Change to 100 top prospects=bad

[clarnzz]

Which pretty much boils down to that if talent is a "perception" as people say it is since it changes. This perception should have a chance to adjust based on performance as scouts usually factor a players minor league performance as a reflection of his abilities. There are certainly exceptions to the rule, but more than likely a young player dominating in the high minors for an extended period of time would be considered "talented" even if he wasn't before. The game engine still allows for this prospect to bust, so why not allow his performance to increase the chance of increasing his true gauge of likelihood of future success, his talent. Relying on a players performance to push his ratings to the max and then hoping the maxed rating pushes the talent to the next level just doesn't cut it.

[draven085]

Originally posted by Henry
I hate to say this - but you have it all wrong...
First of all, a player who is "fair" today may not be "fair" tomorrow. He can get a boost or a bump based on age and/or coaches as well as a random bump.

Maybe I am really dumb, but I don't see how qualifying that bumps are more likely to occur at a younger age (which is as it should be) is not random on the individual level. The bumps still aren't tied to anything aside from the player's age. They are based on nothing about his individual abilities or performance....unless this quote from Markus is true.

Originally posted by Markus Heinsohn
Talent increases/bumps occur when a player is close to the nearest talent rating with his actual ratings. So, maybe Oakland has more players close to the next higher level, and that's why the increases occur. If that is not the case, I suggest simming the scenario 100 times... you will notice that things will even out.

I posted about this in the development thread but didn't get much response there.

[gopads02]

Originally posted by Henry

Given what we have, I think we all agree that Talent - Ratings - and Stats are what we have to work with in OOTP.

We don't work with Stats... in fact we only work with Talent & Ratings in OOTP - that's what I want fixed...and I'm pretty sure I'm not the only one who wants this fixed. I going to make a small hunch here, but I'm guessing anyone who plays this game as long as I have will eventually be aware of this problem and they will eventually want this fixed as well once they realize, they don't need to look at minor league stats at all. They only need to look at current ratings bars and talent bars.

It is an injustice how Minor League Stats are being neglected and ignored, and the only way for it to garner the attention it deserves is for it to have legitmacy- IE a chance at affecting development.

It's kind of ironic how Stats/Performance are the most important notion of this game, and yet it is completely discounted by us in evaluating how a player will develop in the future. That is a seriously unrealistic aspect of the game.

[Cyclone792]

Originally posted by gopads02
What I don't like is now the top 100 is basically misleading us by telling us that we should account for Stats (since it factors it in now when it generates the list), while the game engine itself is not factoring Stats at all when generating the development of players. Notice the disconnect ? The top 100 is for prospects, prospects are meant to be developed. Hence the top 100, whether you use it or not, is a tool for choosing players who should be developed (IE given a chance to fail/succeed). This tool uses stats to evaluate players, therefore it's implying we should use stats to evalute players too! But we shouldn't since stats currently don't affect development of ratings, it is merely a by-product of it.

Gopads, no need to go shooting yourself, I see your point CLEARLY and wholeheartedly agree with it. I'm not sure exactly which post you mentioned it in, but I loved the idea of there being a certain percentage increased probability of a random talent boost when a player performed above expectations in the minor leagues. And the fact of the Top 100 prospects list using stats to evaluate players when the game engine itself does not, creates that disconnect you are talking about.

The 100 prospects list has to go one of two ways: Back to the old system OR to the new system, but only if there is a connection with the game engine in enabling stats to further develop players (i.e. increasing/decreasing the probablity of random talent hits based on statistical performance in the minor leagues). Personally, I would prefer the latter because it would then make minor league statistics useful in evaluating prospects. As it is now with the current system, I rarely use minor league statistics when evaluating players, but long for a system where it would be *useful* to use those stats when evaluating players.

[Henry]

Originally posted by gopads02
What I don't like is now the top 100 is basically misleading us by telling us that we should account for Stats (since it factors it in now when it generates the list), while the game engine itself is not factoring Stats at all when generating the development of players. Notice the disconnect ? The top 100 is for prospects, prospects are meant to be developed. Hence the top 100, whether you use it or not, is a tool for choosing players who should be developed (IE given a chance to fail/succeed). This tool uses stats to evaluate players, therefore it's implying we should use stats to evalute players too! But we shouldn't since stats currently don't affect development of ratings, it is merely a by-product of it.

Ahhhh..... NOW I see this point. I think your then saying, if stats are to be included in the list, they should also be included in development. If that is so, then I agree - but I would also say that they shouldn't be included at all in either place - based on my other post.

I also agree with your stand that both ratings systems should come from the same variable - but "weighted" differently as I said earlier, so they suggest slightly different results. I'm also glad we agree that the list should not be any more accurate that RL.

On the other issue though, I still have a problem. There seems to be a belief that talent is only driven by randomness - and I truly believe a lot of this comes from the fustration felt when the game seems to consistantly "bump" players in the HTML reports.
As I stated in the previous post - it is less random that thought. The "bell curve" represented by Age is paraount in the development curves. Coaches minor league managers, and randomness are a small "adjustement" to this primary variable (age). And the hidden factor cannot be understated. If a player gets a 110% or a 90%, his entire career will be molded from that label.

[clarnzz]

Originally posted by Henry
Ahhhh..... NOW I see this point. I think your then saying, if stats are to be included in the list, they should also be included in development. If that is so, then I agree - but I would also say that they shouldn't be included at all in either place - based on my other post.

I can't believe you just now see this point, it has been the basis of the other side of this arguement for a month and a half now.

[Cyclone792]

Originally posted by draven085
Maybe I am really dumb, but I don't see how qualifying that bumps are more likely to occur at a younger age (which is as it should be) is not random on the individual level. The bumps still aren't tied to anything aside from the player's age. They are based on nothing about his individual abilities or performance....unless this quote from Markus is true.


quote:
--------------------------------------------------------------------------------
Originally posted by Markus Heinsohn
Talent increases/bumps occur when a player is close to the nearest talent rating with his actual ratings. So, maybe Oakland has more players close to the next higher level, and that's why the increases occur. If that is not the case, I suggest simming the scenario 100 times... you will notice that things will even out.
--------------------------------------------------------------------------------



I posted about this in the development thread but didn't get much response there.

I don't know how other people read Markus' quote, but this is an example of how I understood it: Lets say Player A shows us a 5 rating for batting average with AVERAGE talent. His hidden rating is close to that of a 6, but not quite close enough to show us that he is a 6. But since he is rather close to qualifying for GOOD talent, he then gets a bump up to GOOD talent.

Maybe I'm dumb here and that's not what Markus meant. But IF it is what he meant, then it still doesn't explain how some guy in A ball with 1 and 2 ratings has talent boosts across the board that sends every category up a notch. What is REALLY frustrating is when a guy in the minors is putting up great statistics, far more than his talents/ratings indicate that he should have and then that same player has talent drops across the board, killing any shot of that player to ever play in the Bigs (without, of course, a future talent boost).

[Cyclone792]

Originally posted by Henry
Ahhhh..... NOW I see this point. I think your then saying, if stats are to be included in the list, they should also be included in development. If that is so, then I agree - but I would also say that they shouldn't be included at all in either place - based on my other post.

I also agree with your stand that both ratings systems should come from the same variable - but "weighted" differently as I said earlier, so they suggest slightly different results. I'm also glad we agree that the list should not be any more accurate that RL.

On the other issue though, I still have a problem. There seems to be a belief that talent is only driven by randomness - and I truly believe a lot of this comes from the fustration felt when the game seems to consistantly "bump" players in the HTML reports.
As I stated in the previous post - it is less random that thought. The "bell curve" represented by Age is paraount in the development curves. Coaches minor league managers, and randomness are a small "adjustement" to this primary variable (age). And the hidden factor cannot be understated. If a player gets a 110% or a 90%, his entire career will be molded from that label.

But what about when COACHES and SCOUTS are OFF? I understand that coaches should play a role in the development of players when they are enabled, but when they are disabled, what is the formula then? Is it just age and randomness? If so, then every minor league player should have a significantly higher chance of a talent increase rather than decrease because, as you say, a player has a greater chance for an increase up to the age of 27. But what I've observed seems to be the opposite - at most, it appears that talent boosts/drops are even (and one could argue that there seems to be more drops than boosts with players under the age of 27). And without coaches/scouts on, the only other factor appears to be randomness.

[Henry]

Part of the problem with this thread is it's going too many directions - but unfortunately, all of the issues are related to each other.

We don't work with Stats... in fact we only work with Talent & Ratings in OOTP - that's what I want fixed.

It's kind of ironic how Stats/Performance are the most important notion of this game, and yet it is completely discounted by us in evaluating how a player will develop in the future. That is a seriously unrealistic aspect of the game.

Here is one of the major issues. There are just as many on this board that think stats should be included as those that don't. Markus made the change in the first place because of a groundswell of those that felt stats should play a part in the Top 100 list. I think it's now becoming clear that might be ok - IF - they are also a part of the development cycle. Given both options, and consideringthe amount of work it would take to recode the development algorythm, I would think it better to remove stats fromt he propect calculation until some future version.

draven,

Try and picture a bell curve representing the development curve of a player based on age - reaching his peak at 27-33. The random ups and downs based on how close their ratings are to the next level (as Markus mentioned) are decided by that curve. There are some exceptions, of course, but the design of this variance is controlled by that bell curve. Add to that again, the effect of coaches, etc.

In the end - you still have a bell curve (most of the time) but it's ups and downs are simply variances of the "theme". The bell curve is "god" in the sense all the randomness must start there to effect it... thus it's not a totally random roller coaster.

[gopads02]

Originally posted by clarnzz
I can't believe you just now see this point, it has been the basis of the other side of this arguement for a month and a half now.

You know why ? because a month and a half ago, no one ever used terms like "Fog of War" or "Smoke & Mirrors"

[Henry]

Originally posted by clarnzz
I can't believe you just now see this point, it has been the basis of the other side of this arguement for a month and a half now.

I don't see there was a need for that...

[Cyclone792]

Originally posted by Henry

I think it's now becoming clear that might be ok - IF - they are also a part of the development cycle. Given both options, and consideringthe amount of work it would take to recode the development algorythm, I would think it better to remove stats fromt he propect calculation until some future version.

And now we have what I believe is one of the most important aspects of OOTP6 that needs to be worked on! Recode the development algorithm so that statistics DO play a role in the development cycle.

[clarnzz]

Originally posted by Henry
I don't see there was a need for that...

How so? This debate has been hashed for some time now and the point has been brought up by numerous people. I just found it surprising that after 3 threads and over 200 posts that you just now saw the other side of the table's point.

[draven085]

Originally posted by Cyclone792
And now we have what I believe is one of the most important aspects of OOTP6 that needs to be worked on! Recode the development algorithm so that statistics DO play a role in the development cycle.

Absolutely. God, I can't wait for OOTP6

[draven085]

Originally posted by Henry

draven,

Try and picture a bell curve representing the development curve of a player based on age - reaching his peak at 27-33. The random ups and downs based on how close their ratings are to the next level (as Markus mentioned) are decided by that curve. There are some exceptions, of course, but the design of this variance is controlled by that bell curve. Add to that again, the effect of coaches, etc.

In the end - you still have a bell curve (most of the time) but it's ups and downs are simply variances of the "theme". The bell curve is "god" in the sense all the randomness must start there to effect it... thus it's not a totally random roller coaster.

Henry-

I understand what you are trying to say....but like many other I'm arguing that a curve that doesn't take current performance into consideration during development is fundamentally flawed. Because the current curve ignores this crucial portion of data I consider the talent bumps/hits to be random even if they fall within the age curve.

[clarnzz]

Need to make sure you use the proper disclaimer though. That you don't want predictability or every guy who bats .350 in AA for a season turning into the next Pete Rose. Just a small twinge to make searching for that diamond in the rough that has been defying the odds a fun process and a reason to look at something besides his age/talent/rating.

[gopads02]

Originally posted by Henry
Part of the problem with this thread is it's going too many directions - but unfortunately, all of the issues are related to each other.

Wow so you're slowly getting it, henry- how everything connects and in the end leads back to that one simple problem of random talent increases. It's not your fault Henry, it's a very complicated issue/problem to explain and the solution though it is very simple, again is very difficult to explain....(so again don't take anything personal here- we're just working towards the same goal: that is making this superior game even better and more realistic).

Try and picture a bell curve representing the development curve of a player based on age - reaching his peak at 27-33. The random ups and downs based on how close their ratings are to the next level (as Markus mentioned) are decided by that curve. There are some exceptions, of course, but the design of this variance is controlled by that bell curve. Add to that again, the effect of coaches, etc.

In the end - you still have a bell curve (most of the time) but it's ups and downs are simply variances of the "theme". The bell curve is "god" in the sense all the randomness must start there to effect it... thus it's not a totally random roller coaster.

BTW, FYI I bet Draven, myself, and everyone else understood this bellcurve/age factors before you even said what you just said. We have already taken that into account when we're making our request.

I know this & I have no problem how this aspect is not a totally random roller coaster...in fact again I embrace & love it. Now it's your turn to carefully listen to our side again.

Yes, This age-development bell curve shows how when random talent increases/decreases occur they are not entirely random events at all. We get that, we get how as a player is young he is more likely (or in Statistical terms more frequent likelyhood) to get upward talent bumps, how if he's old he's more likely to get downward bumps, and how if he's injured he's more likely to get downward bumps. This bell curve determines the frequency of these random events...

However, our issue is how random the game is in choosing who gets a talent bump up or down. Warning: Using a latin/econometrics/statistics term here, Lets assume a situation of Ceteris Paribus, meaning all things being equal (IE Age, Position, injury status, clutch tendencies, team, their current stage on the age-development bell curve etc.), meaning the only things that are different from Player A compared to Player B is their stats, their talent potential, and their current ratings. As it currently is in the OOTP Game engine, WHO the game player-development engine chooses for receiving an upward tick or a downward talent tick is completely random...Remember this is occuring while Players A,B,C are at the same point on the age-development bell curve. I'm saying this is a problem. It shouldn't be this random!

It's an injustice that player A (14 HRs, HR Talent Fair, Ratings 2) who has the same exact stats as Player B (14 HRs, HR Talent Average, Ratings 4), but whose ratings bars are inferior to that of Player B, will have the same chance of getting an upward talent increase. There is currently no correllation between stat performance and true development opportunities (IE Talent Bar Increases). None, I'm not saying I want total correllation, I'm saying I want some correllation.

This means Player A is overachieving, so he gets 7% chance at a random talent increase, Player B who is performing at a statistical level that is expected of his talent/ratings will get the standard 5% chance, while Player C (A guy whose line reads 14 HRs, HR Talent Brilliant, and Rating 5) should get penalized -2% and will only have a 3% chance at a Talent increase in HRs.

Furthermore on the other end of the spectrum, Player C should have a 7% chance at a Talent Decrease in HRs, and Player B should have the standard 5% chance at a talent decrease, and Player A should only receive a 3% chance at a talent decrease.

You see how under this scenario now, there is now some correllation, while the randomness that makes this game so realistic still exists - just only now the game is more selectively random ie its AI is "Smarter" and more realistic.

Also notice how under this scenario, these probabilities adjustments are layered on top of the Age-Development bell curve, meanining it is taking that bell curve into account. The effect of this is as players age their chances (X%) for "Talent Increases" gradually drop across the board no matter if their overachieving or underachieving. Keeping the stats the same as above, now only changing the players ages to 25. This means Players A,B,C are now all older at age 25, Player A now has a 6% chance to for a talent increase, Player B has a 4% chance at a talent increase, while Player C has only a 2% chance at a talent increase. The correlation with stats-to-expectations still remains, while across the board the Age-development bell curve is taking affect as all players A, B, & C have had a -1% drop in their probabilities.

I hope you're still awake by now Henry. But this is about as clear, logical, and scientific as I can explain it to you. Do you get it yet ?

[Henry]

Originally posted by clarnzz
How so? This debate has been hashed for some time now and the point has been brought up by numerous people. I just found it surprising that after 3 threads and over 200 posts that you just now saw the other side of the table's point.

Oh well, back to training....

[clarnzz]

Originally posted by Henry
Oh well, back to training....

Easy there big fella, is more of a bad thing for people such as myself that were trying to argue that and couldn't express it in a manner for you to see what we were saying.

Please don't take it as an insult, I was just surprised more than anything.

[clarnzz]

The issue after that would be coming up with a way to implement it in a matter to where it couldn't be manipulated and exploited.

The additional chance would have to take into account, minor league level, age and current talent to be able to work I think.

EX:
Player A
Current talent: Brilliant home runs
Age: 23

Level: A ball
Homeruns: 40 (additonal chance of downgrade 5%, addtional chance of upgrade 0%)
Homeruns: 30 (additional chance of downgrade 10%, additional chance of upgrade 0%)
Homeruns: 20 (additional chance of downgrade 15 %)
Homeruns: 10 (addtional chance of downgrade 20 %)

Level AA:
Homeruns: 40 (additonal chance of downgrade : 0%)
Homeruns : 30 (addtional chance of downgrade : 5 %)
Homeruns : 20 ( additional chance of downgrade : 10 %)

Level AAA
Homeruns : 40 ( somehow more resisitant to downgrade)
Homeruns : 30 ( additonal chance of downgrade : 0%)
Homeruns : 20 ( additonal chance of downgrade: 5%)

Player B

Age: 20
Current Home Run Talent: Average

Level A:
Homeruns: 40 (additional chance of upgrade: 10%)
Homeruns : 30 (additonal chance of upgrade : 5%)
Homeruns : 20 (addtional chance of upgrade/downgrade: 0%)
Homeruns : 10 (addtional chance of downgrade : 5 %)
Homeruns : <10 (additional chance of downgrade : 10%)

Level AA

Homeruns :40 (additonal chance of upgrade :15%)


etc...........

I used a raw value rather than something like batting avg/slugging percentage as it can be manipulated by playing time. I do not suggest the percentages I added are correct, I am just trying to explain the fundamentals of the system as I would see it work.