Random debut league problems

[nyy26wc]

I've been running a test historical league, using Gambo/Spritze and random debut. I'm using neutralized stats.

I'm seeing a couple of problems--

1) The players aren't debuting at the right ages.

Some examples--

Lou Gehrig--the database has neutralized stats for him, starting at age 18. Random debut had Gehrig come into the league at the age of 22.

Nolan Ryan--database has neutralized stats for him starting at age
Random debut has him starting at age 22.

Ty Cobb--database has him starting at age 19. Random debut has him starting at 22.


2) The league is dominated by players from the PCL who either never made it to the majors or they only had a cup of coffee.

The league started in 2012. So, 2012 leaguewide stats and finances were frozen in time and have been used for the entire league.

It's 2035. In the first 23 years of the league, there have been 6 players with 600+ HR. Mark McGwire is the all time leader, with 823. What makes him stand out as an outlier isn't as much his total, but the fact that he's the only one of the 6 600 HR hitters who actually homered in the majors.

For example, Larry Lejeune had a cup of coffee with the 1911 Dodgers and 1915 Pirates. He hit the whopping total of 0 HR in his major league career, spanning 24 games. His neutralized stats, from the PCL, are 224 HR. Lejeune is tied for 3rd, with 625. He's in a too close to call race to McGwire's runnerup. The players who are 2nd through 5th on the all time HR list all have between 623 and 626 and are all still active.

I'm going to post Lejeune's screenshots and then have more comments about what it looks like what is happening.

[nyy26wc]

Here's Lejeune's neutralized stats and his OOTP stats.

[nyy26wc]

Isn't neutralized stats supposed to convert what he did in X year and then translate that to today's world?

But, it looks to me like the database converted Lejeune's stats from deadball standards and gave us a neutralized set of stats.

Then, when it came time to import his stats, OOTP looked at that neutralized stats, saw the year he was doing it in, assumed that he was actually reaching those HR figures during the deadball era and then used that as a basis to make him one of the greatest HR hitters of all time.

It sure to me like basing ratings on neutralized stats is having the effect of doubling neutralization.

The league is set to recalc each season, based on historical stats. Here are those settings.

[Spritze]

The Random Debut function auto-neuters the stats it wants to so don't use pre-neutered stats. Double neutering is not necessarily necessary nor optimal. My ex-wives might beg to differ though.

[nyy26wc]

Originally Posted by Spritze

The Random Debut function auto-neuters the stats it wants to so don't use pre-neutered stats. Double neutering is not necessarily necessary nor optimal. My ex-wives might beg to differ though.

I'm running a 2nd set league right now, with all of those other settings, but instead using real stats and not neutralized. The results this time are better.

But, then, when I run my actual league, it's a random debut league, but not a regular random debut league. I want my draft class to be a hybrid of fictional and random debut. So, OOTP will determine the fictional draft classes and I will be importing the IDs for the random debut players through the historical import feature.

So, I presume that, when I do it that way, that's when I would want to use neutralized stats. Is that correct?

[The Game]

Originally Posted by Spritze

The Random Debut function auto-neuters the stats it wants to so don't use pre-neutered stats. Double neutering is not necessarily necessary nor optimal. My ex-wives might beg to differ though.

Did not know this.

[nyy26wc]

Originally Posted by Spritze

The Random Debut function auto-neuters the stats it wants to so don't use pre-neutered stats. Double neutering is not necessarily necessary nor optimal. My ex-wives might beg to differ though.

I also assume that, in the 1st sentence, you had meant "so don't use already-neutered stats."

I'd think that the real stats would be the "pre-neutered" ones.

[VanillaGorilla]

Originally Posted by nyy26wc

I've been running a test historical league, using Gambo/Spritze and random debut. I'm using neutralized stats.

I'm seeing a couple of problems--

1) The players aren't debuting at the right ages.

Some examples--

Lou Gehrig--the database has neutralized stats for him, starting at age 18. Random debut had Gehrig come into the league at the age of 22.

Nolan Ryan--database has neutralized stats for him starting at age
Random debut has him starting at age 22.

Ty Cobb--database has him starting at age 19. Random debut has him starting at 22.


2) The league is dominated by players from the PCL who either never made it to the majors or they only had a cup of coffee.

The league started in 2012. So, 2012 leaguewide stats and finances were frozen in time and have been used for the entire league.

It's 2035. In the first 23 years of the league, there have been 6 players with 600+ HR. Mark McGwire is the all time leader, with 823. What makes him stand out as an outlier isn't as much his total, but the fact that he's the only one of the 6 600 HR hitters who actually homered in the majors.

For example, Larry Lejeune had a cup of coffee with the 1911 Dodgers and 1915 Pirates. He hit the whopping total of 0 HR in his major league career, spanning 24 games. His neutralized stats, from the PCL, are 224 HR. Lejeune is tied for 3rd, with 625. He's in a too close to call race to McGwire's runnerup. The players who are 2nd through 5th on the all time HR list all have between 623 and 626 and are all still active.

I'm going to post Lejeune's screenshots and then have more comments about what it looks like what is happening.

Gehrig and Ryan both entered my league at 18 and 19, respectively. Check your age limits and also check to see if they were not drafted multiple times and did not sign. ADD: Since the three players you cite all entered the league at age 22, I lean towards an age minimum setting being the culprit.

For me, Cobb came in at 34 in the Inaugural Draft.

Re Lejeune, IDK WTF is going on with him in your league. Lejeune had 495 AB in my league, which is a 2012 environment with lower pitching mounds, and hit exactly 0 HR.

Once you reach 2012, turn off your "adjust stats" option, if you want the league to "evolve". Otherwise, any changes to mound height, or whatever, will be factored out. If you want your league output to stay constant from 2012 onwards, leave it checked.

[VanillaGorilla]

Forgot to attach this in above post.

My Lejeune:

[nyy26wc]

Originally Posted by VanillaGorilla

Gehrig and Ryan both entered my league at 18 and 19, respectively. Check your age limits and also check to see if they were not drafted multiple times and did not sign. ADD: Since the three players you cite all entered the league at age 22, I lean towards an age minimum setting being the culprit.

I just checked their player histories--

Gehrig: #1 pick in the entire 2017 draft, signed the next day
Ryan: 1st round pick in the 2030 draft, signed the next day
Cobb: #1 pick in the entire 2029 draft, signed the next day

So, it's not they were drafted multiple times and didn't sign.

I just checked and confirmed that there were no age minimum or maximum settings.

It was a coincidence that those 3 examples all came into the league at 22. Other examples that I just found were Mark McGwire (came into the league at 18, his neutralized stats start at age 20), Ron Guidry (came into league at 19, neutralized stats start at 20) and in the most recently held draft, Ernie Banks (came into league at 17, neutralized stats start at 19).

I've been just looking up players off the top of my head and I can't find any examples of players who went into the draft at the same age that they were when their neutralized stats started.

Once you reach 2012, turn off your "adjust stats" option, if you want the league to "evolve". Otherwise, any changes to mound height, or whatever, will be factored out. If you want your league output to stay constant from 2012 onwards, leave it checked.

For the test league, I was just using 2012 as a constant, to remove that as a variable. I still haven't decided exactly what I want to do for the league stats for when I start the league.

[nyy26wc]

Originally Posted by VanillaGorilla

Forgot to attach this in above post.

My Lejeune:

I don't understand how the same player can produce different neutralized stats.

[VanillaGorilla]

Originally Posted by nyy26wc

Other examples that I just found were Mark McGwire (came into the league at 18, his neutralized stats start at age 20), Ron Guidry (came into league at 19, neutralized stats start at 20) and in the most recently held draft, Ernie Banks (came into league at 17, neutralized stats start at 19)..

THIS has me very interested (and quite geeked).

My McGwire entered at 21.

My Guidry entered at 24.

Both of them signed with the team that first drafted them.

Banks has not entered, yet.

I really like the idea that there is some randomness in the random debut feature as to when the players are chosen to randomly debut actually debut (in regards to their age).

I have never seen any documentation or forum discussion from the team mentioning this. Nonetheless, I find this being a cool feature. If Ichiro usually debuts around 27, but can debut at, say, 20....THAT is very cool! Just checked, he came into the draft at age 26 for my current league.

[nyy26wc]

Originally Posted by VanillaGorilla

I really like the idea that there is some randomness in the random debut feature as to when the players are chosen to randomly debut actually debut (in regards to their age).

I have never seen any documentation or forum discussion from the team mentioning this. Nonetheless, I find this being a cool feature. If Ichiro usually debuts around 27, but can debut at, say, 20....THAT is very cool! Just checked, he came into the draft at age 26 for my current league.

I believe that it's the opposite, that there is only supposed to be randomness in ages for the inaugural draft. But, for the annual draft, since it is a "rookie draft" and you have imported "historical rookies", everyone there comes into the league at the age they were historically when they were a rookie. Then, for this purposes, "rookie" is defined as the first year they have a stat line in the database. If you use the OOTP database, it's the year they made their MLB debut. If it's the Gambo/Spritze database, it's the first year they have a statline for the player.

[nyy26wc]

I decided to run a test. I started a new random debut league in 2012. I simmed to the day after the 2012 draft and then looked at the ages of the 1st round draftees, comparing them to their real life 1st year in the majors.

Let's go crazy and look at the whole entire first round.

The first number will be age drafted in the OOTP draft, the second number will be age in his IRL year in the majors.

First up, a league using the OOTP default database--

1) Barry Bonds, 19, 21
2) Roger Connor, 20, 22
3) Bill Joyce, 22, 24
4) Home Run Baker, 20, 22
5) Beals Becker, 19, 21
6) Joe Harris, 21, 23
7) Bobby Mitchell, 19, 21
8) Cliff Johnson, 22, 24
9) Roger Clemens, 19, 21
10) Herb Score, 20, 22
11) Benny Kauff, 20, 22
12) Jim Burns, 20, ????
13) Mo Vaughn, 21, 23
14) Paul Molitor, 19, 21
15) Frank Fennelly, 22, 24
16) Rob Dibble, 22, 24
17) Sandy Griffin, 23, 25
18) Pedro Guerrero, 20, 22
19) Ryan Braun, 21, 23
20) Chief Meyers, 26, 28
21) Rico Carty, 21, 23
22) Julio Franco, 21, 23
23) Cal Ripken, 18, 20
24) Al Wingo, 19, 21
25) Ray Grimes, 24, 26
26) Chief Zimmer, 21, 23
27) John Montefusco, 22, 24
28) Turk Farrell, 20, 22
29) Robin Ventura, 19, 21
30) Homer Hillebrand, 23, 25

We don't have date of birth for Jim Burns, who is one of the 19th century players with missing data. But, among the other 29 players, all 29 of them entered the random debut draft exactly 2 years younger than they were for their 1st major league season. 29 out of 29 can not possibly be explained by random chance.

So, I was wrong about players entering the random draft at the same age they debuted in the majors. They enter the draft 2 years younger.

Next up, I'll see what happens if I run the same test, this time using the Gambo/Spritze database.

[nyy26wc]

Here's the results, using the Gambo/Spritze database.

1) Jim Kennan, 22, 17
2) Jim Thome, 21, 20
3) Chipper Jones, 22, 21
4) George High Pockets Kelly, 21, 19
5) Magglio Ordonez, 22, 23
6) Bug Holliday, 18, 22
7) Johnny Bench, 19, 21
8) Will Clark, 19, 22
9) Ray Grimes, 24, 26
10) Chuck Knoblauch, 18, 22
11) Moose McCormick, 21, 23
12) Jimmy Wood, 25, 27
13) Chief Zimmer, 21, 23
14) Al Wingo, 19, 21
15) Dale Murphy, 22, 20
16) Randy Milligan, 17, 25
17) Larry Doby, 21, 23
18) Jim Wynn, 18, 21
19) George Pinkney, 23, 22
20) Ed Morgan, 22, 24
21) Johnny Damon, 22, 21
22) Doc Gessler, 20, 22
23) Yogi Berra, 22, 21
24) Lou Whitaker, 22, 20
25) Dizzy Dean, 18, 20
26) Carlos Santana, 17, 24
27) Hank Leiber, 20, 22
28) Brett Lawrie, 17, 21
29) Olmedo Saenz, 17, 23
30) Alvin Davis, 19, 23

No fixed rule here about players being 2 years younger than they were in their real first year.

But, there's also a case of 30 players out of 30 all debuting in the draft at an age other than their 1st year in the majors. That can't be just attributed to random chance.

[nyy26wc]

These tests also raise another question. If the player is debuting in OOTP at a younger age than he has MLB stats, what stats is OOTP using to determine his ratings at that age?

For example, Barry Bonds debuted as a 19 year old in the OOTP debut draft, but didn't have any MLB stats to base ratings on until he was 21. Does that mean that his 19 year old season get his 1st year's stats (when he was really 21), his 20 year old season gets his 2nd year's stats (when he was really 22) and so on? (For these questions, assume that recalc is on.)

And, the case of Chipper Jones is even more interesting. Jones made his MLB debut as a 21 year old, missed the entire season as a 22 year old and then returned as a 23 year old. In the draft using the Gambo/Spritze database, the age chosen for him to debut was the age in his career that he missed the whole season.

[VanillaGorilla]

Originally Posted by nyy26wc

But, there's also a case of 30 players out of 30 all debuting in the draft at an age other than their 1st year in the majors. That can't be just attributed to random chance.

Yes it can. My Guidry entered at 24, just as IRL.

-------

This is awesome work! I never even considered any of this, before.

[nyy26wc]

Originally Posted by VanillaGorilla

Yes it can. My Guidry entered at 24, just as IRL.

-------

This is awesome work! I never even considered any of this, before.

Let's look at the Guidry case.

Guidry entered at 24 IRL. Well, sort of. Guidry was born in 1950, he debuted in 1975. So, during his first season in the majors, his "baseball age" was 24, but by the end of the season, he was 25.

The key question here might be when do you hold your draft? Do you hold the draft at the end of the season?

Let's say that you hold the draft on October 1 or November 1. In that case, if age is computed based on the player's age on October/November 1, Guidry's age for his debut year was 25.

In order to make sure that "baseball age" and actual age were the same things, I set the date of both of these drafts to June 1.

As far as the 30 out of 30 not being the same, I was wrong about it couldn't be just random chance. I just decided to just click on a few random 2nd rounders, just to see if maybe I could find one who was the Guidry of my league. (Maybe I also did that just to proscrastinate a little longer before I have to get to sleep.) I did find 1--Johnny Damon.

So, in the light of all of these other cases, maybe Damon and Guidry are just among the outliers of our respective leagues.

Also, what database where you using? This list shows constant variation in Gambo/Spritze, but an absolute rule of 2 years younger with OOTP default. If it turns out that my OOTP default is producing an absolute rule and yours isn't, then we have another mystery on our hands.

[VanillaGorilla]

Originally Posted by nyy26wc

Let's look at the Guidry case.

Guidry entered at 24 IRL. Well, sort of. Guidry was born in 1950, he debuted in 1975. So, during his first season in the majors, his "baseball age" was 24, but by the end of the season, he was 25.

The key question here might be when do you hold your draft? Do you hold the draft at the end of the season?

Guidry has a birthday of Aug 28. My draft occurs around Nov 1. He was 24 when drafted and his rookie season is also as a 24 year-old.

Originally Posted by nyy26wc

Also, what database where you using? This list shows constant variation in Gambo/Spritze, but an absolute rule of 2 years younger with OOTP default. If it turns out that my OOTP default is producing an absolute rule and yours isn't, then we have another mystery on our hands.

Am using OOTP "Garlon and Spritze"...OOTP default with nuetralized stats selected.

[David Watts]

Ugh, I am out of town and can't look at my game until Monday. This is going to bug me all weekend.