Questionable AI decisions

[Dyzalot]

Originally Posted by jaysdailydose

This thread is now 16 pages long, but I went back and looked.

Exactly like I figured, your Randolph that they put at 3B has an 18 arm, compared to an 11 for Sanchez. Arm is generally looked at as the most important tool for a third baseman.

Can easily see why they moved him to third, and Sanchez to second now.

See what I mean? Most of these things you are describing can be explained with even a cursory look at the ratings. An 11 arm on the 10-20 scale is actually extremely low for anyone with 3B experience and true 3B ability.

What the numbers show me is a man who is comfortable playing third but doesn't really have the tools to play there, and a backup coming in who doesn't have a ton of experience but has all the tools to play the position and then some.

No. Perez was at third with a "16" arm and an overall of "14". Randolph may have a slightly better arm but he has almost no experience and he has better range, error and turn dp ratings, all of which seem to me to be more important factors at 2nd than they are at third.

[jaysdailydose]

Originally Posted by Dyzalot

No. Perez was at third with a "16" arm and an overall of "14". Randolph may have a slightly better arm but he has almost no experience and he has better range, error and turn dp ratings, all of which seem to me to be more important factors at 2nd than they are at third.

That I misread, but I can still see why they did it. The range and the error ratings are definitely valued higher at third than second by the AI.

Randolph would be an elite third baseman with more experience, so I can see why a manager would make that call, even with Perez instead of Sanchez (which I apologize for misreading.)

I can see why you would think that the positional rating would be more important, however, and I don't mean to imply that it doesn't have value itself. It does. It is still totally explainable, and really not even strange, to see the decision.

[Dyzalot]

Originally Posted by jaysdailydose

That I misread, but I can still see why they did it. The range and the error ratings are definitely valued higher at third than second by the AI.

Randolph would be an elite third baseman with more experience, so I can see why a manager would make that call, even with Perez instead of Sanchez (which I apologize for misreading.)

I can see why you would think that the positional rating would be more important, however, and I don't mean to imply that it doesn't have value itself. It does. It is still totally explainable, and really not even strange, to see the decision.

We'll have to agree to disagree then. I don't believe MLB managers put the utility infielder at third where he has barely played and moving the starting third baseman instead of putting him at 2nd. I also disagree with what is valued more where. I certainly believe that range and error rate are more appreciated at 2nd than at third.

[MKG1734]

Originally Posted by endgame

I tend to favor a Zen approach and offer it up often: seek first to understand and only then to be understood. It's a simple governor for growth.

Your post did not apply to what I was posting; however, this part of your post applies to what I am attempting to do.

After getting back online and checking and re-reading this thread, it appears that jaysdaileydoes - in particular - seems to think I am saying the game is egregiously incorrect in what it is doing (without seeking to find out WHY it is doing something). This is not the case.

What I AM attempting to do is figure out HOW the game works.

For example, jaysdaileydose continues to say that (from my previous example of Players A, B, and C) that Player B and his plus-plus-power (65) is being played over a player with much stronger hitting ratings on the whole in Player C. On top of this fact, Player C has substantially stronger statistics when compared with Player B. ESPECIALLY when vs. RHP only is considered.

Now, I understand all the previous; however, jaysdaileydose seems to think I am ONLY wanting the Player C to play over Player B. This is not true. While I disagree with the AI manager playing Player B over Player C, I am confused with how the game is working.

For example, and this seems to perhaps be a point of confusion with my posts ...

If the players are used for STATS ONLY - with 100% of the AI Evaluation based on the current year - the lineup does not change. The AI continues to play what is a suboptimal overall player statistics-wise (as well as ratings wise).

Now, this lineup permeates thru EVERY setting Ive tested on the AI Evaluation. So....does this mean that the AI Evaluation settings mean nothing when it comes to lineup creation?

If the answer is they don't...then okay. What DOES create lineup settings?

If the answer is they do...then okay...but why does the lineup stay the same thru each setting? For example ... if stats mean 100% ... why would the game play a player who is clearly having a much worse year when compared to a player having a much better year?

For those who haven't followed, let me make this easy:

The game has 3 players who are essentially equal at both SS and 2B (all three can play both positions essentially equally)

Ive since finished the season ... here are the year-end stats (HR and AB/HR in parentheses)

Total:
Player A: 376 PA, .283 - .336 - .477 (15, 23/1), 2.2 WAR
Player B: 313 PA, .257 - .294 - .429 (12, 25/1), 1.5 WAR
Player C: 326 PA, .318 - .374 - .505 (7, 43/1), 3.3 WAR

vs. RHP:
Player A: 322 PA, .282 - .332 - .480 (13, 23/1)
Player B: 203 PA, .228 - .266 - .394 (8, 24/1)
Player C: 243 PA, .330 - .374 - .542 (7, 32/1)

**NOTE: Player B was brought up later in the season, which accounts for the fewer Plate App.; however, Player B was a starter vs RHP and vs LHP

Furthermore (with all defensive ratings essentially the same):

Player A:
Con: 50/55
Gap: 70/70
HR: 55/55
Eye: 40/55
AvoidK: 45/45

Player B:
Con: 45/55
Gap: 50/60
HR: 65/75
Eye: 45/50
AvoidK: 30/35

Player C:
Con: 60/75
Gap: 70/75
HR: 50/65
Eye: 45/60
AvoidK: 55/60

If the game is set up to play 'stats only' - why is Player B starting over either Player A or Player C? Its a legitimate question ... and understanding its answer will either help me understand what the game is doing or help me adjust a setting (but which setting?)

Under 'stats only' isn't Player B's 65 HR Power ignored in favor of stats? If so ... why is Player B starting?

[jaysdailydose]

Originally Posted by Dyzalot

We'll have to agree to disagree then. I don't believe MLB managers put the utility infielder at third where he has barely played and moving the starting third baseman instead of putting him at 2nd. I also disagree with what is valued more where. I certainly believe that range and error rate are more appreciated at 2nd than at third.

Back to MLB managers, again...

For one, in near every scenario, that kid would be playing third base. They aren't going to waste a near-HOF type of arm at second base.

You can disagree with me on what is valued more, but if you think range is more valuable at second than third, we're not going to agree on much of anything. Second base is the easiest position to play in the infield outside of first. Obviously, errors are bad anywhere -- but range is FAR more valuable at third base than second considering any ball hit down the line is a double if you don't range to it. This is not the case at second base.

Regardless of whether you disagree with me or not, there is nothing inherently wrong with what the AI did -- you just don't agree with it. Others will. That's baseball.

If you don't like what happens, as said previously -- manage the games yourself. You obviously don't like what the AI managers do in thousands of situations...

You must really think Buck Showalter is an idiot starting a guy with no second base experience and DHing his utility infielder then... (see Steve Pearce and Jimmy Paredes.)

There is NOTHING wrong with the AI in this instance.

[jaysdailydose]

Originally Posted by MKG1734

Your post did not apply to what I was posting; however, this part of your post applies to what I am attempting to do.

After getting back online and checking and re-reading this thread, it appears that jaysdaileydoes - in particular - seems to think I am saying the game is egregiously incorrect in what it is doing (without seeking to find out WHY it is doing something). This is not the case.

What I AM attempting to do is figure out HOW the game works.

For example, jaysdaileydose continues to say that (from my previous example of Players A, B, and C) that Player B and his plus-plus-power (65) is being played over a player with much stronger hitting ratings on the whole in Player C. On top of this fact, Player C has substantially stronger statistics when compared with Player B. ESPECIALLY when vs. RHP only is considered.

Now, I understand all the previous; however, jaysdaileydose seems to think I am ONLY wanting the Player C to play over Player B. This is not true. While I disagree with the AI manager playing Player B over Player C, I am confused with how the game is working.

For example, and this seems to perhaps be a point of confusion with my posts ...

If the players are used for STATS ONLY - with 100% of the AI Evaluation based on the current year - the lineup does not change. The AI continues to play what is a suboptimal overall player statistics-wise (as well as ratings wise).

Now, this lineup permeates thru EVERY setting Ive tested on the AI Evaluation. So....does this mean that the AI Evaluation settings mean nothing when it comes to lineup creation?

If the answer is they don't...then okay. What DOES create lineup settings?

If the answer is they do...then okay...but why does the lineup stay the same thru each setting? For example ... if stats mean 100% ... why would the game play a player who is clearly having a much worse year when compared to a player having a much better year?

For those who haven't followed, let me make this easy:

The game has 3 players who are essentially equal at both SS and 2B (all three can play both positions essentially equally)

Ive since finished the season ... here are the year-end stats (HR and AB/HR in parentheses)

Total:
Player A: 376 PA, .283 - .336 - .477 (15, 23/1), 2.2 WAR
Player B: 313 PA, .257 - .294 - .429 (12, 25/1), 1.5 WAR
Player C: 326 PA, .318 - .374 - .505 (7, 43/1), 3.3 WAR

vs. RHP:
Player A: 322 PA, .282 - .332 - .480 (13, 23/1)
Player B: 203 PA, .228 - .266 - .394 (8, 24/1)
Player C: 243 PA, .330 - .374 - .542 (7, 32/1)

**NOTE: Player B was brought up later in the season, which accounts for the fewer Plate App.; however, Player B was a starter vs RHP and vs LHP

Furthermore (with all defensive ratings essentially the same):

Player A:
Con: 50/55
Gap: 70/70
HR: 55/55
Eye: 40/55
AvoidK: 45/45

Player B:
Con: 45/55
Gap: 50/60
HR: 65/75
Eye: 45/50
AvoidK: 30/35

Player C:
Con: 60/75
Gap: 70/75
HR: 50/65
Eye: 45/60
AvoidK: 55/60

If the game is set up to play 'stats only' - why is Player B starting over either Player A or Player C? Its a legitimate question ... and understanding its answer will either help me understand what the game is doing or help me adjust a setting (but which setting?)

Under 'stats only' isn't Player B's 65 HR Power ignored in favor of stats? If so ... why is Player B starting?

The 65 might be ignored. The 15 jacks in 300 plate appearances from a young shortstop would not be. Potential is also still considered by the AI as well. The stats are already showing him out to be a very productive power hitter at a young age.

Why don't you open the editor and see what makes your manager tick, for real, instead of wondering... it might shed a lot of light on why he is making these decisions and it is obviously bothering you a great deal.

[MKG1734]

Originally Posted by jaysdailydose

The 65 might be ignored. The 15 jacks in 300 plate appearances from a young shortstop would not be. Potential is also still considered by the AI as well.

vs. RHP:
Player B: 203 PA, .228 - .266 - .394 (8, 24/1)
Player C: 243 PA, .330 - .374 - .542 (7, 32/1)

So, with Stats Only ... the game chooses to play (even removed the manager now) Player B over Player C vs RHP ...

Notice, Player C's slightly less HR power, statistically, but much greater hitting and even 'power hitting' ability (542 vs 396 SLG).

So ... I'm still stuck with zero answers. Does AI Evaluation actually AFFECT the lineups? If it is supposed to ... and I can show this issue as above .... is it working properly? Or does it not, and the lineups are set up purely 'under the hood' with no customizable setting to really change it? (since the lineups stay the same with any evaluation setting I choose...)

[jaysdailydose]

I'm done here. You don't listen to anything.

If you hate that they are playing them where you don't like, manage yourself. Or get a grip and do what I told you to do and edit the damned manager and SEE why he's doing what he is doing rather than wondering.

Simple, yet starting to think you'd rather argue.

[MKG1734]

Originally Posted by jaysdailydose

I'm done here. You don't listen to anything.

If you hate that they are playing them where you don't like, manage yourself. Or get a grip and do what I told you to do and edit the damned manager and SEE why he's doing what he is doing rather than wondering.

Simple, yet starting to think you'd rather argue.

I just told you I REMOVED the manager (turned them off). And the lineup setting that the GAME chooses remains the same. Which means the GAME is choosing this lineup as well. Under EVERY AI Evaluation setting. And I'm asking ... how could that be? Bc if stats only matters ... why is THIS the result? With no manager and no ratings taken into account?

[jaysdailydose]

Originally Posted by MKG1734

I just told you I REMOVED the manager (turned them off). And the lineup setting that the GAME chooses remains the same. Which means the GAME is choosing this lineup as well. Under EVERY AI Evaluation setting. And I'm asking ... how could that be? Bc if stats only matters ... why is THIS the result? With no manager and no ratings taken into account?

And I've told you several reasons why it could be...you ever think the game just views baez as the better player and prospect or maybe even thinks he's the only one who should play short?

Man, if you get this bent out of shape over every similarly rated player you're in for some tough times. These players are literally so close to each other skills wise it could literally just be baez's potential. If you hate it so bad, just do the lineups yourself.

Again, you seem to just want to complain over something 99% of us would go "well, my scouts and manager think he's the better player."

[jaysdailydose]

I think your biggest problem is realizing that stats-only is a term used by the community for taking player skill ratings out of the equation. It doesn't mean that stats (or ratings the other way) is the only thing used for a full evaluation of the player. I'm not a dev, but to think it is solely making decisions off just the stats (or just the ratings) is pretty absurd, I think?

For instance, you can set for a full stats-only setup, but if you don't turn the ratings off you still see them. The computer still has to look at potential, injury and fatigue status, matchups, day/night, righty/lefty and thousands of other factors.

If you think you know better than the cpu, then manage. That is why you get the option.

[MKG1734]

Originally Posted by jaysdailydose

And I've told you several reasons why it could be...you ever think the game just views baez as the better player and prospect or maybe even thinks he's the only one who should play short?

Man, if you get this bent out of shape over every similarly rated player you're in for some tough times. These players are literally so close to each other skills wise it could literally just be baez's potential. If you hate it so bad, just do the lineups yourself.

Again, you seem to just want to complain over something 99% of us would go "well, my scouts and manager think he's the better player."

Don't you see what is going on here? Its apparent you dont. Let me help you:

Me: AI Evaluation set to 100% current year stats, no managers
You:

well, my scouts and manager think he's the better player

Me: All 3 players are equally good at SS and 2B (with Player C the best)
You:

you ever think the game ... thinks he's the only one who should play short

You:

These players are literally so close to each other skills wise it could literally just be baez's potential

Me: Doesn't 'Stats only' remove this from the equation (ratings)? .... still haven't received an answer .... on TOP of that, Player B's potential is LESS than Player C's ... which negates your statement here

You:

you ever think the game just views baez as the better player and prospect ... ?

Me: Doesn't "stats only" remove this from the process? Shouldn't the game be playing a lineup based on stats and not on potential (since ratings are not taken into account) .... still haven't received an answer for this, either. And, again, Player B's potential is LESS than Player C's ... which negates your statement here

[jaysdailydose]

Originally Posted by MKG1734

Don't you see what is going on here? Its apparent you dont. Let me help you:

Me: AI Evaluation set to 100% current year stats, no managers
You:

Me: All 3 players are equally good at SS and 2B (with Player C the best)
You:

You:
Me: Doesn't 'Stats only' remove this from the equation (ratings)? .... still haven't received an answer .... on TOP of that, Player B's potential is LESS than Player C's ... which negates your statement here

You:
Me: Doesn't "stats only" remove this from the process? Shouldn't the game be playing a lineup based on stats and not on potential (since ratings are not taken into account) .... still haven't received an answer for this, either. And, again, Player B's potential is LESS than Player C's ... which negates your statement here

No, stats only does not make the ai ignore potential, which I have stated.

[jaysdailydose]

Nor does it negate my statement as Baez has the potential for 40+ Homer power. I don't know why I am even still talking, in this amount of time you could have opened the editor and seen what made the manager make the decision (before you dumped him) or look in the players themselves in the editor and see what you might be missing.

Yes, there are hidden variables you can't see. But I'm honestly kind of baffled that you think the evaluation process is so simple and basic.

[MKG1734]

Originally Posted by jaysdailydose

No, stats only does not make the ai ignore potential, which I have stated.

Thank you.

stats-only is a term used by the community for taking player skill ratings out of the equation

Ok, I am assuming you mean the individual skill attributes such as hit / hit for power / etc. Now, with player skill ratings taken out of the equation we are left with two things:

Stats
Overall Rating / Potential Rating

If this is what you mean ...

Player C is better than Player B with regard to both. So ... with managers off, and the above information taken into account ... what is prompting the game to play Player B? I'm sincerely trying to figure it out so that I can adjust / turn off this setting.


With regard to the editor:
Player C's current and potential is vastly superior when compared with Player B's

[jaysdailydose]

Also, if you are playing with scouting on, your scouting could be completely off and the ai could be working from that. So many variables, stop looking at it like it should just be cut-and-dry.

Or just manage yourself. You obviously think the AI is doing a worse job than you would... and you may even be right!! Only one way to find out!

[MKG1734]

Originally Posted by jaysdailydose

Also, if you are playing with scouting on, your scouting could be completely off and the ai could be working from that. So many variables, stop looking at it like it should just be cut-and-dry.

Or just manage yourself. You obviously think the AI is doing a worse job than you would... and you may even be right!! Only one way to find out!

Scouting set to 100% accurate

And Im attempting to watch the game for errors to see what I can fix bc I like to play in what amounts to 'god mode' where I don't control a team and simply watch the league develop. However; this is hard to do when I know the game is making odd decisions with regard to the settings. Therefore, I'm attempting to identify what setting would cause the AI to make such a decision when the settings should, essentially, cause it to make the opposite decision. Not to mention the editor (if there are under the hood mechanisms at work) indicate that the player who is NOT playing is far superior than the player who IS playing.

To me....that means something is not working properly. Maybe I'm just much more detail oriented than the average OOTPer to notice things that are amiss.

[jaysdailydose]

Originally Posted by MKG1734

Scouting set to 100% accurate

And Im attempting to watch the game for errors to see what I can fix bc I like to play in what amounts to 'god mode' where I don't control a team and simply watch the league develop. However; this is hard to do when I know the game is making odd decisions with regard to the settings. Therefore, I'm attempting to identify what setting would cause the AI to make such a decision when the settings should, essentially, cause it to make the opposite decision. Not to mention the editor (if there are under the hood mechanisms at work) indicate that the player who is NOT playing is far superior than the player who IS playing.

OK, I see what the problem is now. You realize that when you leave the ratings on, and then evaluate it off the statistics, that the overall rating is based on said stats, right?

In actuality, Baez could be the way better player -- but be having a down year, and your scouts are seeing that.

I hate that you guys alienated The Wolf, as my friend is honestly far better at explaining this part than me.

The best thing I can do for you since you seem to really need to get to the bottom of this is to go take this season and go look in the editor at everything. You'll be able to see far more of whats going on and it might shed a lot of light into what is going on.

Baez is showing as a crappy rated OVR because he had a **** year. The CPU scouts are telling the AI that Baez is the better player, so he's getting the nod -- but its pissing you off because the rating (and the production) is telling you that Russell is better (which he is playing better right now.)

I would bet that when you move on into the next season, that Russell will have a ratings jump, and Baez's will decline, and your AI will start putting Russell in the lineup. I won't say I guarantee it, but its a pretty good bet.

I still think you should go play around in the editor a touch, if nothing else but to get a better understanding of how everything works of each other.

[jaysdailydose]

The main reason why so many people that went to stats-only went to stats-only is because evaluating off of the ratings leads you to have omniscient GM's. They see the decline before it is happening, and the ratings drop. You know instantly when the guy falls off the cliff. (Obviously not very realistic.)

With stats only, you don't see the decline (or the lack of a peak) until after its happened.

I'm guessing in your scenario, now that we have far more background, is this:

Baez is the better rated player to the scouts. The engine is going to still rely a bit on ratings, as none of these players have a lot of stats to work off (which is the reason why most stats-only posts recommend going four years into the future and using feeder leagues.)

So, you bring Baez up, and the CPU sees the player it scouted. He only got a little bit over 300 ABs when you had him up (going off recall) so he'd probably be prime for a drop in his contact/eye ability (say under-the-hood, if you will) but now your scouts aren't analyzing the ratings, they are analyzing the stats, so they don't see it coming right away.

Like I said in the last post, I would anticipate a bump for Russell and a drop or at least staying the same for Baez, and I anticipate that Russell will be your starter. We'll see how it works out.

[MKG1734]

The best thing I can do for you since you seem to really need to get to the bottom of this is to go take this season and go look in the editor at everything. You'll be able to see far more of whats going on and it might shed a lot of light into what is going on.

I understand your entire above post, and believe me, Im not intentionally attempting to annoy you. However, I have to continue this bc of your above statement.

I'm not brand new to OOTP. I understand everything with regard to the editor / ratings within the editor, etc. Which is why I'm most confused here .... the editor indicates Player C (Russell) is far, far superior to Player B (Baez). Which is essentially the opposite of what the game is doing. With the actual ratings with the game ALSO being the opposite of what the settings indicate the game is being asked to do...

Hence why I'm frustrated. Nothing about this makes sense. From any angle. Which makes me worry that the game is not working properly, inherently.