I have a ton of guys under achiveing (based off Ability/Last Seasons Stats) and 2 Over Achieving (based on Ability) and it has my Line Up very worried.  http://www.whatifsports.com/HBD/Pages/Manager/EditLineups.aspx  My 2-3 guys shouldn't really even be in the Majors but are hitting very well early on.  My #3 last year is Leading Off.  My Power Hitters in the 4-5-6 spots are now 7-8-9 because they can't hit enough BA to warrent going in the 3-4-5-6 spots.

  I am trying to get more Scoring Production, so right now they are where the CURRENT STATS suggest they should be playing.  When the numbers go more 'normal' or 'expected' I will shift again.  I am trying to ride the 'hot hitters' as it is, and not even look at Ability.  BUT looking at Ability shows my Line Up is all crazy.  How would you set these guys up?  If you can't see Player Links on that Page, I can post them.

We cannot see the lineup page for your team.

Regardless of player links, set your batting order in terms of ratings, not stats. Hot streaks are purely statistical and are not built into the game-- just because a guy has been tearing it up for a week does not mean in the slightest that he will continue to do so. It just means he keeps winning the statistical lottery every game but next game his best chance is still to perform exactly as his ratings dictate.

Posted by MikeT23 on 6/24/2010 10:47:00 AM (view original):
And, for the record, a lot of owners have a different opinion than prezuiwf.   While we agree that the ratings run the game, it's silly to stick with something for 162 games because "the ratings dictate it."

Why? Do you believe that the players on your team put up numbers based on something other than their ratings?

All players have 'up' years and 'down' years. If I have a playoff-capable team, after ST I set my lineup based on the ratings and my personal expectation/experience with that player and/or those ratings. After ~25 games, I shuffle the guys with better OBP toward the top and slide the guys with a 'poor start' down, but not too much unless I had doubts about the guy & his ratings in the beginning. At ~50 games, I discard my pre-season thinking and re-order the lineup based on current results and make a note of their stats "as of..." At ~75 games I review the notes and the recent results (up or down) and make minor adjustments. This is also when I decide "I need to make a trade as X is having a poor season". At ~100 games, I carefully monitor fatigue and begin to watch for the classic 'reversion to the mean' for players out-performing their ratings, making minor adjustments. At~130 games, it's time for call-ups and lots of fatigue management, so lots of lineup juggling. I typically want my best possible lineup for the final 10 division games if it's close to be at 100%. 

So while I base everything on ratings, it's also about recognizing when a guy is over or under his expected performance and either pulling the proverbial trigger or having lots of patience. 

There are no egos to contend with, so my long-time #4 hitter isn't going to have his slump prolonged if I drop him to 6th for two weeks. 

I do my lineups in two ways . . .

Up to the All-Star game, my lineups are set by ratings, or by last season's stats, or a combination of both. Whatever "feels" right to me. Either way, it's pretty much set for the first 91 games.

After the All-Star game, my lineups are set by stats; specifically by splits. I feel that by that time, I have a significant enough sample size to make some choices based on stats, and will tinker with my lineup accordingly. The deeper we get into the season, the closer the stats should normalize to where they belong, and guys will naturally fall into place. But if a guy has drastically overperformed in the first half, and continues to overperform in the second half, he'll stay in the heart of the lineup and I'll try to take advantage of what I know is just a random statistical anomoly. Likewise with an underperformer. He'll stay in the lineup, but may be dropped to the seven or eight hole until he earns his way out of there.

I'll typically change my lineups to break LOSING streaks.  If I notice a scoring drought, I'll check the splits and overall stats and re-adjust the lineup to optimize OBP at the top and OPS in the middle.  Or I might give a sub a shot at starting for a series if he has better ratings for either fielding or hitting.

I set my lineup on Opening Day based on ratings, and ratings alone (although if I've had my team for several seasons, I'll tend to work off past performance, too), but I will almost always tinker if someone is "cold" or "hot", because part of me knows that I don't know 100% how the ratings interact, so I have to go based on results... especially when the results last for several hundred PAs.

I will often re-evaluate how I rate players based on performance, since sometimes subtle differences in ratings can have significant impacts.

I base my lineup on ratings and then to a lesser degree their stats.

But if I have 2 players who are identical in every single non hitting rating and attribute except that one has higher hitting ratings across the board I will put the better hitter in a more favorable position in the lineup every time.  I don't care if the hitter with the poorer ratings is outhitting the other guy by alot, the better ratings (to the extent I can interpret them) will win out every time for me.

I've taken too many statistics courses and have learned too much about randomness to fall into the trap of attributing too much significance to small sample sizes.

"Cold" and "Hot" streaks are random and offer no predictive ability for future performance

schuyler101... I think we've had this conversation before!

Given enough time and data, sometimes hot/cold streaks are just an owner misreading the ratings. Perhaps I assume I'm wrong too quickly?