Historical Random Debut

[actionjackson]

Originally Posted by Markus Heinsohn

That's not a bug, that's simply the way it works. Player stats are simply the result of individual ratings and the competition. Individual ratings are normalized to a neutral era (which is pretty simple mathematically), so the competition is the culprit here, and since it is random, funny things may happen at times. It's not unrealistic at all.

I'm sorry Markus, but I must respectfully disagree. Look particularly at this page of this thread. People posted their results using OOTP18, and I posted mine using a 60+ season long OOTP16 historical random debut. OOTP16 produced very realisic league-wide totals, and individual player seasons. OOTP18 had the very realistic league-wide totals, but there were a lot of super-duper individual seasons in it. Somewhere between OOTP16 and OOTP18 something changed, and it's been echoed many times both in that thread, and in this one.

[actionjackson]

Originally Posted by silvam14

That’s kind of what I’m thinking. If you are playing a game where Derek Jeter is facing Cy Young, how realistic can you expect things?


Sent from my iPhone using Tapatalk

Derek Jeter is facing Cy Young, but he's also facing a lot of Joe Schmoes. Everybody is in there, and the talent distribution is actually pretty comparable to a regular historical game because there's a huge range in talent. The high-end talent is (as is typical IRL) in short supply, and as the talent level drops, the pool increases. I use a five round draft in my RD leagues, and that's more than enough to keep them competitive. You get 44 players per team off the top, and that number keeps increasing, even with just five rounds worth of new players every year.

[JaBurns]

Originally Posted by actionjackson

I believe the weight distribution is static. In other words, with five year, double weighted, and the player's season being 1985, 1983 would be worth 1/6, 1984 would be worth 1/6, 1985 would be worth 1/3, 1986 would be worth 1/6, and 1987 would be worth 1/6. If that makes any sense. At least all the fractions add up to one, so my math isn't off.

No your math is 100%. I was thinking of the other ratings were it is x% potential, y% current year stats, z% stats one year ago, a% stats two years ago. I believe you may gotten Markus’ attention and it will be interesting to see what happens when I start on Monday with a new quest for my perfect way to play OOTP.
I am continuing my debate with myself on which year to lock stats, in many ways I don’t think it matters because I will change my thinking many times. I just got to make a decision and stay with it...

[pstrickert]

I ran some tests during beta and OOTP19 still seems to run a bit hot. It might be an improvement over OOTP18, however. In the past, when I've tried random debut leagues, I've always limited the range from 1920 to Present. Including dead ball era players (and earlier) seems to result in too many outliers. But maybe that's just me.

[actionjackson]

Originally Posted by pstrickert

I ran some tests during beta and OOTP19 still seems to run a bit hot. It might be an improvement over OOTP18, however. In the past, when I've tried random debut leagues, I've always limited the range from 1920 to Present. Including dead ball era players (and earlier) seems to result in too many outliers. But maybe that's just me.

I just can't bring myself to limit the field. No Cy Young. No Honus Wagner. No Nap Lajoie. No Ross Barnes. No Cap Anson etc etc etc. I have to go from Minimum year to Maximum year. I just can't seem to draw the line so I take 'em all.

Also, as long as it's better than OOTP18, that'll be a step in the right direction.

[JaBurns]

Thank you for the analysis, I don’t want to eliminate players, but is there an easy way to excluded dead ball era players?

Would one way to combat the hot issues be to modify the TCR? Just a thought, I am going to attempt a random debut league and hope for the best because it’s my preferred way to play in my solo leagues.

[Edster007]

Originally Posted by actionjackson

Come join the Historical Random Debut Geek Club. This is just one thread from the OOTP18 - Historical Simulations section of the forum. There are about 3 or 4 others down there. Come hang out with us. We'd be honoured to have such a modding legend in our presence.

Thanks to the link I spent most of Saturday reading that thread. Starting a HRD tomorrow. I have only played the app based version of this game but have a little more time to get back into computer baseball. Hope you're up for a few questions....

[JaBurns]

Originally Posted by Edster007

Thanks to the link I spent most of Saturday reading that thread. Starting a HRD tomorrow. I have only played the app based version of this game but have a little more time to get back into computer baseball. Hope you're up for a few questions....

These guys are great and have made my transition to random debut fun! I will answer questions that I can but the real gurus are David Watts, pstrickert, and actionjackson.

Welcome to the dark side as actionjackson says; my first advice would be don't overthink it, which is my biggest problem. You will probably have to go through may iterations to find what you want, saving your initial setup as a template helps me.....

[actionjackson]

Originally Posted by Edster007

Thanks to the link I spent most of Saturday reading that thread. Starting a HRD tomorrow. I have only played the app based version of this game but have a little more time to get back into computer baseball. Hope you're up for a few questions....

Always. If you don't get me, there are more of us. Another one for the dark side. Welcome.

[David Watts]

Originally Posted by Markus Heinsohn

That's not a bug, that's simply the way it works. Player stats are simply the result of individual ratings and the competition. Individual ratings are normalized to a neutral era (which is pretty simple mathematically), so the competition is the culprit here, and since it is random, funny things may happen at times. It's not unrealistic at all.

So when you look at baseball seasons throughout history, you see tons of guys getting 230+ hits a season 240, 250, 277? Sure, players have done so over history, but 200 hits is still the benchmark in real life. The batting races year in and year out feature 2 or more guys hitting .360 and above. .400 means very little. As shown in the bug forum this stuff doesn't only happen in random debut leagues either. Stats are hot in straight historical as well. And in one of those threads I posted a fictional guy. Can't remember how high his average was, but dude was hitting 70+ bombs a year while stealing 100+ bases. The great thing now is historical guys string together 2,3,4 or more super seasons, often at a level unseen in real baseball(put it this way, the real Miguel Cabrera's run from 2010-2013 would be nothing more than Lamar Mundane in a OOTP18 league). Of course the super years usually mean OOTP has to force suckitude on players at the end of their career.

I had Rogers Hornsby hit .485 in a straight historical with no development. Pretty sure it was a real lineups setup. I know it was a real transaction setup.

Like I said before, this wouldn't bug me as much if the game never had the ability to create realistic stats. It did. OOTP16, 15, 14 etc. But, over the last two versions something changed. I think a game so proud of it's ability to create league numbers so close to reality, would be embarrassed to produce individual seasons that make a mockery of real baseball's single season leaderboards. And please don't think I'm against outlier seasons. I just detailed two of them(George Scott and Greg Luzinski in my Southern League thread in the Historical section). Outlier seasons make this game incredible. But, it loses it's greatness when in order for George Scott to have that outlier season he has to string together 3 or more seasons hitting above .350.

Ugh, sorry for whining once again. It just sucks to have your favorite way of playing the game ripped out from under you.

[David Watts]

Originally Posted by pstrickert

I ran some tests during beta and OOTP19 still seems to run a bit hot. It might be an improvement over OOTP18, however. In the past, when I've tried random debut leagues, I've always limited the range from 1920 to Present. Including dead ball era players (and earlier) seems to result in too many outliers. But maybe that's just me.

I never use players from before 1901. In my current league I used 1905 as my minimum(I can't ever deny myself Mr. Cobb) and I really like what I'm seeing. In the past when I used guys from prior to 1901, I found the pre 1900 pitchers to be super heroes. I begged for the minimum and maximum feature back when random first came out. Screamed like kid on Christmas when Markus said he would add it to the game.

[Markus Heinsohn]

Originally Posted by David Watts

So when you look at baseball seasons throughout history, you see tons of guys getting 230+ hits a season 240, 250, 277? Sure, players have done so over history, but 200 hits is still the benchmark in real life. The batting races year in and year out feature 2 or more guys hitting .360 and above. .400 means very little. As shown in the bug forum this stuff doesn't only happen in random debut leagues either. Stats are hot in straight historical as well. And in one of those threads I posted a fictional guy. Can't remember how high his average was, but dude was hitting 70+ bombs a year while stealing 100+ bases. The great thing now is historical guys string together 2,3,4 or more super seasons, often at a level unseen in real baseball(put it this way, the real Miguel Cabrera's run from 2010-2013 would be nothing more than Lamar Mundane in a OOTP18 league). Of course the super years usually mean OOTP has to force suckitude on players at the end of their career.

I had Rogers Hornsby hit .485 in a straight historical with no development. Pretty sure it was a real lineups setup. I know it was a real transaction setup.

Like I said before, this wouldn't bug me as much if the game never had the ability to create realistic stats. It did. OOTP16, 15, 14 etc. But, over the last two versions something changed. I think a game so proud of it's ability to create league numbers so close to reality, would be embarrassed to produce individual seasons that make a mockery of real baseball's single season leaderboards. And please don't think I'm against outlier seasons. I just detailed two of them(George Scott and Greg Luzinski in my Southern League thread in the Historical section). Outlier seasons make this game incredible. But, it loses it's greatness when in order for George Scott to have that outlier season he has to string together 3 or more seasons hitting above .350.

Ugh, sorry for whining once again. It just sucks to have your favorite way of playing the game ripped out from under you.

As I said, it's always a matter of how the league is made up, what players are in the league, what is the competition.

Regarding the mode in comparison to OOTP 16, there was no change that I am aware of, other than the usual tweaks to AI or league LTM calculations and defensive ratings. The ratings neutralization was not touched, so player ratings should still be very similar. I can absolutely see no reason for outliers other than small sample size or simply bad luck because the competition in the league is weak by chance.

Edit: A few tips:
-Exclude players prior to 1901
- Always use "career mode", so that at least 3-year ratings recalc is used.
- Set fielding ratings recalc period to 3 years, too.
- Play around with the "adjust players minimum AB" etc settings

I just tested a league (1984 stats settings) and simmed a few years in OOTP 19 and everything looks fine to me.

[David Watts]

Originally Posted by Markus Heinsohn

As I said, it's always a matter of how the league is made up, what players are in the league, what is the competition.

Regarding the mode in comparison to OOTP 16, there was no change that I am aware of, other than the usual tweaks to AI or league LTM calculations and defensive ratings. The ratings neutralization was not touched, so player ratings should still be very similar. I can absolutely see no reason for outliers other than small sample size or simply bad luck because the competition in the league is weak by chance.

No problem, as long as Windows doesn't make OOTP16 obsolete I can make do.

So, Rogers Hornsby played against competition so much different than he did in real life while participating in a real transactions league, that he was a .485 hitter?

Shooting fish in a ocean here, but could it be a lack of Lahman issue? Wonder if we were able to plug the Lahman into 18 if we would be getting the same monster seasons. Who knows, but denying nothing changed is like denying the existence of taxes, Oprah and butter.

[Lukas Berger]

Originally Posted by Markus Heinsohn

- Play around with the "adjust players minimum AB" etc settings

As I've mentioned before when discussing this with you guys, this is a big thing to get right, even more so in a random debut league because of how the talent can vary. When you're playing a random debut league the default may not be the best option.

If you want to limit outliers and keep the top players from performing really well in a random debut league, then you'll want to be sure to set this pretty low.

[Lukas Berger]

Here's an example of what Paul got simming out a historical game a few builds ago in the OOTP19 beta:

Originally Posted by pstrickert

1901-2016 / 1-yr recalc / real lineups & transactions
___________________________________________

.400+ AVG

real: 13x
sim: 12x (last time: 1947)
____________________________________________

60+ HR

real: 8x
sim: 3x (high: 70)
_____________________________________________

600+ HR (career)

real: 9x
sim: 9x (high: 729)
_____________________________________________

3000+ Hits

real: 31x
sim: 37x (high: 4309)
_____________________________________________

150+ RBI

sim: 45x
real: 22x (high: 179)
_____________________________________________

350+ K (pitching)

real: 5x
sim: 7x (high: 379)
_____________________________________________

100+ G (pitching)

real: 1x
sim: 3x (high: 101)
____________________________________________

So as you can see, the stats for regular historical play aren't running particularly high in OOTP19.

Now whether that's true for your specific random debut method of playing with the 1984 statistical environment, I have no idea. As as I know you're the only person who plays that specific way and I don't think anyone has tested that out during beta.

[Markus Heinsohn]

Originally Posted by David Watts

No problem, as long as Windows doesn't make OOTP16 obsolete I can make do.

So, Rogers Hornsby played against competition so much different than he did in real life while participating in a real transactions league, that he was a .485 hitter?

Shooting fish in a ocean here, but could it be a lack of Lahman issue? Wonder if we were able to plug the Lahman into 18 if we would be getting the same monster seasons. Who knows, but denying nothing changed is like denying the existence of taxes, Oprah and butter.

Always use career mode for random debut. It may be that replay mode is set as default for you, if it is, change it. If you have replay mode for random debut you will get outliers because a guy who hit .350 in 200 AB will get ratings which are too high. This does not happen in career mode where the recalc base is 3 years.

And Hornsby had a well balanced league as competition in real life. In random debut it may happen that he plays against many more scrubs (if you got unlucky with the import)...

[Lukas Berger]

Originally Posted by Markus Heinsohn

Always use career mode for random debut. It may be that replay mode is set as default for you, if it is, change it. If you have replay mode for random debut you will get outliers because a guy who hit .350 in 200 AB will get ratings which are too high. This does not happen in career mode where the recalc base is 3 years.

I think he is using career mode though...

From what I understand, the issues David is getting don't tend to be with low ab guys from rl having great seasons either, it's more with already pretty great guys who got lots of abs having even better seasons like Hornsby, Ron Santo, George Scott types.

[Jeff1787]

"Always use career mode for random debut. It may be that replay mode is set as default for you"

Where is this option: Career mode vs. Random mode??

[Markus Heinsohn]

Originally Posted by Jeff1787

"Always use career mode for random debut. It may be that replay mode is set as default for you"

Where is this option: Career mode vs. Random mode??

First step in the league creation wizard, below the option where you enable random debut.

[Markus Heinsohn]

OK, ran a test in OOTP 19... random debut, career mode, simmed from 1970 to 1980. I had only 5 players hitting .350+ for a season (Ed Delahanty had the best and 2nd best at .374 and .372, the others were Gehrig, McGriff, Babe Herman). All other stats look great, too. No unreasonable outliers. HR high was 49 (Mays).