[AngelinOF]

Is anyone else getting some screwy line-ups by the CPU? For some reason I am seeing the CPU play people out of position typically in the outfield. An example:

Providence currently has 3 outfielders
LF Harry Stovey: He's rated at all 3 OF positions with LF (50/940) being his best position by far since CF and RF has ratings of 10/900.
CF Paul Hines: He's rated to play LF and CF with both positions being about average 50 range and .940 FP.
RF Mike Tiernan He's rated for all OF position fairly equal between 50-60 for all 3 and .950 FP.

For some reason the CPU is putting Stovey in LF, Tiernan in CF, & Hines in RF (a position he has no rating for). It refuses to put Tiernan at RF his best position. It seems like it should put the players at the position they are listed as since that's their best positions but its ignoring it for all 3 players.

There are also a couple of other teams with similar problems, Cincinnati is playing Sam Crawford in LF and John Dobbs in RF both out of position, Crawford has more range/FP at RF then Dobbs and Dobbs is a better LF. Crawford's arm is also almost twice what Dobbs is 97 to 57 something that really should be looked when determining who plays RF.

Cleveland has Tom Donovan playing RF even though he has no rating at RF and they have a RF who has almost the exact same ratings nor do they have the RF set up to be a defensive replacement.

The Phillies are for some reason playing a 1B (Klondike Douglass) in RF (5/915) and Ed Delahanty at 1B (10/975). They'd be much better suited to play these two at their actual position since they have the similar ratings at those positions Douglass 60/983 at 1B and Delahanty 55/980 at RF.

Pittsburgh insists on playing Fred Clarke and Ginger Beaumont out of position, Clarke in CF and Beaumont in LF, neither have any ratings at the position they are in the lineup at.

St. Louis is playing Jesse Burkett in CF even though he has no rating at that position, and the CF they have on the roster is playing Burkett's natural position.

Washington has their corner outfielders flipflopped also, with Pop Foster playing RF and Sam Dungan playing LF. At the positions they are playing their ratings are 6/940, but their natural positions are 50/950.

[GoldenGlove]

You beat me to it. It looks like it chooses the best fielder for left and the other fielder plays right whether those positions are the primary positions or not.

[swampdragon]

I believe Golden Glove is correct. It's actually worse than that. The AI looks at all the players who can play at catcher, picks the best one, and starts him. The AI then moves on to 1B, etc, etc. It doesn't revisit the decisions. So Frank Chance, rated at catcher but a Hall of Fame 1st baseman, was picked in a 1901 replay to be the Cub catcher. Chance was actually the RF for much of the 1901 season, but the AI didn't consider putting him there, either. Instead, it ran out of outfielders and put Johnny Kling out in RF. This was not a problem in version 5.

[Dan Theman]

This often happened with the in-game logic for ootp5. Whenever a pinch hitter would come in, the computer would rarely reassign the outfielders to create the best overall defense.

[AngelinOF]

Yeah the Cubs line up is a mess for me too, I missed them the first time. Chance is catching in place of Kling and they have a back up RF at 1B who just barely has a fielding rating at 1B (1/989). This isn't a situation to get a better bat into the line up either since this players ratings are 50/30/3/10/73. They would be much better off playing Chance at 1B and Kling 53/36/8/11/74 behind the plate. The defense would also be alot better.

[jbmagic]

here a solution



goto every cpu controll team and do there lineups, depth chart, team strategy.....and it will be lock ...as long u dont click hit Run computer Manager on all teams or entire team it will be lock...if u do the cpu will change there lineup the way they want....

[AngelinOF]

No the CPU resets its line-up after each save if you make a change so in order for that to work you would have to do that prior to simming each day's games.

What Swampdragon said makes sense, I ran a few tests giving the best players in the line-up ratings at catching/1B, meaning the fictional teams #3 & #4 hitters. And in each case the team would put these players at C & 1B and ignore they're normal position and place a less than suitable player in their position.

[swampdragon]

The Cubs have a regular with no set position in Charlie Dexter, and the AI has never handled those terribly well. I remember OOTP5 would frequently bench Honus Wagner for the 1901 Pirates. But this business of handling each position in order is new. If Kling and Chance are both going to be in the lineup, Kling should catch. I would understand if the AI decided to put Chance instead of Kling at C to get the extra bat in the lineup and put Dexter out in RF, but that's not what's happening.

[GoldenGlove]

1st Idea:
This would probably slow the game down, but maybe if it could create a lineup based on primary positions. Then go through each bench player for that lineup and match him up with the positions that he is suited for. If he is better than a player at one of his secondary positions, he is inserted. This of course is flawed as well, but it may be better than it is currently.

2nd Idea:
Create a formula for the offensive ratings. Then use this formula to place the highest players into their best fielding position. As soon as a player is found to have no fielding position open in the defensive alignment, go through his positions and find the higher rated player that can be moved to an open position. If nobody can move, then the next player is looked at until the lineup is filled.

[jbmagic]

just an idea


goto team stragedy

and do individual player tendencies and pick the players u dont want to have no pinch hitter for and move the slider to less.....

and in the depth chart put the sub % to zero or 1 % so your starter dont get tire and will seldom get sub for...

[GoldenGlove]

Its more of an issue of simming and AI generated lineups for computer teams. When playing solo, I create lineups first and then generate depth charts. The AI for both OOTP5 and OOTP6 are flawed currently, although both are reasonable. I'm just looking for improvement.

REVISED 2ND IDEA:
1) Create a formula for the offensive ratings. A hidden system with weights on each rating.
2) Then order this rating.
3) Place the 1st player at his best fielding position.
4) Place the 2nd player at his best position. If it is taken by the 1st player, place the 2nd player at his best secondary position.
5) Continue down the list. If all of a player's defensive positions are filled, put him at DH or go on to the next player if there is no DH.
6) After the lineup is filled, compare the DH to each player at his fielding positions. If he has a better fielding ability, swap the players.

[sporr]

I can at least let you know why this is happening. Markus revised the engine so that secondary positions would be taken into account. If you remember, OOTP would often value primary position listed above all others which sometimes led to poor AI decisions (3 star players rated at all OF positions, but all with RF as primary and 2 would be traded to fill the CF and LF voids for example). Obviously, some more tweaking needs done to get this more accurate. I'm confident Markus will come up with a solution.

[GoldenGlove]

I know thats why it has changed, but I don't know if having it search through positions from C to RF is the answer.

[sporr]

I agree.

[sixfour210]

Originally posted by swampdragon
I believe Golden Glove is correct. It's actually worse than that. The AI looks at all the players who can play at catcher, picks the best one, and starts him. The AI then moves on to 1B, etc, etc. It doesn't revisit the decisions. So Frank Chance, rated at catcher but a Hall of Fame 1st baseman, was picked in a 1901 replay to be the Cub catcher. Chance was actually the RF for much of the 1901 season, but the AI didn't consider putting him there, either. Instead, it ran out of outfielders and put Johnny Kling out in RF. This was not a problem in version 5.

This explains why no matter how bad Chipper Jones is rated at 3B, he will always start there ahead of DeRosa.