Very nice.

Posted by contrarian23 on 7/22/2023 8:19:00 PM (view original):

Posted by cubbies84 on 7/22/2023 12:06:00 PM (view original):
i just dont understand how 100 games in teams are what they are, except when they arent.

I just built a table in Excel to simulate 100 seasons of a team with a "true" winning percentage of .550.

Here are just a few examples:
Season 1:
Record after 100 games: 58-42 (.580)
Record in games 101-162: 32-30 (.516)
Season 3:
Record after 100 games: 47-53 (.470)
Record in games 101-162: 40-22 (.645)
Season 4:
Record after 100 games: 57-43 (.570)
Record in games 101-162: 32-30 (.516)
Season 7:
Record after 100 games: 57-43 (.570)
Record in games 101-162: 28-34 (.452)
Season Season 9:
Record after 100 games: 59-41 (.590)
Record in games 101-162: 26-36 (.419)
Season 13:
Record after 100 games: 56-44 (.560)
Record in games 101-162: 43-19 (.694)

I could go on and on and on...those are just from the first 13 sets of 162 games.

This is as pure a simulation as you can create. No accounting for fatigue or quality of opposition or team matchups or manager decisions or anything - just run the random variable over and over again.

Anyone who thinks there is ANYTHING unusual about a simulated baseball team's performance fluctuating susbtantially after 100 games needs to learn some basic probability and statistics.

interesting, but isnt this sim accounting for fatigue and just lousy player selection or lousy park factors? I get i can lose any game at any time, but my question is how does a team playing .700 ball, playing one at .350 with fatigued pitchers in a park that doesnt suit their players suddenly sweep? or that it happens two series in a row with those same variables. Everyone and everything has ratings, these ratings when properly deployed should give some advantage over another team who hasnt done is as well? no?

Sure. And over thousands and thousands of games those things would lead to a demonstrable advantage. But over 3 or 6 or 10 or 20 games the random nature of statistical processes can easily outweigh them. Heck even 162 games is a small sample size when it comes to the number of variables that play into the SLB algorithm, as schwarze’s post above should make clear.

We have seen the worst team in baseball sweep the best team in baseball many times in real life. We have seen an 83 and 82 win team make the world series (Cards and Mets respectively) with the Mets winning the world series. We see it more in the Sim because the sim plays many more games than real life. We have 24 seasons going on at one time. Both in real life and the Sim, the randomness in small sample sizes wins out. It is not who you play but, when you play them.

I'm just 40-38 in the current 162-game cycle. Looks like regression has arrived.

In today's pm session, my $255M won 6-4 (total of 10 runs).

My other five games produced a total of 18 runs, combined.... 0-1 (L), 4-2 (W), 1-3 (L), 0-1 (L), 5-1 (W)

My $70M team started 51-41, scoring more than 3 runs in 52 of 92 games. Since then, 10-25, scoring more than 3 runs in 7 of 35. But ya, nothing wrong with the SIM.

I rarely focus on streaks. There are too many variables to make any really meaningful evaluation. Does a current losing streak coincide with playing higher quality teams? Did your earlier winning occur b/c your opponents were playing their weaker pitchers, and now that the season-long inning fatigue has somewhat dissipated and now those teams aren't throwing those pitchers as much. Was your team lucky in 1-run games and now that luck has regressed?

But mainly, like was mentioned earlier, 100 games (even 162 games) really isn't a big enough sample to determine the true strength of team. If I go 50-30 in my first 80 games, then go 30-50 in my next, who's to say my team isn't more than a .500 team that just got lucky early? Playing my same team vs the same exact opponents in another 162-game season might see my team go 40-40 each half, or even 30-50 in the first half, and 50-30 in the second. As I said before, this variability has been proven with my "same roster challenge" theme. Imagine watching 24 teams, with the exact players play each other in a 162-game season and seeing teams win as many as 95 games or as few as 65 games. Would you consider the sim broken in that scenario?

Dateline July 23rd. Tampa Bay GM Peter Bendix filed a lawsuit against Major League Baseball today claiming, "This game is rigged. On June 30th we were 57-28, since then we've gone 4-14 to blow our division lead and fall into second place. How can a team play better than .667 ball for more than half the season and then suddenly we can't beat anyone? This makes no sense whatsoever, and we intend to get to the bottom of it in court."

Posted by contrarian23 on 7/26/2023 12:11:00 PM (view original):
Dateline July 23rd. Tampa Bay GM Peter Bendix filed a lawsuit against Major League Baseball today claiming, "This game is rigged. On June 30th we were 57-28, since then we've gone 4-14 to blow our division lead and fall into second place. How can a team play better than .667 ball for more than half the season and then suddenly we can't beat anyone? This makes no sense whatsoever, and we intend to get to the bottom of it in court."

The 2022 Dodgers also demand justice. No way a team can win 111 games in the regular season and lose in the first round of the playoffs.

I have a team in juice's league, that has put together the following streaks. (The comments in parentheses is how I felt about the team at that point in time)

Started 9-6 over first 15 games (Hey, this team is pretty good)
Went 7-12 over next 19 games - Cumulative record 16-18 (Meh, probably just a .500 team)
Lost 12 straight - Cumulative record 16-30 (I guess I overestimated how good this team is)
Went 5-6 over next 11 games - Cumulative record 21-36 (Yep, this team sucks)
Went on an 18-1 tear, including 14 in a row - Cumulative record 39-37 (I was unlucky early in the season, this team is decent)
Went 8-5 over next 13 games - Cumulative record 47-42 (In first place now, I knew this team was good)
Went 7-17 - Cumulative record 54-59 (What the f*ck - I thought this team was good)
Currently on an 8-1 run - Cumulative record 62-60

Maybe this team is really a .500 team. Who knows. Streaks happen. The sim is not broken b/c of winning/losing streaks (although don't get me started on modern day RPs)

My last 3 cycles: 1-5, 1-5, 1-5.

Perhaps the correction has arrived?

A week ago we finally made it into the cage. It's been a slow death since. Sure was nice to visit, tho.

I really hate my $70M team...

Here's one for you Jtpsops - my AL leading 110 mil team was just swept by the worst team in the whole league - an under .300 team.

Posted by mildnhazy on 7/28/2023 7:19:00 AM (view original):
Here's one for you Jtpsops - my AL leading 110 mil team was just swept by the worst team in the whole league - an under .300 team.

I haven't complained about the quality of teams I'm losing to. Just the complete collapse of teams that were doing well for quite a while.