OOTP Scouting

Non-league talk in here. Please make NSFW *links* and not pics.
User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Fri Aug 21, 2020 5:38 pm

In other words, the definition of "Talent" is different in game than it is to us humans (and OOTP scouts):

Talent to the game engine: "A marker for how far and in what direction I should I should progress this player's ratings now."
Talent to a game scout: "Where can this player be expected to peak at?"
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Fri Aug 21, 2020 6:22 pm

The OOTP Gods and TCR….


I’ve not played in an online league that totally cranks the Talent Change Randomness (TCR) to full bore. I’ve heard various reports on how it feels, but I can say that in the old days I had data to show this factor did make considerable differences to the fortunes of players. I’m tempted to say we should do that, but I don’t want to screw up this great thing we have here for a pet peeve or lab project (grin).

Ted’s comment has me thinking of this because, IMHO, I think this is one of the holes in the game’s development logic: players who crash don’t come back up often enough. Some of this could be “fixed” in the development logic, and some of it really can’t—things like adding new pitches for pitchers, for example, are (in my way of thinking about it) only indirectly related to the development model (in which Stuff, Movement, and Control are the direct inputs, and pitch types are indirect).

Anyway, and advantage of cranking TCR is that it logically increases the chance of players having their Talents (rating vectors) adjusted multiple times, both up and down. It is still, to my memory, not really “real” but it’s a little closer. It’s problem again, however, is that development will generally be downward even in “normal” leagues, and most human GMs get down in the dumps when development reports come in with bad news.

The real answer to this is to play stats only—without ratings. I’m not suggesting that for the BBA, BTW. But that is the answer. And the “beauty” of that is that—assuming you have your minor league Totals well set, the feel of the game is very real. It does, however, take a lot more focus by us GMs to watch performance rather than rely on ratings and arrows and whatnot.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Fri Aug 21, 2020 6:23 pm

Well...that was a fun way to waste an afternoon.

It makes me want to test the dev curves again. :)
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

User avatar
Jwalk100
GB: FL Pacific Division Director
Posts: 3039
Joined: Sat Jan 11, 2020 9:42 pm
Has thanked: 1817 times
Been thanked: 771 times

Re: OOTP Scouting

Post by Jwalk100 » Fri Aug 21, 2020 7:07 pm

I play in @HerbD stats only league (The Heartland Baseball League). I find that stats only is refreshing because you are rid of the clutter of ratings. I have no use of my excel spreadsheets in that league which are set up for ratings. I only use my numbers spreadsheets which use "Moneyball" formulas.
Image
ImageImageImageImageImage

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: OOTP Scouting

Post by Ted » Fri Aug 21, 2020 7:36 pm

Jwalk100 wrote:
Fri Aug 21, 2020 7:07 pm
I play in @HerbD stats only league (The Heartland Baseball League). I find that stats only is refreshing because you are rid of the clutter of ratings. I have no use of my excel spreadsheets in that league which are set up for ratings. I only use my numbers spreadsheets which use "Moneyball" formulas.
I've thought about trying stats only. The problem is, then I'd want scouts. Like actual scouting reports. Or the information to make my own. And OOTP just doesn't provide that. You can have true ratings, shitty scouts, or nothing.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Fri Aug 21, 2020 8:10 pm

I enjoyed the stats only league I was in quite a bit... but it does take a different mindset.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

User avatar
Jwalk100
GB: FL Pacific Division Director
Posts: 3039
Joined: Sat Jan 11, 2020 9:42 pm
Has thanked: 1817 times
Been thanked: 771 times

Re: OOTP Scouting

Post by Jwalk100 » Fri Aug 21, 2020 8:14 pm

Ted wrote:
Fri Aug 21, 2020 7:36 pm

You can have true ratings, shitty scouts, or nothing.

That sums it up for me playing this game.
Image
ImageImageImageImageImage

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Fri Aug 21, 2020 10:25 pm

Correct.

For me, the best way is pretty much where we are now. "True" ratings, done at a a rating scheme that provides a decent Fog of War, messing a bit with the PCMs and the Talent Randomness.

I like 20-80, but I admit I like the idea behind the 2-8 range even better just because it expands the performance bands a little.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Mon Aug 24, 2020 1:06 pm

So, yeah...you got me excited about career paths, and so I've spent a day working with my old scripts and finding all sorts of more issue...which isn't the point here, but what the heck. Never let a chance to beat a dead horse pass. Anyway, just as an example of what I mean by saying that "dead prospects" aren't necessarily dead when they lump early, here is a small sample of 5 hitters who had Eye Talent lumps in their teens, then went on to get some very healthy bumps...

Eye-TAL-5-Careers.PNG

The data comes from a long sim with a somewhat out of the box game settings (player dev = 1.0, Talent changes standard). By itself, these five mean nothing--but the main point is that even in today's OOTP players can and do recover from issues early in their development.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: OOTP Scouting

Post by Ted » Mon Aug 24, 2020 1:21 pm

RonCo wrote:
Mon Aug 24, 2020 1:06 pm
So, yeah...you got me excited about career paths, and so I've spent a day working with my old scripts and finding all sorts of more issue...which isn't the point here, but what the heck. Never let a chance to beat a dead horse pass. Anyway, just as an example of what I mean by saying that "dead prospects" aren't necessarily dead when they lump early, here is a small sample of 5 hitters who had Eye Talent lumps in their teens, then went on to get some very healthy bumps...


Eye-TAL-5-Careers.PNG


The data comes from a long sim with a somewhat out of the box game settings (player dev = 1.0, Talent changes standard). By itself, these five mean nothing--but the main point is that even in today's OOTP players can and do recover from issues early in their development.
Yeah a few guys do recover. I hold on to pretty much anyone up until around age 19 unless I can't find a spot anywhere. Most guys who ever had any sort of potential I tried to keep until around 21, but you can't wait for em all to bounce back. Still, as you've noted. The whole model isn't really that great. Should either be more variability in the potentials overall, or very little lumping until a guy proves he isn't developing. The more I think about it, I wonder if that's really the way to go. 17-21 or so is when you should see marginal and non prospects bump, and blue chippers should have very little movement in potentials. Then as they age they should maybe lump their potentials as it appears they won't develop. Lots of ways to be better. And as you keep pointing out, a lot of it has to do with OOTP ratings not really representing scouting reports the way we think of them. They probably should, though.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Mon Aug 24, 2020 1:26 pm

I'm not really on the page that ratings should represent scouting reports. That's actually the real issue with us humans.

"Scouting reports" and ratings need to be distanced, else we see the sausage getting made--and since we humans can never see the reality of what makes humans work in real life, that's a source of angst.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: OOTP Scouting

Post by Ted » Mon Aug 24, 2020 1:51 pm

RonCo wrote:
Mon Aug 24, 2020 1:26 pm
I'm not really on the page that ratings should represent scouting reports. That's actually the real issue with us humans.

"Scouting reports" and ratings need to be distanced, else we see the sausage getting made--and since we humans can never see the reality of what makes humans work in real life, that's a source of angst.
I guess that I'm unclear as to what the purpose of a potential rating that's not a scouting report is. It's basically the game telling you how good a player might get. Isn't that what a scouting report tries to do? It's just that with a game, we know the true talent potential, whereas scouts have to estimate it. If the potential rating isn't the game's version of a scouting report, what is it then?

If not to be a scouting report, why have a rating that functions so differently from anything in real life? Real players don't have a "true talent potential". That's not a thing unless you believe in an all knowing god (and even then mere mortals do not have access to that info). Real players have a collection of physical and mental attributes, along with a certain degree of skill, that people look at and try to determine how good the player can get. That's what scouting is, right? Why would you write a game such that there is a "potential" that doesn't represent what that scouting is trying to do? What does that model add? It would be one thing if the potential ratings was fixed, and you could say that the game gives you the option of "Seeing behind the curtain" or letting you predict the future with certainty, basically viewing players as god.. But the potential changes. So it's not there for that purpose. So then what is potential rating if not a scouting report on what we think a player will become?

As far as some distance between current ratings and scouting reports, maybe I can see it. Actual ratings are in fact, true talent level. So sure, have that different from scouting. But again, I then assert that's the role of the potential rating.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Mon Aug 24, 2020 1:52 pm

I say that because in realty, the players in question don't change any in their skills. If you replaced OOTP "Talent" or "Potential" with "How fast this guy is learning right now relative to his current state," you would get a better picture of what is actually happening. Instead, we humans want it to mean "what a scout says." In many cases we can make it work well enough to keep the fiction going. But OOTP development does not work like Scouts work--hence the issue.

Here are the top three of those players' career Eye Skill (rating) versus his rate of learning ("Talent").

Note that in none of these (admittedly small sample size) cases does the kid in question ever get worse...instead he either slows his learning or accelerates it.

The next question, then, is how often do these things happen, which we can look at, too.

EYE-TAL-RAT-3-CAREERS.PNG
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Mon Aug 24, 2020 2:02 pm

OOTP--like almost all sports management games--needs a software parameter to steer development with...and most call it something like "talent." But that's not how humans work (as we both said). "Talent" is a rating the software developer needs to manage skill development (in this model, anyway ... there are other ways, but they aren't as ... um ... simple or obvious). Once we have "Talent" in this form, it becomes easy to forget that this is not actually the player's most likely outcome, hence mistake a scouts job as "identifying where that parameter is right now."

This is the root of why I think scouting in OOTP will never feel very good if you look into the details.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: OOTP Scouting

Post by Ted » Mon Aug 24, 2020 3:24 pm

RonCo wrote:
Mon Aug 24, 2020 1:52 pm
I say that because in realty, the players in question don't change any in their skills. If you replaced OOTP "Talent" or "Potential" with "How fast this guy is learning right now relative to his current state," you would get a better picture of what is actually happening. Instead, we humans want it to mean "what a scout says." In many cases we can make it work well enough to keep the fiction going. But OOTP development does not work like Scouts work--hence the issue.

Here are the top three of those players' career Eye Skill (rating) versus his rate of learning ("Talent").

Note that in none of these (admittedly small sample size) cases does the kid in question ever get worse...instead he either slows his learning or accelerates it.

The next question, then, is how often do these things happen, which we can look at, too.


EYE-TAL-RAT-3-CAREERS.PNG
Hmm, unless you know that's how the code works, I'm not sure I buy the vector idea. I've had plenty of guys stay 2 of 8 in a rating for years, then lump to 2 of 2, then 1 of 1. I don't see how the potential talent can be a measure of how rapidly a skill is increasing relative to current state if some guys just don't move. I'm also pretty sure I've had a player lump an actual rating either without dropping potential, or with dropping the potential by one. Say was 2/8, now 1/7 or 1/8. I don't have the data to prove it, so I could be misremembering. Although, relative ratings may me muddying the waters. I can't say if those incidents were pre or post 2037 midseason.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: OOTP Scouting

Post by Ted » Mon Aug 24, 2020 3:31 pm

Either way, I think we've gotten to the point of dithering over game mechanics and semantics. OOTP's model doesn't represent scouting. So the scout really can't work the way most of us would expect, as you say. And yeah, I get that the software developer needs a model for where a player is likely to end up, but OOTP's mixing it with publicly visible ratings and then having their in game scouts describe those ratings is just a mess.

Were I starting from scratch, I wouldn't use this model. But if we want to talk about it in relation to what OOTP has, what's being used at POT now would be hidden, except as commish. It wouldn't change. The publicly visible version of "pot" would be a scouting report that would be a median outcome of any hidden "true talent pot". Then as the player developed better or worse than the median track for said hidden pot, you'd upgrade or downgrade the publicly visible one. Or something like that.
Last edited by Ted on Mon Aug 24, 2020 3:51 pm, edited 1 time in total.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Mon Aug 24, 2020 3:40 pm

All of that is fundamentally sound enough as far as I can tell.

This is where I throw my very strongly worded statement in that I have no idea how the game actually does it's mechanics. My views are my own, built up from about two decades of testing the crud out of the results the game gives. There is no doubt that you will find players whose ratings fall with or without a "Talent" change. For my cash, that's an indicator of randomness around a mean, but I could well be wrong in that guess.

This thread is about scouting, though, and these posts are meant to portray a world that is just as you just noted: that given how the game appears to use the software parameter called "Talent" in its development process, the scouting algorithm should have some other goal than to use identifying "Talent" as the scout's goal. Otherwise, it's really just a slave to the pernicious "God" inside OOTP, and will always feel wrong.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19807
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1981 times
Been thanked: 2900 times

Re: OOTP Scouting

Post by RonCo » Mon Aug 24, 2020 5:24 pm

Going off the deep end like I do...

Here is output of some of my old scripts on this v21 test league. I have to compile them all by hand right now, which sucks. So I'm trying to get into code so it just comes off the run automatically.

In this case it's "average" progression of the EYE rating by age, including some idea of the number of big changes in a season, by age.

Note this is Ratings, Not "Talent" ... but I grab all the raw material for a bunch of stuff regarding both talent and ratings progression (including unique pitch progression). It's just a matter of getting my brain around the process of gathering it all together well. Still, this is "true development" in that it describes how our OOTP humans actually acquire skill over time.

The point in this chart is that on the whole, there's not a bunch here to say anything is "wrong" ... I mean, I think this chart says that players grow about like you'd expect them to overall, though (to what I think is at least some of Ted's point here) the number of wild swings are not massively high. That said, this is Eye we're talking about. I'm not sure I have a great idea of how it should be. (*)

(*) When I chain ratings development together in a way Tom Tango described on his site, though, I get similar looking curves.

And this is a bit of a generic league, with TCR at "normal." In addition, there are still OOTP screw ups that cause me to dump a chunk of data that I can't grab properly ... so buyer beware.
EYE-RAT-PROGRESSION.PNG
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

Post Reply Previous topicNext topic

Return to “Off Topic Chatter”

Who is online

Users browsing this forum: No registered users and 5 guests