Whats the use of current ratings?

[BPS]

Originally Posted by CecilCooper

I think one thing that a lot of people are missing is in this game you don't get the day to day interaction with your players that you get IRL.

In spring training in OOTP land you don't get to see, "Wow, player X looks like he aged 10 years in the offseason." "Pitcher Y's control is all over the place right now, but his velocity seems to be there and he has good movement on his pitches." In real-life, going on past statistics does no good, when compared to what you see in real-time with your own eyes. Granted, eventually the statistics would let you know this, however in real life a manager may not have ever let it get that far.

That in my mind, is the reason that I can see the need for some measure of current ratings.

That's how I see them.

In the real world, observers can try to pass judgment on the skills of players by taking into account, at minimum: (1) past statistics, (2) current statistics, and (3) non-statistics based assessments of the player.

The latter is what reported ratings provide. It might be determined by lots of different things: subjective determination of good/bad luck a player might have had during games, observations of the player outside of games in, say, practice, the impact of current physical condition of the player, etc. The ratings might be seen to represent the relative ability of the player if everything, luck etc, averaged out.

Of course, in both real life and OOTP the reported ratings are often off by some amount, or even by a great amount. To this untutored mind, ratings add something important to the game. They represent the information you'd have about players IF you actually saw them in front of you during games, practice, off the field, etc.

[StyxNCa]

Originally Posted by CecilCooper

I think one thing that a lot of people are missing is in this game you don't get the day to day interaction with your players that you get IRL.

I

Now, if we could get Markus to realize we can't have real life communication with coaches and players during the game maybe we could get a popup asking if we want a runner to advance.

[HonusWagner]

Originally Posted by CecilCooper

I think one thing that a lot of people are missing is in this game you don't get the day to day interaction with your players that you get IRL.

In spring training in OOTP land you don't get to see, "Wow, player X looks like he aged 10 years in the offseason." "Pitcher Y's control is all over the place right now, but his velocity seems to be there and he has good movement on his pitches." In real-life, going on past statistics does no good, when compared to what you see in real-time with your own eyes. Granted, eventually the statistics would let you know this, however in real life a manager may not have ever let it get that far.

That in my mind, is the reason that I can see the need for some measure of current ratings.

You make an interesting and valid point. Absolutely a sim is going to be limited. This is what I was getting at with using player development ratings, to see if a player is generally improving or not.

Yet, using your scenario, of how a player is performing --
1. IRL, sure, a manager might notice 'something' is wrong with xxxx, but would the manager take any action until that 'something' manifests itself on the field? Likely not. So one could still use just stats to establish recent trends in performance. IRL, if something is wrong wiht a player but that player is still performing ok on the field, then I dont see how a manager will remove him, if hes still contributing to a winning effort.

2. In OOTP, how does knowing how a pitcher's control or stuff rating mean anything? IRL of course such things could be observed as you say and worked on to improve. But it makes no difference in OOTP. OOTP could use an overall rating and it would still be the same thing. As it is now, if you are seeing green in all current rating pitching categories youre good to go.

This isnt how it works IRL. RL requires history. Scouts, then managers and coaches, spend time getting to know a player. Through this, they learn what he is capable of iow his potential. Thats it. HIstory and potential. These are established via record of performance and impression made. The latter is based on observation. Youre right, OOTP obviously cannot provide observation. But it does have scouting impression (based on quality of scout) and it does have record of performance in stats. So I still dont see the need for current ratings, unless they are to make it easier for the OOTP manager.

3. It is still unclear and has not been demonstrated what current ratings actually are or what they mean. I did one brief test last night using a players current ratings on, using the top scout, a team trainer as scout, and no scout - and through the course of a season there was no change in the players current ratings.

[RonCo]

Interesting discussion. You need current ratings in order for the engine to provide results. I like using relatively poor resolution for ratings levels (1-10, or 2-8) because that just gets you into a band, and from there you have to rely on stats to make a determination. [a "5" power hitter who manages 8 HRs is a tweener--he could be a very low 5 or a high five who got unlucky] for just the reason HW is pointing at.

The better question is "why do you need an artificial potential rating to shoot for." Progressions could, and should be based on a series of things. The scout's potential rating certainly needs to exist, but the game engine _nneds_ a current rating to make results it doesn't actually doesn't _need_ a potential. All it needs is a current age, current skill level, and an expected improvement (along with a random engine and several modifiers), and the player's career path could be mapped out on an "as you go" basis. Then the scout's job would be to predict, based on a current assessment, where that path is most likely to go.

Either way, I agree that the game is most challenging when you play without ratings on. If you have lots of time, it's also much more fun.

However, having ratings at hand is just like playing the old board games where you could see everything. It's also a time-saver. Some folks just don't have hours and hours to spend doing a deep-dive to really understand stats.

In addition, the absolute best use of stats really requires the minor leagues to be well-set-up so those stats can tell a contextual story. This is doable in OOTP--but it's really, really hard, and you have to know a lot about how the game works for it to make sense. If Markus to fix this (also doable), then the use of minor league stats could become a huge enhancer to the game as HW wants to play it.

[HonusWagner]

Originally Posted by BPS

That's how I see them.

In the real world, observers can try to pass judgment on the skills of players by taking into account, at minimum: (1) past statistics, (2) current statistics, and (3) non-statistics based assessments of the player.

THeres no such thing as current statistics. All statistics are past statistics

The latter is what reported ratings provide. It might be determined by lots of different things: subjective determination of good/bad luck a player might have had during games, observations of the player outside of games in, say, practice, the impact of current physical condition of the player, etc. The ratings might be seen to represent the relative ability of the player if everything, luck etc, averaged out.

One is free to imagine current ratings to represent whatever he likes. My issue is with how they are used by the game and, more important, that I do not have the option of removing them if I wish to have a more immersive sim experience. Because if I remove them then there are only potential ratings (which is what I want) for immediate pitching and batting (as I have mentioned in a previous post).

Of course, in both real life and OOTP the reported ratings are often off by some amount, or even by a great amount.

How do you know this? This is what Ive been trying to find out and demonstrate but have not been able to.

[RonCo]

Originally Posted by HonusWagner

3. It is still unclear and has not been demonstrated what current ratings actually are or what they mean. I did one brief test last night using a players current ratings on, using the top scout, a team trainer as scout, and no scout - and through the course of a season there was no change in the players current ratings.

Current ratings are the player's skills, generally mapped to a resulting rate. These are then used to bounce random numbers against in order to achieve results.

For example, a pitcher's Stuff (and a little Control and Movement) are mapped to a K/AB rate matrix. A hitter's AvoidK rating is also mapped to a K/AB rate. These two numbers, along with a value calculated in the League Totals, are compared somewhere in the results engine and a unique probability of a strikeout is determined for these two players (adjusted for ballpark and whatnot). This probability gets used to resolve the at-bat.

[RonCo]

Originally Posted by HonusWagner

THeres no such thing as current statistics. All statistics are past statistics

It would be fair to interpret his comment as "recent stats," which is obviously what he meant.

One is free to imagine current ratings to represent whatever he likes. My issue is with how they are used by the game and, more important, that I do not have the option of removing them if I wish to have a more immersive sim experience. Because if I remove them then there are only potential ratings (which is what I want) for immediate pitching and batting (as I have mentioned in a previous post).

I agree there should be maxes, or perhaps even better most ratings should be composites of lower-level physical ratings...like defense is a composite of arm, range, and error. OOTP is a hodge-podge of rating styles that is a little disquieting at times.

[HonusWagner]

Originally Posted by RonCo

Interesting discussion. You need current ratings in order for the engine to provide results. I like using relatively poor resolution for ratings levels (1-10, or 2-8) because that just gets you into a band, and from there you have to rely on stats to make a determination. [a "5" power hitter who manages 8 HRs is a tweener--he could be a very low 5 or a high five who got unlucky] for just the reason HW is pointing at.

I will have to try that, using poor resolution. Another improvement that would be easy and quick to code would be to have a quality adjuster, like with scouting ratings, for current ratings. Or, just provide an option to turn off current ratings and provide potential ratings in all or most categories.

The better question is "why do you need an artificial potential rating to shoot for." Progressions could, and should be based on a series of things. The scout's potential rating certainly needs to exist, but the game engine _nneds_ a current rating to make results it doesn't actually doesn't _need_ a potential. All it needs is a current age, current skill level, and an expected improvement (along with a random engine and several modifiers), and the player's career path could be mapped out on an "as you go" basis. Then the scout's job would be to predict, based on a current assessment, where that path is most likely to go.

The engine would still work if the current ratings were hidden. THe potential rating is not artificial - it represents an important aspect of the recruitment factor IRL - the opinion of the scout. Whether or not the scout is right is not the issue. My point is allowing for the ambiguity that is found IRL.

Either way, I agree that the game is most challenging when you play without ratings on. If you have lots of time, it's also much more fun.

However, having ratings at hand is just like playing the old board games where you could see everything. It's also a time-saver. Some folks just don't have hours and hours to spend doing a deep-dive to really understand stats.

Exactly. Your use of board games as analogy is accurate. I dont care if people want to use ratings. God bless. But the option ought to be there for a deeper sim experience, if one chooses.

In addition, the absolute best use of stats really requires the minor leagues to be well-set-up so those stats can tell a contextual story. This is doable in OOTP--but it's really, really hard, and you have to know a lot about how the game works for it to make sense. If Markus to fix this (also doable), then the use of minor league stats could become a huge enhancer to the game as HW wants to play it.

This is an interesting point you bring up about minor league stats. Im very new to the game so I know nothing about ths aspect. Why dont they provide a meaningful context of player development/performance?

[HonusWagner]

Originally Posted by RonCo

Current ratings are the player's skills, generally mapped to a resulting rate. These are then used to bounce random numbers against in order to achieve results.

For example, a pitcher's Stuff (and a little Control and Movement) are mapped to a K/AB rate matrix. A hitter's AvoidK rating is also mapped to a K/AB rate. These two numbers, along with a value calculated in the League Totals, are compared somewhere in the results engine and a unique probability of a strikeout is determined for these two players (adjusted for ballpark and whatnot). This probability gets used to resolve the at-bat.

So with the ratings generating stats then one would be smart to cut out the middle man (in this case, stats) and refer to his players variables (ratings). This is exactly the problem. Stats are window-dressing. You dont even need to know a players history. Just look at his current ratings.

[RonCo]

Originally Posted by HonusWagner

The engine would still work if the current ratings were hidden. THe potential rating is not artificial - it represents an important aspect of the recruitment factor IRL - the opinion of the scout. Whether or not the scout is right is not the issue. My point is allowing for the ambiguity that is found IRL.

I've never been particularly fond of scouts in OOTP, but it's all tied into the way OOTP uses the parameter called potential. I'm not suggesting there is no need to project potential--but in OOTP "potential" is not really "potential" as we know it. That parameter can and does change most weeks for many players. What OOTP "potential" actually is in the code is an indicator of how quickly and in what direction the player is most likely to grow.

In this sense, it's a "fake" potential.

But "potential" really should be a human-like projection of future progress based on both "recent" stats and other physical elements of the player in question. Instead, OOTP "potential" is about this strange software parameter that controls "current" or "near future" growth, but that can change a ton.

Realize that when I say that game does not _need_ potential, I mean the inner-workings of the machine does not need it. In fact, if it did not exist, it would make development of a real potential projection more easily understood.

This is an interesting point you bring up about minor league stats. Im very new to the game so I know nothing about ths aspect. Why dont they provide a meaningful context of player development/performance?

They do provide context, but the league totals (and their relationship to ratings) are so strange in the OOTP minors that the context is hard to determine. It is a common complaint that minor league stats are not realistic.

[HonusWagner]

Originally Posted by RonCo

Current ratings are the player's skills, generally mapped to a resulting rate. These are then used to bounce random numbers against in order to achieve results.

For example, a pitcher's Stuff (and a little Control and Movement) are mapped to a K/AB rate matrix. A hitter's AvoidK rating is also mapped to a K/AB rate. These two numbers, along with a value calculated in the League Totals, are compared somewhere in the results engine and a unique probability of a strikeout is determined for these two players (adjusted for ballpark and whatnot). This probability gets used to resolve the at-bat.

In this example, a pitchers stuff would change according to the scale selected by the user (ie 1-5, or 1-100) or does the engine use the highest scale? For example, when I use the 1-5 scale for ratings, Ichiro and Martin Prado are both rated 5 for contact; when I use the 1-100 scale, the engine rates Ichiro 100 and Prado 81. So the larger scale is more representative ORL. So if one uses the 1-5 scale, the engine would not treat Ichiro and Prado the same?

[RonCo]

OOTP's scale is 1-250 no matter what scale you show the data in, but the rating scale is keyed to 1-200. So if you show ratings on a 10-point scale:

1 = 1-20
2 = 21-40
3 = 41-60
4 = 61-80
5 = 81-100
6 = 101-120
7 = 121-140
8 = 141-160
9 = 161-180
10 = 180-200

In the example you show Ichiro is a 200, Prado is a 162 (the 1-100 scale is 2-points per rating point). So in the 1-10 scale, Prado would be a very low 9, Ichiro would be a very high 10.

On a 1-5 scale, Ichiro and Prado would look the same to the user (both "5") but would perform very differently. Hence the reason to use lower-gradient ratings because they obsure "truth." And at that level, this obsuring _probably_ gives an even better fog of war than any OOTP scouting algorithm does. Hence I turn scouts off and use a low-res rating scheme as a general rule.

[RonCo]

Same thing for "potential." When on a 1-10 scale, a "7" potential could be a 121 or a 140...a pretty big gap in performance. Why do I need a scout to obscure that?

[RonCo]

In the old V5 days, Potential used to be "Poor," "Fair." "Average," "Good," and "Brilliant." (oh, for the old days. )

I always looked at is at if:

Poor = 1-40
Fair = 41-80
Average = 81-120
Good = 121-160
Brilliant = 161-200

[HonusWagner]

Excellent feedback!

RonCo, I really appreciate you taking the time to engage in this discussion and provide these insights. The way you describe it there is much room for ambiguity that renders the ratings all but meaningless. Since the bulk of players will be average - that is, either a 2 or 3 - just going by ratings creates a great deal of fog, as you point out.

I also see your point re artificiality of potential ratings -- it sounds like potential is arrived at via a random number generator (unless based on historical stats) rather than using player atributes as variables. I have noticed that a players rating will change dramatically over the course of a season or two. A top draft pick will be released in a couple of years. Perhaps this is intended to mimick real life as after all not all top picks become dominant players. Yet, these things should be based on attribute variables rather than luck, if I understand your point.

So essentially, with the 1-5 scale, you use the ratings as a very rough estimate and use the stats for what they are meant for.

What is your relation to the game? How do you know these details? Is there information available one can read up on?

[HonusWagner]

Originally Posted by RonCo

In the old V5 days, Potential used to be "Poor," "Fair." "Average," "Good," and "Brilliant." (oh, for the old days. )

I always looked at is at if:

Poor = 1-40
Fair = 41-80
Average = 81-120
Good = 121-160
Brilliant = 161-200

That is more realistic. Unfortunately, changes were not progress in this case. When a scout is sitting in the bleachers watching a high school prospect pitch he isnt asking himself if the kids potential is a 83, or maybe a 87, or could be a 44 after that last pitch. He's qualifying an impression that is summed up and expressed using vague, but hopefully meaningful, words and gives a summary that equates to sign em or forget em. A scout would likely laugh at the idea of using a numbered graph to indicate exactly what the kid is going to do. Still, the 1-5 scale seems to be closest to imitating the previous (better) rating system.

[RonCo]

I'm assuming that even inthe old system the game was still based on 1-200 (250), but tha the game just didn't show you numbers.

[RonCo]

Originally Posted by HonusWagner

What is your relation to the game? How do you know these details? Is there information available one can read up on?

I have been fiddling around with it for 6+ years, and have spent some years as a beta tester--though I'm not a beta at present and my knowledge will probably be somewhat obsolete when v10 is released.

[HonusWagner]

Good stuff. So perhaps as a beta alumni you might have clout and let Markus in on some important tips re improvements

[HonusWagner]

Originally Posted by RonCo

I'm assuming that even inthe old system the game was still based on 1-200 (250), but tha the game just didn't show you numbers.

of course, it remains a computer-based, data-driven sim

but I think presentation goes some ways in providing the experience, the suspension of disbelief, not just cosmetically, but what the user bases decisions on. But obviously you agree with that.