as much hassle but to be able to look over box scores at least from each minor league game and also, the AI needs to be fixed so they aren't just randomly moving players from league to league... like dolf said, have occasional super prospects who skip those minor league levels for the most part, but have the AI progress players through the levels instead of just randomly inserting them wherever they feel like....

[Clay Dreslough]

The newest patch pretty closely matches Dolfanar's suggestions. I'll get it posted here soon.

Clay

[dolfanar]

Cool. If it helps for the future, I ran a 50 year sim in 8.25 and took a hard look at how my assumptions would work long-term. I tweaked my initial setup to the following. I understand that you've probably modified my assumptions already, but figured you could use this to compare against.

Age 16
All players go to R

Age 17
Overall rating <= 75 go to R
Overall rating > 75 go to A

Age 18
Overall rating <= 70 go to R
Overall rating > 70 OR <= 80 go to A
Overall rating > 80 go to AA

Age 19
Overall rating <= 65 go to R
Overall rating > 65 OR <= 75 go to A
Overall rating > 75 OR <= 85 go to AA
Overall rating > 85 go to AAA

Age 20
Overall rating <= 60 go to R
Overall rating > 60 OR <= 70 go to A
Overall rating > 70 OR <= 80 go to AA
Overall rating > 80 go to AAA

Age 21
Overall rating <= 55 go to R
Overall rating > 55 OR <= 65 go to A
Overall rating > 65 OR <= 75 go to AA
Overall rating > 75 go to AAA

Age 22-23
Overall rating <= 60 go to A
Overall rating > 60 OR <= 70 go to AA
Overall rating > 70 go to AAA

Age 24-25
Overall rating <= 55 go to A
Overall rating > 55 OR <= 65 go to AA
Overall rating > 65 go to AAA

Age 26-27
Overall rating <= 60 go to AA
Overall rating > 60 go to AAA

Age 28-30
All players 28-30 should go to AAA

Age 31+
Overall rating > 60 go to AAA
Overall rating <= 60 Cut & Retire (In retrospect Sky Dog has a point, and at this point the player is toast, and I've seen way too many of these guy's after a few years clogging up the minor league arteries or worse ex-Major Leaguers who hang on for years past anything logical or realistic)


This in conjunction with the AI choosing only AAA level players (when available) for Major League call-ups and the "25 rule" where the AI would cut any player beyond 14 batters and 11 pitchers at the AAA level based on a age-priority basis as mentioned above.

[polarisslbm]

Dolf, in your table, at which point (ratings wise) would you call up players to the majors?

[dolfanar]

Originally posted by polarisslbm
Dolf, in your table, at which point (ratings wise) would you call up players to the majors?

Again this is AI only.

The idea is that the AI would only look at players who are atleast AAA caliber when choosing it's major league lineup. Even if that means playing players out of position. this is to stop the AI from calling up players straight out or Rookie and A ball. IF the team runs out of AAA players, then it could dip into AA, and then if no more AA, then A, and so on.

So in theory any player could be called up, but only a truly depleted team would ever go much past AA for a player. This would push the average ML debut age up to about 22-23, where it should be.

[henry296]

Interesting result of this sysytem. Since all older players are in AAA, the Yankees have 47 players in AAA at the end of the 2006 season including Giambi, Mussina, Pavano, Lofton, Mike Hampton.

Todd

[dolfanar]

System hasn't been fully put in place yet. See Clay's post ("The newest patch pretty closely matches Dolfanar's suggestions. I'll get it posted here soon").

[henry296]

Originally posted by dolfanar
System hasn't been fully put in place yet. See Clay's post ("The newest patch pretty closely matches Dolfanar's suggestions. I'll get it posted here soon").

Misintpreted his comments when he post 8.25... I'll call that feedback for 8.26.