Rehashing the Reliever Issue

[RangerJoe]

[quote:019d36d7d1]I'm sure this is gonna be a hot debate, but I'm gonna get to the bottom of this, and I'm ready to take it on.

As you know, I have ventured into Strat-O-Matic's proprietary game engine code and have been trying to figure things out. I've put traces in the bullpen logic functions and have been reviewing them for a little while now. What's jumped out at me is this following theory:

Most of the perceived failings of the computer's bullpen management are actually the result of small, overworked, or over-restricted bullpens.

The game engine was originally written to manage teams which are constructed similarly to real-life teams - i.e. with 11-man staffs, and 5- or 6-man bullpens. It does an effective job with them, and produces realistic results over the course of a full season. As I'm following along with the leagues that you're playing now, all I'm seeing everywhere are small (4-man), overworked bullpens, and every other pitcher set to "Avoid" something or other. Thus, in a given relief situation, the computer manager does the best it can with very few choices, and that only makes it harder for each successive decision it has to make, and then you'll see it seemingly "ignore" some of the settings you've assigned. If you have 1 guy set as a closer, 2 guys set to avoid lefties, and another guy set to avoid being used before the 8th, that doesn't leave the computer a lot of options in a crucial 6th inning situation.

It is my opinion, and that of some prominent members of the SOM Online community, that if you carry a large, realistic enough bullpen (at least 5 eligible guys, preferably even 6), that you'll see that HAL will do a good job managing it.

This said, I am giving you the opportunity to convince me otherwise. I could say, "there's no changing the code," but I WON'T. I am prepared to make necessary changes, IF I am convinced that there is a problem. I think this is a fair approach. I'm starting another thread in which you can post specific examples of what you think are poor relief decisions.

Thanks,
Bernie
[/quote:019d36d7d1]


This is from a thread Bernie began over at the "SOM Baseball Online" board last summer. During the past few days, there have been some added responses, briniging it back to that board's forefront. This issue came up in one of my leagues today, so I thought I would bring it over to the ATG II board, illiciting some responses from the standpoint of the historical game. The opinions below are my own, and are certainly not omnipotent. :wink:

While Bernie's explanation makes some sense, I have some disagreement. For instance, my limited experience has been that the more pitchers one carries the more choices HAL is given to rest a reliever that was used in (say) two previous games; explaining Doug Bair's appearance in a save situation while Newhouser sits. So I don't believe carrying 11 or 12 pitchers is the answer either. That is not to say that carrying that many pitchers isn't a good strategy on occassion, as I know of managers who effectively carry more pitchers on a regular basis. I just don't believe that it cuts own on HAL's strange bullpen moves.

In a lot of recent ATG leagues, I have been able to get HAL to behave
most of the time, by how and when set-up's and closer's boxes are checked, but it is still less than an exact science. One explantion might be in the difference between ATG and the modern 0X game, where CL is a designated position, and where the 1 inning fatigue is more common. In these cases it makes little sense to have a closer enter before the ninth. One example is the time I had Gagne and checked not to enter before the ninth and not to be used in a blowout; the result was a string of 26 straight appearances with a save. Many, if not most of the stud relievers at ATG carry a 3-inning fatigue factor, so naturally most managers do not want to limit the designated closer to the 9th inning. That also tends to confuse HAL about whom to bring in when. Thus, in my experience, the use of a closer and set-up is less an issue over at the 0X game than over here.

One last thing about carrying more pitchers. This explanation doesn't explain how many of us were able to put together supberb pitching staffs, with low ERA and WHIP, when we were able to use only 8 pitchers at ATG I. Actually it was seven pitchers, when you consider that one space was usually held by Harper, Medich, etc. So IMHO, in a game featuring many pre-1980's pitchers who carry a SP8 or 9 factor, larger staffs isn't necessarily a better thing.

[Free Radicals]

I always try to carry at least a 5 man pen , with a stud from both sides and then the cheapies. I only designate like a Garber or Perranoski as 2-3 innings . The rest I have on quick hook 1-2 innings , sometimes avoid whichever depending on the pitcher. I have found that by narrowing Hal's choices by going click crazy on settings only confuses him . By having a 'main guy' for Hal to go to with a backup 'main guy' it eases the pressure put on the pen . Hal makes some bonehead move sometimes it would appear but if you look closely at whats going on in the pen you'd see that the reason maybe Bair comes in is because Newhouser has pitched 3 of the last 4 games and and his fatigue is showing . I used to complain about Hal and came to the conclusion he does what he can with what you give him and most times his moves are the correct one , if it works out or not .

[Valen]

[quote:62000b014a]One last thing about carrying more pitchers. This explanation doesn't explain how many of us were able to put together supberb pitching staffs, with low ERA and WHIP, when we were able to use only 8 pitchers at ATG I.[/quote:62000b014a]
I always carried 10+ pitchers in ATG1 on the theory that the extra relievers helped prevent anyone from pitching tired. Looking back I rarely was among the league leaders in ERA in spite of having what I considered some pretty good staffs. I had became frustrated about the teams carrying only 8 pitchers and as you say in reality less always doing better than me.

So in short I have to agree with premise that carrying more pitchers does not help HAL truly get better results.