September call-ups - needs tuning

[BleacherBum]

I've played a couple of seasons now on 1.66, and I'm concerned about the way the AI is using the new 40-man roster feature. Every computer managed team in my league filled its roster to the maximum 40 on the first day possible. What could they possibly plan on doing with all these players? If you check how this option is used by real MLB teams, you'll find that most teams only call up a handful of players in September. Especially teams in a pennant race, since they can't afford to run auditions when they are fighting to win a championship.

In my current Puresim season, I'm in the pennant race, and my primary competition is now using inferior minor league players, both hitters and pitchers. I checked that these players are accumulating ABs and IPs, at times when games are in doubt, and that these guys just aren't major league caliber yet. While its certainly going to make it easier for me to claim the title, I'd prefer they put their best players on the field.

Is there a way to shut this off (an option) until it gets programmed properly with respect to the AI?

[puresimmer]

While it does usually take the roster up to 40, the AI should not be replacing top major leaguers with inferior minor leaguers unless the players are really tired or a game gets out of hand.

Can you post a sample game log or something. It might help me figure out what I am missing.

1.66 is considered BETA, so I certainly am open to tweaking this AI, I'm sure there is tuning to be done.

Shaun

[motnahp]

I touched on this a bit in the Public Beta area, under the topic of "Taking one for the Team". I did not have any data to back up my fears, but apparently you do. IMO, when the seldom-used players were (mostly) made into 5-point scrubs (on the 1-100 scale), it opened up some cans of worms. When pitchers with "5 Stuff" and "5 Control" take the mound, a series of bad things happen.

It's feasible that a 5-point batter could be badly overmatched by MLB pitching, but his ineptitude does not scream out at you in a boxscore the way a 5-point pitcher's poor performance does. It doesn't take too many instances of a pitcher allowing 15 runs in an inning, walking 10-15 men, for realism to vanish.

Batters, too, can cause problems for a manager. For my 2006 Braves, I have Scott Thorman on the roster as a bench player. In his rookie season of 2006, he batted 128 times and hit .234 with 5 homers. Since he imported as a 5-point scrub, he really has no chance of getting close to his real-life stats. For my assn., he went hitless in his first 16 at-bats, before getting a single. Do I play him more often and hope his statistics will normalize? The words, "Strike Three.....Yerrrr Out" ring in my ears whenever I consider him as a pinch hitter. I believe he struck out at least 10 of his first 16 ABs.

My point is that both the pitchers and batters need some tweaking, with the emphasis placed on the pitchers, since their performances as 5-point "scrubs" can be downright frightful and association-poisoning.

[BleacherBum]

I'll pull together some examples and post later.

However, I'm curious if you agree that 40 men on a big league bench is unrealistic as compared to how MLB teams operate?

Also, is there some easy way, either XML or database setting, to turn this off until it gets straightened out? My recent division title by 1 game is a hollow victory.

[DonBraswell]

My concern is that if a player comes into the league with 5 point ratings, he never developes into a Major Leaguer. In the 60's alot of players ( Maris, Colivato, Killebrew.etc) came up as 5 pointers and never improved until I tinkered with their ratings. Gentile and pitchers like Score, Koufax and others. Never perform close to their real abilities. It seems that players who were marginal to begin with never develop. Maybe Real Major Leaguers should import with higher ratings, 30's to 40's. I don't know how to fix this. I also see the AI promoting, drafting and using players I wouldn't even consider. I guess it's me.
Don

[Beach23BoyP]

I now don't know what feature of PS I like the "least" ......

Animated ball flight or Sept call-ups?

[KG Erwin]

Funny -- several guys kept clamoring for September call-ups, without fully understanding the potential drawbacks.

I'd just as soon have an option to turn it off, and I would turn it off.

[motnahp]

I considered adding this same thought to my earlier post. As players import, they start with the "baseline" they're imported with. If that baseline happens to be "5-point scrub", they just won't develop their true talents.

I believe that the program change which made the seldom-used player a 5-point scrub to begin with must be the culprit. Reverting back to the way it was (a year ago?) might be the way to go toward solving this.

FYI, I am all for the September call-ups and would love to see some form of this retained.

[Beach23BoyP]

I wish that both the animated ball flight and Sept call-ups were optional.

[BleacherBum]

ORIGINAL: KG Erwin
Funny -- several guys kept clamoring for September call-ups, without fully understanding the potential drawbacks.

Every new gameplay option raises a question: Will the AI handle it well? Or be confounded by it?

What's interesting is that I don't see how I would use it myself, since I sim most of my games. Putting the extra players on my roster just results in my "field manager" making more decisions that I disagree with.

[puresimmer]

I think you guys will LOVE this feature in the next patch. BleacherBum had some really good points. I went back through that AI and made some serious tweaks. Here is the note from the readme in the forthcoming 1.67 update:

* AI - Improved September call-up AI. Teams will no longer just fill the majors with 40 players, some will maybe only bring a very few players. A key factor the AI will consider here is where it is in the standings. Teams that are hopelessly out of the race will be MUCH more likely to bring 5-10 players whereas teams in the thick of a pennant race will be super judicious around call ups.

Thanks for the feedback!

[puresimmer]

ORIGINAL: puresimmer

I think you guys will LOVE this feature in the next patch. BleacherBum had some really good points. I went back through that AI and made some serious tweaks. Here is the note from the readme in the forthcoming 1.67 update:

* AI - Improved September call-up AI. Teams will no longer just fill the majors with 40 players, some will maybe only bring a very few players. A key factor the AI will consider here is where it is in the standings. Teams that are hopelessly out of the race will be MUCH more likely to bring 5-10 players whereas teams in the thick of a pennant race will be super judicious around call ups.

Thanks for the feedback!

One thing I should point out. This AI won't kick in until the next time you pass 9/1. Meaning if you are already in mid-September with your current 1.66 association, it will stick with whatever it currently has on the Major league roster.

[LetsPlay2]

Now that sounds good !!
 
Cool !!

[BleacherBum]

ORIGINAL: puresimmer
* AI - Improved September call-up AI. Teams will no longer just fill the majors with 40 players, some will maybe only bring a very few players. A key factor the AI will consider here is where it is in the standings. Teams that are hopelessly out of the race will be MUCH more likely to bring 5-10 players whereas teams in the thick of a pennant race will be super judicious around call ups.

Thanks for the feedback!

I applaud the quick turnaround on this - thanks Shaun! This will at least reduce the number of guys who are sitting on the bench for no reason.

I'm pulling together specific data, but what I found generally is that:
1) Many of the 15 call-ups per team are middle relievers, most who didn't play at all during September. It seemed one or two of these guys were tapped for mop-up duty, and for the most part that was handled correctly. There were a couple questionable situations though, ones where the game was not in hand. For instance, I saw a 19-year old kid pick up a save for a pennant contender.
2) A few position player call-ups per team have favored status for some unknown reason, and end up coming into games on a daily basis as late substitutions, often in key situations. This is how I noticed the problem in the first place. One of my opponents had replaced his star outfielder with a mediocre minor leaguer, in a couple of game situations. A little further investigation showed that this scrub was appearing in a lot of Sept. games.

So because of #2, I would still love the option to turn it off. In my opinion, if the guys in #2 are good enough to play in key games, then they should be good enough to make the 25-man roster. It seems the criteria used to determine the roster is different from what is used to make game decisions.