@GoozeFn_PSN said:

@Dolenz_PSN Interesting, but there’s a lot of nuance in this game that won’t get considered, like a CF playing out of position on a corner equally to a secondary. I have no idea how these programs work, but it’s pretty interesting.

True and since I don't play online and I don't have most cards I can't possibly cover every nuance. This program started as a way for me to automate the legends and flashback counts per team chart and ballooned out of control.

Good news, after discussing it with the AI it thinks we could implement a feature to allow users (assuming I release it and people actually want to use it) to easily change the weights behind the Meta Scores and run reports based on that new weighting. That means you could change it and run reports and then change it again and run more reports, eventually ranking each position by your own weightings. It would be far more personalized that way

I may try this out after the first big roster update and I have those changes implemented.

@GoozeFn_PSN said:

@Dolenz_PSN You almost have to do it by position though. Defense really doesn’t seem to matter at all for 1B. Does anything affect a 1B ability to pick a throw in the dirt? At catcher arm strength/accuracy, pop time and blocking are all that matter. Reaction and speed matter the most in the OF. Ability, reaction and arm at SS. It’s very different from position to position.

I'll think about it for the future. But I feel like it would really impact the Squad building logic.

Here is that explanation straight from the program itself (I am under no delusion that many if any will ready this)

The Best Overall Squad builder uses a "puzzle-piece" algorithm designed to maximize the total combined rating of your 13 active hitters (9 starters, 4 bench). Here is the breakdown of how it decides who plays where:

1. The "True Rating" (Meta Score)

Instead of just looking at the number on the front of the card (OVR), the system calculates a Meta Score. This weighs the most important stats (Power, Contact, Speed, and Fielding) to determine a player's actual impact. A 99 OVR with low speed might actually be "worse" than a 97 OVR with elite speed in this model.

2. The Positional Hierarchy (Why players move)

The algorithm doesn't just pick the top 9 players; it runs a positional optimization:

Primary Spots First: It first identifies the "Locks"—players like catchers (C) who rarely have secondary positions.

The Secondary Shuffle: If you have two elite Shortstops (SS), the algorithm checks if one has a secondary position (like 2B or 3B). It compares two scenarios:

Scenario A: SS #1 at SS, and a mediocre 2B at 2B.

Scenario B: SS #1 at SS, and SS #2 at 2B (taking a small penalty to his defensive rating).

If Scenario B results in a higher total team score, it moves the player to the secondary spot.

3. The DH (Designated Hitter) "Safety Valve"

The DH spot is handled last as a "best available" slot.

Relegation: A player is relegated to DH if they are an elite hitter but their Fielding stat is lower than the other players who qualify for their position.

Example: If you have two elite 1B/3B types, the one with the better "Arm" and "Reaction" stats will get the 3B glove, while the "slugger" with lower defense is moved to DH to keep their bat in the lineup without hurting the team's defensive average.

4. Bench Logic

Once the 9 starters are optimized to create the highest possible score, the remaining players are sorted by their Meta Score. The top 4 remaining enter the bench.

Summary: The algorithm is looking for the highest mathematical sum. It will happily move a superstar to their secondary position or the DH spot if it allows a slightly lower-rated "specialist" (like a defensive wizard) to fill a hole that no one else can play, resulting in a better overall team total.

@GoozeFn_PSN said:

@Dolenz_PSN To be clear, I enjoy these posts from your program. I think defense matters, but it depends on the position. Good defense is most important at SS and OF IMO. Not sure if you have the ability to give it weight in that way.

I agree that defense matters more for the positions up the middle but I just did not want to try to dive into the quagmire to try and give them all some set value. So right now it evaluates every position the same way.

I am thinking about releasing the program (give people a chance to download and run it locally) after we get our first big roster update and I redesign the screens for that. After these posts I thought that what the program may be able to pull off would be a settings page where users could override the default weightings for Meta, Offensive Meta and Pitching Meta and set up their own weighting.

I still don't know if I would want to do it by position though as then you have to worry about secondary positions also. I already have some complexity built into the Squad Generator where it will frequently put guys in their secondary positions.

The whole Meta scores calculations in my program was originally designed just as a way for me to differentiate between the vast gulf we see between early 99s and late 99s as we progress through the year.

@GoozeFn_PSN said:

@Dolenz_PSN I’m always happy to see that there are still things humans are better at than computers. Cabrera is more valuable than Pujols and Pedroia doesn’t belong anywhere near this list.

To each their own. I never make claims that my program is a benchmark any person should use when constructing their squads. It is simply doing math. It is a reporting tool and, while I used AI to program it, it does not even try and use AI to evaluate the cards . It can't take into account each person's personal preferences or what attributes they value most at each position. It doesn't try to quantify a batters swing or whether they bat lefty, righty or switch. I realize these things are important to players but I choose not to try and quantify them.

Pujols vs Cabrera
You can see beneath the card image that the program considers Cabrera the better offensive player (110.6 to 105.78). But when you then expand it to include Fielding and Speed it gives Pujols a slight lead (96.05 to 95.97). It is a very negligible amount though (.08) and essentially is a dead heat.

Pedroia gets a big overall boost from his defense although he also has very good Contact and CLT. He has little power so I understand why people would not use him. It's obvious you prefer power at the position. Nothing wrong with that.

But again the program is just doing math.

@bjd10048_MLBTS said:

You’re not smarter than anyone.

Yes he is.

I have about 20 Ish at bats with the 95 miggy and have him parallel up to a 98 and so far I’m not really impressed by the swing. I’m not saying it’s necessarily bad but I would say tulo low and Pujols are both better when you factor in attributes defense

Maggie has felt a tad slow and not really generating the power I was hoping for