alternate historical question

[thehef]

Let's say I wanted to start a historical game, with minors (for the sole purpose of accessing minor league players; I don't care otherwise what will happen in the minor leagues). Let's say I want to start in 1955, and after the initial import of players to historically correct teams, I will turn off historic evolution. Then I'll create, say, four more teams and put two each in the AL & NL. For these four teams I'll populate them with the best of the best players toiling in the minors and some free agents. Then I make sure the AL & NL have appropriate schedules, maybe make a few other tweaks...

So my question: For best results, do I need to do anything with League Totals, Modifiers, the era_stats or era_modifiers files, etc.? Or just roll with it?

Thanks in advance.

[Brad K]

That's a real hard one to think through. Because you have four complete teams of all AAA players you're actually going to have 20% of the playing time going to AAA. In a league where the expansion teams draft from the existing teams the expansion teams get some players with MLB experience for the starting lineup. The same number of AAA players come up as in your scenario but they're distributed among all the teams mostly in non starting roles.

OOTP doesn't perform well in some situations at the extremes. I think you'd see that if you played with this setup, but can't predict what it would be.

[thehef]

So I did a test, starting with 1953...

- Initial import of players to historically correct teams
- I moved the LA Angels and Seattle Rainiers from the PCL to the AL, and the Hollywood Stars and San Francisco Seals to the NL.
- For all four teams above, I did my best to align their rosters as they were in 1953 (OOTP has some bugs re player import, so some players weren't available; in those cases, I plucked a similar player from some other minor league.)
- I found schedules that would work for the updating alignments of the AL, NL, and PCL, and imported those. (162 game season for MLB.)
- I did nothing with League Totals, Modifiers, the era_stats or era_modifiers files...
- 1-year recalc, no development, injuries very low, position player fatigue high; Rule 5 draft & trading set to Off. Everything else set to defaults other than things that don't matter (such as baseball cards, uniforms, WPA graphs...).
- I simmed the game one month at a time, manually making significant historical transactions.

Results highlights:
- Standings are as I expected: Angels were bad (53-109); the other three "expansion" teams were atrocious (45 wins for Hwd, 44 for SF, 30 for Seattle). Dodgers and Yankees won their leagues, with somewhat higher win totals (112 and 107, which I attribute to the existence of two very poor teams in each league).
- Leaderboards looked normal. A few outliers here and there, but nothing that I wouldn't see in a strict replay.
- With the caveat that I have little understanding of things like League Totals, Modifiers, the era_stats or era_modifiers files, the Historical Simulation Accuracy panel showed all green numbers for 1953. Curiously, though, after an initial look at this - immediately after having advanced to the offseason - the data on this panel went away.

At any rate, it's a small sample size and I may find some things that don't look right if I do more digging, but so far, this seemed to work pretty well. Gives me some ideas for some alternate world replays I might try...

In fact, I might try this same thing again, except I'd make the remaining PCL teams (Oakland, San Diego, Sacramento, Portland) the farm teams of the four PCL teams that I placed in the majors. Perhaps having double the volume of AAA-level players, those four teams I placed in the majors might ascend to actually expansion-level, as opposed to being so terrible... We shall see...

[luckymann]

Interesting hef. Did you just delete the rest of the minor leagues / teams / players as well as turn off auto evolution?

[Brad K]

As with W/L records, the spread between the stats of the best and worst players should be greater than historical. Concerning player stats looking normal, try an analysis by groups, for example, the players ranked 6 to 25 in home runs.

[thehef]

Originally Posted by luckymann

Interesting hef. Did you just delete the rest of the minor leagues / teams / players as well as turn off auto evolution?

Nope. I turned off auto-evolution (which, IMO, since it enompasses so much more, is mislabeled as "Automatically Expand League"), but I left the other minor leagues as is. In some cases, when I was moving around some players, if I stumbled upon a key player who should've been on a high-level (AAA or AA... Int'l Lg, American Assoc, Texas Lg, etc.), I would move him to that team. But generally I just left those other minor leagues alone.

Originally Posted by Brad K

As with W/L records, the spread between the stats of the best and worst players should be greater than historical. Concerning player stats looking normal, try an analysis by groups, for example, the players ranked 6 to 25 in home runs.

Will do, and will report back.

[Brad K]

Please do. It will be interesting and informative regardless of the results.

[thehef]

Top 25-ish HR leaders actually looked really good. In cases where a guy's OOTP HR totals were significantly higher than IRL, it was mostly because he received more AB's than IRL. Two guys near the bottom of the list below - Kelleher & Davis - IRL played only partial seasons in AAA. So based upon what they did IRL, their OOTP totals seem pretty realistic to me.

I'll check out another Top 25 for another category if ya want. Just LMK.

[Brad K]

Would you mind posting their ABs? Thanks.

[thehef]

Originally Posted by Brad K

Would you mind posting their ABs? Thanks.

Sure thing. I just updated the image in the post above.

[Brad K]

Kelleher and Davis omitted from analysis due to lack of real life MLB play.

[thehef]

Originally Posted by Brad K

Kelleher and Davis omitted from analysis due to lack of real life MLB play.

Re that analysis, nicely done. Thanks!

So what's your interpretation of those numbers? I'm thinking it turned out pretty good, with the idea being that the top HR hitters should hit a tad more HR's since, roughly speaking, 22% of the time they are facing pitching that can be loosely thought of to be AAA level.

[Brad K]

I think the top hitters should hit a few more as they did.

But look at this, how many RL top 25 HRers are missing from the top 25 in the game. 1/2 short of half!

[Brad K]

OH, I forgot. I'm supposed to have advertising.

The above posts sponsored by...

THE NEW MOTORAMIC CHEVROLET

[thehef]

Originally Posted by Brad K

I think the top hitters should hit a few more as they did.

But look at this, how many RL top 25 HRers are missing from the top 25 in the game. 1/2 short of half!

Damn. I wanted to go back and see how those guys - the ones missing from the list - did, but I accidentally deleted it while doing some disk cleanup. I think I have a backup from the beginning of the season and can re-run though. Stay tuned...

[thehef]

Originally Posted by thehef

Damn. I wanted to go back and see how those guys - the ones missing from the list - did, but I accidentally deleted it while doing some disk cleanup. I think I have a backup from the beginning of the season and can re-run though. Stay tuned...

I re-ran my 1953 test and my HR leaders looked very similar to the previous test. As for the guys you noted who were missing...

Originally Posted by Brad K

But look at this, how many RL top 25 HRers are missing from the top 25 in the game. 1/2 short of half!

I think you were looking at the wrong year. Your numbers are for 1955, not 1953, so most of those guys that you have as missing IRL 1953 either didn't play in the majors or barely played.
EDIT: Which is understandable because in my OP I gave 1955 as the "for example"... then later, when I decided to actually test it, I went with 1953, which I mentioned but I can see how that wouldn't be clear enough. Sorry for any confusion

[Brad K]

1953

[Brad K]

1953 top 26 real life vs game

[Brad K]

Originally Posted by thehef

I re-ran my 1953 test and my HR leaders looked very similar to the previous test.

If you're willing to post the numbers for that I could analyze the additional sample. Also we'd have the full game data on the real life leaders.

[thehef]

Originally Posted by Brad K

If you're willing to post the numbers for that I could analyze the additional sample. Also we'd have the full game data on the real life leaders.

Here ya go. I had advanced the game to the late-1954 offseason for other testing purposes. So while the stats below are for 1953, Frankie Kelleher does not appear even though he hit 22 HR's. That's because he retired after 1953. So there may be others who don't appear on the list because they, too, retired. If you find anyone missing, LMK and I'll look them up... Also, included the next 26 so you don't have to leave the lines blank for some guys who don't appear in the first 26... (And while I'm certainly not asking you to do any work, if you want the 1954 sim numbers, I'll give 'em to ya, too )