[Charlie Hough]

I've been compiling a list of pretty significant issues with OOTP, and these include problems with both historical games and with general AI or sim behavior that could impact current MLB games as well.

This list was compiled based on multiple historical saved games and across multiple seasons within those games. In each case, I was using real historical minors, but I don't think that impacted or created any of these issues.

1. The AI must STOP leaving starting pitchers in the game for far too long, even when the pull starting pitcher AI slider is set to maximum. The AI will leave starters in the game to give up 7 runs instead of pulling them, even with the slider all the way to the maximum of quick pulling of starters. No manager is ever going to do this if there is a full bullpen of rested relievers available. The AI needs to base pulling pitchers on runs and earned runs allowed and not just fatigue and historical era strategy settings.

2. Too many quality free agents remain unsigned by spring training and the start of the season, and too many of them suddenly accept minor league contracts with major league options. AI teams should be signing these players before it gets to this point, and I can't figure out why they're not doing it. Their contract demands aren't high, they're often better than other players on teams' 40-man rosters, and their latest stats and ratings are very good. If they were signed in a more timely fashion, then they wouldn't end up lowering their demands and settling for a minor league deal with a major league option.

3. The AI leaves far too many experienced major league players as free agents after the entire off-season and spring training, and even throughout the subsequent season. It could easily sign them to minor league contracts, but it doesn't do that. Many of them are good enough to be on an MLB roster and have better ratings than some of the AI teams players at their positions. But the AI does not recognize this. AI teams could sign these players to improve their talent at certain positions or to assign these players to AAA and later call them up in case of injuries or poor performance, but there are always a number of veteran players that the AI ignores and leaves as free agents. For example, how does Milt Wilcox, with 16 wins during the previous season and decent ratings, remain unsigned with a low contract demand? How do outfielders who are good enough to be starter or a team's best reserve outfielder go unsigned?

4. The AI waives players far too frequently and risks losing younger players or players with good talent because it keeps waiving some of the same players multiple times throughout a season. The AI will often waive players with good ratings and good prior-year or even current-year stats, yet it will keep players on the roster at the same position that are not as good. Then another AI team or the human team will claim some of these players, and they are lost with no compensation.

5. League total modifiers are sometimes incorrect or inappropriate for starting pitcher stamina and sometimes for relief pitcher stamina as well. I am seeing this when using a player's total career as a basis for stamina, so maybe the sim is trying to adjust for the higher overall pitcher stamina. But it ends up creating situations where starting pitchers and relivers cannot throw a realistic number of pitches relative to their stamina ratings and their era without starting to become fatigued too early and then becoming so fatigued that they can't pitch to a realistic number of innings compared to what they regularly did in real life. For example, I've seen the sim set starting pitcher stamina at around .830 for 1981, which is far too low. I end up having to manually review the league total modifiers each season to ensure accuracy.

6. Similar to the above, league total modifiers for balks are typically way above what they should be. For historical games in the 1980s and 1990s, I consistently see the sim set the modifier at 1.2+, which results in way too many balks being called. I have to manually change this modifier after it's calculated before each season.

7. Occasionally, the AI makes insane decisions, such as placing a young Ozzie Smith on irrevocable waivers when he is by far the best shortstop in the organization. I had to use the option to lock Smith and one of his teammates on the 1981 San Diego Padres because the AI kept wanting to waive and demote them and use players who had worse ratings at their positions. I've seen similar problems with Alfredo Griffin, the top shortstop in the Toronto Blue Jays organization in 1981, being declined arbitration and being allowed to become a free agent despite having a low contract demand.

8. In some cases, counter to point #2 above, players have an overly inflated expectation that they should be signed to a major league contract extension when they're currently on a minor league deal and their ratings and recent history do not justify this. For example, a player who is currently in the minors, has a minor league contract, isn't good enough to be on 40-man roster, and has never played in the majors or has not played in the majors over the past couple of seasons, should not be demanding a major league contract.