Looking for sympathy not advice!

I WANT my record to be reflected by the pythag record. There's nothing worse than being +75 in net runs but playing .500 ball in the win/loss column.

I was in an ATG league where I finished the season with a +55 run differential, 2nd highest in the league, but yet I finished with a 75-87 record, the least number of wins.

The reason for the +55 run differential: In six of my wins, I outscored my opponents 88 runs - 10 runs. Once that 78 run differential is taken into account, I then had a -3 differential for the other 156 games.

http://365.strat-o-matic.com/team/schedule/1109252

Sure, I get that.

A lot of games are played here every day. Heck, I just posted in another thread a manager who lost a finals game to me when his pitcher threw a no-hitter and lost 2-0! Another manager just posted a game 7 perfect game. Strange things will happen. Streaks will happen. And often the only explanation is nothing more than than the usual statistics which can be parsed within the laws of large numbers.

But for the discussion here, it would be like finding out my pitcher threw a no-hitter and lost in the finals game right after it happened to the other guy. A true "evener upper".

Weird things happen. But counter-balancing weird things shouldn't seem to be happening as often as they APPEAR to be happening this particular season.

It's certainly hard to understand how a 51-27 team then goes 32-52 or a 41-59 team then goes 43-19. And those are just the most extreme examples. The question, I think, is does Strat correct for extremes in player performance? If there was a baseline for a player's performance in a league and when the player exceeds a certain range tied to his expected performance he gets nudged back. If that were true then we would see extreme swings because Strat is correcting for too little or too much dice luck. I get that luck can swing back and forth but what is a bit unsettling is the way it at seems that a consistently good team has now become a consistently bad team. An example is looking at the record of my 83-79 team dividing it into 20 game segments:

12-8 0-20
15-5 21-40
11-9 41-60
15-5 61-80
7-13 81-100
8-12 101-120
7-13 121-140
8-14 140-162

Basically, my team was consistently good in the first half with no 20 game segment under .500 in the first half. In the second half it was consistently bad with no 20 game segment over .500. If it's just random dice roll luck then you would not expect the bad and good luck to evenly divide between each half. Of course that is only one team and by itself does not prove anything but very odd. My team scored 373 runs in the first half and gave up 297; in the second it scored 318 and allowed 416. And made only a few changes to the team, the only significant ones changing pitchers around game 120 (after things went bad) and those changes actually were beneficial.

I don't have an opinion on whether Strat has any built-in correction but odd things where terrible teams over a significant period of time suddenly become very good teams or vice-versa make one wonder if they happen often enough.

Actually, I did find a rational explanation--my team had 225 injury days in the second half as opposed to 133 in the first half. Forgive me Hal!

So strat is going to somehow make it where a hitter that starts off to a amazing start to the season will magically collapse in the second half of the season. A .340 hitter in 81 games will be a .260 the last 81 just to bring him inline with his life stats? That more or less would lead me to drop this .340 hitter around mid season before the rolls somehow bring him back down. I still argue that the predicted record is a bad idea. If I start a season and see that I will finish the season 15 games out and know that strat will do everything possible to have the played records match up with their predicted ones that would make it very hard to keep interest or not make wholesale changes because what would the point be to have it play out as is?

Seb,

The pythag record, or predicted record, is a Bill James formula where net runs predict how many wins you will have in a season. It is completely independent of SOM Online.

If you take a large sample of teams, you'll find the pythag records form a perfect bell curve, mostly between +10 and -10 games from the actual record, with most teams within a game or two. SOM does NOTHING to move actual records to match the pythag record.

Further, SOM does nothing to "govern" or "average out" season outliers among individual players. This discussion has been addressed many, many times both online and off, with beta testers and other insiders.

Having said all that, as you can see from the posts above, it sure SEEMS like at least a few independent observations are being made where strange and statistically unlikely events appear to be occurring, esp when it comes to counter-balancing hot/cold streaks. However, this is all purely anecdotal and in good fun, i.e. "shared misery."

I've done pretty well with the 2015 set. That alone should signal that there is something wrong with it.