OOTP19 Ratings and AI Real Talk

[KBLover]

Hey OOTPers, let's discuss our thinking, reasons, and settings for ratings, scouting, and AI evaluation.

Ask questions, share settings, share good/bad results you especially liked, all that good stuff.

[bravesfan1984]

Thinking of setting my team strategy to sabermetric. What do you guys use? Since I'm looking for a new manager, I'd like him to be a sabermetric one, but the pickings are slim. Would it be that much of a detriment if I hired Girardi, who I think has a traditional setting?

Sent from my SM-G965U using Operation Sports mobile app

[KBLover]

Hard to say since performance will depend on a lot of factors, but since he's got a high reputation, he might keep you from the strategy you want to set (check how much control he allows over his strategy).

If you have your heart set on it, I might would pick another guy that would let you set strategy and still has a good development influence and relationship with your players (especially your core ones).

That said, it might not matter a ton (a debate that exist IRL as well) and a manager with good ratings and relationship might offset any downsides.

Keep in mind, you can see a sample lineup and pitching staff from the manager to get an idea of how he uses your existing roster. This is especially important if he retains control over the lineup.

[canes21]

I'll bring this post into here.

Quote: Originally Posted by WaitTilNextYear

All this convo about settings probably goes better in the OFFICIAL OOTP thread, but since we're discussing it here (and there's only like 5 of us posting anyway, lol)...my AI eval settings are 60/20/15/5. I put a heavier weight on ratings because I don't want to bamboozle the cpu out of talented, young players that are having a down statistical year/month etc. I feel like this puts a good weight on the talent under the hood with the statistical performance thrown in as a kicker. I have scouting on "low" as well. I like how that works with the 1-100 ratings scale. I don't use the 20-80 scale because of too much repetition with everyone being a 45/50/55 etc.

My settings are 35/35/20/10 and I use the 20-80 scale with "Very Low" scouting accuracy. The main thing I wanted to talk about in this post is the comment about repetition with everyone being a 45/50/55. That's actually part of what I love about the scale. I've always hated 0-100 scales in games. I always used the letter system in College Hoops 2k8.

I love the combination of scouting not being accurate and the ratings scale only having 13 places you can land on it between 20 and 80. I love the "fog of war" it brings to the game. For every superstar we know is good in the league there are 20 guys who are 45 or 50's. Guys that one organization values at a 60, one values them at 35. The league is built off of these guys, not the superstars. I love the impreciseness of the scale. In real life it's really hard to really quantify what makes non-superstar player X "better" than non-superstar player Y. He may have quicker hands, but by how much and how does that translate to more success and how much success.

I feel like with a 0-100 rating, I personally get hung up on looking at a 85 overall player and wondering why he's playing worse than an 84. Yet, for whatever reason when I see a 45 outperforming a 40 with the 20-80 scale I just think, maybe my scouts got it wrong, maybe his strengths are the right combination for him to fit into a perfect role on that team.

In no way am I saying a 0-100 scale is worse than a 20-80 scale. Use whatever you want, but I just personally love the Very Low/20-80 combination as it brings in the extreme lack of a sure thing in talent evaluations that I love about sports. I dislike precise systems representing imprecise things. I know there are solid numbers under the hood, but at least with the settings I use, I can forget that and reason with whatever happens better than I can with a 0-100 scale. I don't know why that is, but it is just how I am.

[KBLover]

I like 20-80 for similar reasons. It's a good amount of uncertainty without feeling like I have zero reliable information.

Scouts are fallible, and scouting is certainly an imprecise science, though I think it's getting better, especially at the majors with all the data being collected and being able to actually quantify things that were purely "feel" and eye test.

That's why I keep scouting on High accuracy. The inherent uncertainty of the grade scale + the vagaries in performance inherent to baseball itself is enough for me. I'm already scratching my head as to why red hot Ohtani suddenly ended up with a very average ERA as his ratings are increasing. Then he turns around and posts a 2.01 postseason ERA... will the real Ohtani-san please stand up?
Something isn't consistently clicking between him, his skill set, and the team and/or his usage. And he's not the only one between the franchises I have.

Guys with high avoiding K's striking out a lot. Guys with 70 contact having trash BABIPs hurting their production. And too many pitchers to list. So there's plenty of unknowns to deal with in my book.

Plus, I'm pretty sure my scout, despite his ratings and the money he's allocated, is being wrong, at least at first glance. I saw it when I bought Yoshida again. I thought "wow she's doing much better this time" but first report in my system is "nope, still struggling". Drafts will be a crap shoot for sure because money is going to international scouting since I'm signing compensation FAs probably forever until the team rebuilds. I won't even go there with the uncertainty of international scouting...

Seen that with some trades, too. And when I see a change immediately after acquiring someone, I wonder if it's real development or "correcting scouting error".

[WaitTilNextYear]

The reason I like the 0-100 scale more than the more common 20-80 "scouting scale" is basically for variety.

If two players are both rated 50 on the "scouting scale" they are basically the same. If two players are rated 51 and 53 on the 0-100 scale, they are still basically the same but I enjoy the visual variety of having the different numbers.

Also, the development report updates provide more to unpack with a 0-100 scale because a lot of times an attribute will go up 1 point or down 2 points so you can get an idea of where a player is trending (even if barely) instead of them staying at the same increment of 5 within the "scouting scale."

I don't really think setting it at 0-100 vs 20-80 gives me much more of an advantage because I also turn off OVR/POT stars/numbers, set scouting to low/very low, and set trading to very hard. Any advantage I get from a bit more pixelation in the ratings is cancelled by these other settings. I think I get adequate fog of war this way.

If I really didn't want to get hung up on ratings or pay attention to them at all, I'd just go straight stats only. The main reason I don't tend to use stats only much anymore is because I simply don't like the look of the player cards without ratings bars/colors/numbers on them. Call me superficial, lol.

[KBLover]

Are two 50's basically the same?

I was under the impression that 50 (or any of the ratings) represented a range and that range, potentially, could be different (one player closer to a 45 and one closer to a 55)?

Or is that wrong?

[WaitTilNextYear]

You can't tell that from the ratings though. You are inferring that. If there are two 50's, then they are identical within the context of that ratings scale. You can imagine anything you'd like in terms of whether a 50 is actually a hidden 52 or a closet 47. If that's how you prefer to get your fog of war, then that's one way to do it.