Given that he's a DH, the comment must be referring to out-of-ballpark activities...

I've never played HD before. Do seasons in HD not start unless all of the teams have human coaches?

Posted by dedelman on 3/12/2019 3:46:00 PM (view original):
Given that he's a DH, the comment must be referring to out-of-ballpark activities...

This is the best comment I have heard in years.

Posted by WillEphs on 3/12/2019 4:58:00 PM (view original):
I've never played HD before. Do seasons in HD not start unless all of the teams have human coaches?

Correct, it starts when the league is full. Some leagues fill much quicker than others. Your best bet to play quickly is to join a public league with few openings. Private leagues that fill quickly don't typically take players who haven't played HBD before. There are exceptions to this, but I don't know which leagues those are.

Sorry for the long post, but this topic really interests me. I apologize for my nerdiness.

It is difficult to compare real life to the sim because in real life, we typically need to see a players stats before we evaluate how good a player is. Whereas in the game, most of us evaluate players by ratings.

Nonetheless, lets compare coin flips to at bats.

In real life, flip a coin 600 times (to simulate 600 at bats). Do this for each of the 800 (32 teams *25 players) players in majors in HBD. Chances are, for a nice sum of players, a nice sum of the coins will land on heads much more often than tails and vice versa.

However, if you flip a coin 600 times for each of the 800 players over a ten season period [STILL IN REAL LIFE], there will STILL BE SOME, but much less that are still having the same good/bad luck over that ten season period.

Yet, in the game (HBD). I HAVE NEVER seen a player sustain this kind of good/bad luck over a long period of time. Have you? Have you seen a AAAA guy that has been in the majors with sustained (i.e. 8+ years) success? Or have you seen a player with great ratings that consistently does not perform to ratings for 8+ seasons? I have not. In real life coin flips it is rare, but it does happen. However, in HBD coin flips, I have NEVER SEEN IT.

There are two things that can explain this.

1. When evaluating players, most owners either can't look at stats because the player is a prospect and stats don't exist. Or, the owner looks primarily at ratings when evaluating players. If most owners look at ratings, maybe there have been AAAA players that have hit 800 OPS over 8+ seasons, but since we only look and sort by ratings, even if there was a player that was a AAAA player (55-60 hitting across the board ratings) that has an MLB career OPS of 800, we would never know because he would probably retire at the age of 30 because nobody signs him. However, I rule this out, because of the other side of the coin. I still have not seen or heard of a player that has 75+ across the board hitting ratings but has an OPS of 650 over 8+ seasons. If there was no code to protect against this, it would happen. It would be rare, but it would happen. So I think there is code to protect against this. Also, it would make sense that they protect against this. Because people will be people. An owner with a player that consistently under performs his ratings will put in a ticket. They don't want many tickets on this.

2. There is some type of progression/regression equation used to bump/nerf player performance. At first, I thought it might be a seasonal ten percent bump/nerf, but I really have no idea what the timing of it is. I just think that it is there. if it wasn't, from time to time we would see guys over/under performing over long periods to time.

My Kansas City Sprints in Rice Not Lynn started the season 1-14 and are 30-14 since. Shall I just assume Bench Coach Sammy Lawton has been betting on Sprints' games?

Posted by tlowster on 3/12/2019 8:51:00 PM (view original):
Sorry for the long post, but this topic really interests me. I apologize for my nerdiness.

It is difficult to compare real life to the sim because in real life, we typically need to see a players stats before we evaluate how good a player is. Whereas in the game, most of us evaluate players by ratings.

Nonetheless, lets compare coin flips to at bats.

In real life, flip a coin 600 times (to simulate 600 at bats). Do this for each of the 800 (32 teams *25 players) players in majors in HBD. Chances are, for a nice sum of players, a nice sum of the coins will land on heads much more often than tails and vice versa.

However, if you flip a coin 600 times for each of the 800 players over a ten season period [STILL IN REAL LIFE], there will STILL BE SOME, but much less that are still having the same good/bad luck over that ten season period.

Yet, in the game (HBD). I HAVE NEVER seen a player sustain this kind of good/bad luck over a long period of time. Have you? Have you seen a AAAA guy that has been in the majors with sustained (i.e. 8+ years) success? Or have you seen a player with great ratings that consistently does not perform to ratings for 8+ seasons? I have not. In real life coin flips it is rare, but it does happen. However, in HBD coin flips, I have NEVER SEEN IT.

There are two things that can explain this.

1. When evaluating players, most owners either can't look at stats because the player is a prospect and stats don't exist. Or, the owner looks primarily at ratings when evaluating players. If most owners look at ratings, maybe there have been AAAA players that have hit 800 OPS over 8+ seasons, but since we only look and sort by ratings, even if there was a player that was a AAAA player (55-60 hitting across the board ratings) that has an MLB career OPS of 800, we would never know because he would probably retire at the age of 30 because nobody signs him. However, I rule this out, because of the other side of the coin. I still have not seen or heard of a player that has 75+ across the board hitting ratings but has an OPS of 650 over 8+ seasons. If there was no code to protect against this, it would happen. It would be rare, but it would happen. So I think there is code to protect against this. Also, it would make sense that they protect against this. Because people will be people. An owner with a player that consistently under performs his ratings will put in a ticket. They don't want many tickets on this.

2. There is some type of progression/regression equation used to bump/nerf player performance. At first, I thought it might be a seasonal ten percent bump/nerf, but I really have no idea what the timing of it is. I just think that it is there. if it wasn't, from time to time we would see guys over/under performing over long periods to time.

You can't compare at bats to coin flips because the algorithm is far more complex then that. Looking at just a batter and taking everything out of the equation lets do some simple math (which is probably wrong but I'm just doing this for comparison.


Johnny Wang is on the San Diego Whatchamacallits and is up to bat.
Johnny's stats are
Contact: 80
Power: 30
VsL: 60
VsR: 76
Eye: 86

Johnny steps up to the plate. The sim first looks to see if Johnny makes contact with the ball or strikes out. So lets take contact + eye divided by 2. Giving us a total of 83. Pick a number at random between 1 and....... lets say 110 (even guys with 100 contact and 100 eye probably still strike out). If the number was between 1 and 83 Johnny made contact and did not strike out.

Next lets see if the contact hit is a hit or an out. We look at VsR because Johnny is facing a right handed pitcher. Pick a number between 1 and 110 again at random. If the number is between 1 and 76 Johnny's contact will be at least a base hit.
Now lets see how many bases Johnny got off of his hit.

Pick a number at random again between 1 and 110. If the number is between 1 and 30 Johnny got a double. If the number is between 1 and 15 Johnny got a triple and if the number is between 1 and 7 then Johnny hit a homerun.

Now, again, this is a very very very watered down version of the HBD algorithm but this is why there are no outliers. Its not a flip of a coin or a 'true/false' statement. There are constant numbers and calculations going into each at bat.

AND even given Johnny Wang's ratings, I would say it's completely not out of the question that he could strike out 20 times in a row, or hit .200 for three quarters of a season.

"There are constant numbers and calculations going into each at bat." I agree. That also means "the other guys get paid too." The other owner is doing his best to put the best players he can get up against Johnny Wang. When we say small sample size, it seems to me that very few owners in this game take into account the opposition. The other teams are always changing. Their players are getting better or getting worse. The pitching rotations are changing. The teams are not the same as they were last season. It's entirely possible that Johnny Wang goes a month or two facing only the league's other good teams, AND the other team's SP1s and SP2s, AND never gets to beat up on any AAAA pitchers.

And THEN, only THEN, do we bring in the randomly-generated "coin flips."

Perfectly reasonable explanation.

Can somebody explain how a team that has an expected winning percentage of .576, ends up with an actual winning percentage of .469? My guess is that there are a lot of factors at play as well. My Wichita team in Earl Weaver was 2nd in the MLB in ERA, WHIP, Slugging given up, and OBP given up. First in the AL in runs given up. Also, was above average in every major hitting category. They were not world beaters at the plate, but they OPS'd .761 as a team at the plate this season.

I suppose it could be just dumb luck (1-run game record was 15-30). It could be that the AL that I am in is the most competitive league I have ever been in. Five of the six playoff teams are the same every year, but even the guys that don't make it are still full of talent up and down their line up and pitching staff. It could be that I was fifth in the league at GIDP (on offense) and first in the league in caught stealing (on offense) and maybe the games that I lost by one or two runs were the games where I had guys get caught stealing in high leverage spots or I had a guy up with the bases loaded and one out and he grounds into a double play.

I am probably just coming across as whiny here, but having a 0.060 difference between your expected winning percentage and your actual winning percentage is bad luck. Having a .107 difference is downright ridiculous. It would be easier to swallow if I had an explanation. Has anybody else ever seen a .107+ bad luck season where your expected winning pct is that much higher than your actual?

Final Line:

Team	Owner	W-L	PCT	GB	Division	1-Run Games	Extra Inning	Spring Training	RS	RA	Exp PCT
Honolulu Lunas	dewersansoda	99-63	.611	-	20-10	22-16	8-6	7-11	881	702	.607
Wichita Seven Nation Army	tlowster *	76-86	.469	23.0	14-16	15-30	8-9	8-10	769	652	.576

The Expected win PCT is actually the most watered down formula WIS does (which is odd to me). All it is is some combination of your runs scored divided by your runs against or something. I've never actually figured out the formula but if you look at other teams across the league, if you have more runs for then against, your EXP PCT will be over 500. If they are equal, it will be at 500, if you have more Runs against then for it will be below 500. Sadly, you can't really use this information for much.

Pretty much what I think. It's a statistic, but how useful a statistic is debatable. "My team might be playing better than it looks. Or worse than it looks." So how you act on that, or don't, is up to you.

It's the pythagorean expectation: (runs scored)^2 / (runs scored^2 + runs allowed^2). It's a historically unbiased estimation of expected winning percentage in baseball. Think of it as a line of best fit with usual error. After a fair amount of games it can be used as a gauge of team performance, removing luck from the equation. That being said, certain strategies can make you deviate from the prediction. For instance, if you have a terrible mopup guy relative to the league, your predicted % may be downwards biased.

^it also appears the exponent of 2 has been refined to better fit what has actually historically happened. Based on my team's record, HBD doesn't use two. There are a few different ways to calculate the exponent (some taking into account games played, so it changes) not sure what HBD uses.

How much is slot money for each slot?

Well, it blew my mind this past season in Earl Weaver. Not sure how I can be 2nd in all pitching categories and 11th in Offensive OPS, but still be five games under 500. But, I guess if my batters were beating up on crappy mopup guys or below replacement level pitchers all season, it makes more sense. Yet, it doesn't make me fell any better.