Two oddities - one maybe patch-related?

[AdamLindsBeard]

Two odd things I'm noticing that I thought I'd toss out for consumption or conversation:

A couple of patches ago, if I asked my scout to update me on a player, I'd get that update the next day, or maybe a day after. Huge help if I was thinking of grabbing someone off waivers.

Today, because suddenly the scouting accuracy on pretty much everyone tops out at "Average", and most are "Low", I ask for updates more often, and just got word back that the three players I've requested updates on will take two weeks to complete.

# # #
Second thing:

I have a reliever who, 4 appearances into the season, has an ERA of 0.00, a WHIP of 0.39 and is projected to finish the year with a 1.9 WAR.

...and he gets DFA'd by the AI every 3-4 days. Four times now. He's angry because of the transactions.


?!?!?!?!

[Sweed]

Originally Posted by AdamLindsBeard

Two odd things I'm noticing that I thought I'd toss out for consumption or conversation:

A couple of patches ago, if I asked my scout to update me on a player, I'd get that update the next day, or maybe a day after. Huge help if I was thinking of grabbing someone off waivers.

Today, because suddenly the scouting accuracy on pretty much everyone tops out at "Average", and most are "Low", I ask for updates more often, and just got word back that the three players I've requested updates on will take two weeks to complete.

# # #
Second thing:

I have a reliever who, 4 appearances into the season, has an ERA of 0.00, a WHIP of 0.39 and is projected to finish the year with a 1.9 WAR.

...and he gets DFA'd by the AI every 3-4 days. Four times now. He's angry because of the transactions.


?!?!?!?!

On the second thing first. A RP with 4 appearances could almost luck his way into those numbers. The team, in OOTP through your scout, knows (within the fog of war) what type of player this is talent wise. Could be he is way overperforming. Also into consideration would be what kind of player the AI is replacing him with.

The first? Some were reporting scouting reports were coming in too quickly compared to the expected time stated on the scout page. Perhaps this was adjusted in the patch as you noted? Or could be a busier time of year for your scouting staff? More things to do delaying reports?

[AdamLindsBeard]

You might be onto something with the RP, however...

Compare these stats (and I'm looking at my fictitious 2021 Blue Jays):

Phelps: 7.2 IP, 5.9 K/9, 0.00 ERA, 0.39 WHIP

And he gets DFA'd every 3-4 days... versus THESE two guys:

Clase: 3.2 IP, 7.4 K/9, 7.36 ERA, 1.50 WHIP
Kintzler: 6.1 IP, 8.5K/9, 7.11 ERA, 1.74 WHIP

Now... the number there that is the outlier is the K/9... but other than that, Phelps is better in every possible statistic.

But here's where, to me, it gets interesting...

The other difference? The ratings for Stuff / Movement / Control

Phelps: 62 / 45 / 50
Clase: 51 / 71 / 69
Kintzler: 42 / 68 / 77

But STILL... shouldn't the AI be DFA'ing players who are having a lousy year?

Is that a setting I can tweak somewhere, so the AI makes decisions on results instead of ratings?

EDITED TO ADD: Could I repair that behavior by tweaking the "Player Evaluation AI Settings"? Would that impact which players the AI is choosing to DFA?

EDITED AGAIN TO ADD: Opened the whole "weighting" can of worms, read the 600 threads on the board on the subject, playing around with different settings. Thanks for the train of thought.

[Sweed]

Originally Posted by AdamLindsBeard

You might be onto something with the RP, however...

Compare these stats (and I'm looking at my fictitious 2021 Blue Jays):

Phelps: 7.2 IP, 5.9 K/9, 0.00 ERA, 0.39 WHIP

And he gets DFA'd every 3-4 days... versus THESE two guys:

Clase: 3.2 IP, 7.4 K/9, 7.36 ERA, 1.50 WHIP
Kintzler: 6.1 IP, 8.5K/9, 7.11 ERA, 1.74 WHIP

Now... the number there that is the outlier is the K/9... but other than that, Phelps is better in every possible statistic.

But here's where, to me, it gets interesting...

The other difference? The ratings for Stuff / Movement / Control

Phelps: 62 / 45 / 50
Clase: 51 / 71 / 69
Kintzler: 42 / 68 / 77

But STILL... shouldn't the AI be DFA'ing players who are having a lousy year?

Is that a setting I can tweak somewhere, so the AI makes decisions on results instead of ratings?

EDITED TO ADD: Could I repair that behavior by tweaking the "Player Evaluation AI Settings"? Would that impact which players the AI is choosing to DFA?

EDITED AGAIN TO ADD: Opened the whole "weighting" can of worms, read the 600 threads on the board on the subject, playing around with different settings. Thanks for the train of thought.

There certainly are a ton of variables that can go into an OOTP decision. Same as real life

Yeah, the evaluation settings can make a difference. After many years of playing and reading through those threads I finally stuck on 25/25/25/25. That was from a thread in v20, I think?. Haven't seen a reason to switch with my slow play 'em all out style.

Having said that I think, with a new season and limited stats, the AI modifies those weights. IE Phelps' great stats are valued within the context of the small sample size. A good thing IMHO. Like real life, a player could just start great when the team knows his approx real ability and will not be swayed from that evaluation until a large sample size has been established.

I would evaluate those guys as
Clase hands down the best. Phelps and Kintzler are almost a wash to me. I hate BB and HR allowed and those are Phelps' weakest points though, stuf can overcome some of that.

Kintzler on the other hand will get hit while limiting HR and BB. If I have a plus to great defense I might lean towards him over Phelps.

The DFA every 3-4 days is the most troubling thing to my mind rather than
who the AI is selecting in that pool of P's. I'd rather it make a decision and stick with it especially with the small sample size. Could be this is just one of those "hardly any difference in the players" situations where the AI gets stuck in a loop? AI keeps teetering between wanting a power P vs. a control P?

[AdamLindsBeard]

Fun followup...

Made some AI tweaks so that now I'm running with 45/35/15/5.

Played a game. AI brought Phelps in to pitch the 7th and 8th, and here are his numbers from today's game:

2 IP, 2H, 0R, 0BB, 3K.

Our pitchers won the the "Team Effort" achievement for combining on a 6-hit shutout.

Oh - and right after the game? AI responded by DFA'ing Phelps, who's now projected to have a 2.1 WAR at the end of the year.

He's my most dependable arm out of the bullpen so far this year, and the AI can't get rid of him fast enough.

[AdamLindsBeard]

Here's another fun wrinkle in the "What the hell is the AI thinking?" adventure that I'm on.

When I noticed that Phelps had been DFA'd, I looked to see what other things had happened.

Turns out it also juggled my starting rotation.

It took Tyler Chatwood:

G 3 GS 3
IP 17 / HA 14 / BB 6 / K 8 / WHIP 1.18 / ERA 2.12

...and replaced him with TJ Zeuch:

G 4 GS 0
IP 7 / HA 7 / BB 3 / K 6 / WHIP 1.43 / ERA 3.86

When my inner voice started yelling "WTF", I went and looked at their ratings for STU / MOV / CON / STA and found this:


Zeuch: 40 / 56 / 55 / 60
Chatwood: 58 / 70 / 32 / 40

Because we're Toronto, in a park where Movement > Stuff > Control among EX GB pitchers, the 23-point difference in Control is easily wiped out by the advantage Chatwood has in Movement and Stuff. So then we're left with Stamina, where Zeuch has a 60.

Clearly Chatwood is a better choice to start; by averaging 5.2 IP per start, he's demonstrated he can pitch consistently into the sixth inning.

So what's the deal? And how do I combat THAT one? Do I need to go to the Player Strategy and FORCE the AI to use him as a starter?

[HoustonGM]

I do find that each year there's a handful of players that seem to get DFA'd every couple days for a while. Sometimes even one day after they previously got DFA'd. Performance entirely aside, that does seem like an odd quirk of the AI.

[AdamLindsBeard]

Originally Posted by HoustonGM

I do find that each year there's a handful of players that seem to get DFA'd every couple days for a while. Sometimes even one day after they previously got DFA'd. Performance entirely aside, that does seem like an odd quirk of the AI.

Yeah, I mostly throw stuff like this out there in case there's something obvious that I'm missing in terms of my setup, or if it's something the devs want to flag as something worth checking.

(For example, I never thought about tinkering with the weighting until today... and I've been playing the game for years.)