How Bad Can You Be?

I'm on the Celebration of the Winter Solstice Holiday Break, and as usual I'm over-thinking my Strat-O teams.

So I have a theoretical question that may or may not have a definitive answer.

How bad can a manager be, averaging over all teams played, if he played every team like he intended to win (but he had the worst luck and didn't win)?*

* By that, I mean, you play a full salary team, without excessive in-season drops, without playing a bunch of 5s in the field, without playing 50 cent starting pitchers, etc etc. Assuming the top managers win 55% of the time, does this mean the "average" manager wins only 45% of the time, or does he always hold a 50/50 proposition?

I'm following the New Observation Thread and wondering about in-season drops, which in my opinion have always been the most visible statistic paving the road to underperformance. If it's not, why do you think otherwise good teams using the full allotment of salary lose?

Let me ask the question this way: In a universe of managers using only the HAL picks 'em function, you would see a bell shaped curve of results very tight within the 45% wins to 55% wins range. What would the teams on the 45% wins side look like and why?

These are three very different, not entirely-related questions that aren't easily answered:

By that, I mean, you play a full salary team, without excessive in-season drops, without playing a bunch of 5s in the field, without playing 50 cent starting pitchers, etc etc. Assuming the top managers win 55% of the time, does this mean the "average" manager wins only 45% of the time, or does he always hold a 50/50 proposition?

The big question here isn't the original question of "how would an average manager play?" The big question is actually: What constitutes an "average" manager and how do you determine what an average manager is? I would love to see somebody actually determine that.

I'm following the New Observation Thread and wondering about in-season drops, which in my opinion have always been the most visible statistic paving the road to underperformance. If it's not, why do you think otherwise good teams using the full allotment of salary lose?

This one is kind of Holy Graily. There are a multitude of factors shaping and influencing a good team's failure. If anybody can actually determine what does, they should sell more guides than JoetheJet...

Let me ask the question this way: In a universe of managers using only the HAL picks 'em function, you would see a bell shaped curve of results very tight within the 45% wins to 55% wins range. What would the teams on the 45% wins side look like and why?

Unless somebody has actually played a few leagues with HAL-picked teams, and few if any have, he can't really answer this. There are ways to possibly speculate how and why HAL would pick teams, but none that could answer the question alone. Considering HAL's greatest concern in a league where HAL picked every team would be equal distribution, I wouldn't assume his selection process would be very rigorous. So, predicting that process would be pretty speculative.

No, there are no easy answers, I agree.

That's why I'm asking for the speculation.

If you have played x number of teams and you are sitting at 45% wins, should you use the HAL picks 'em function only, and if you did, what do you THINK would be different than what you were doing previously?

Is it possible to be worse than .500 without TRYING to be?

A player should only make such a decision with a substantial knowledge of the efficacy of HAL's picking process. Considering how many people still whine about HAL ruining their teams while never giving credit for HAL's assistance in their success, people generally have a low esteem for HAL's skills. So, for most players, I assume they would not hand over the reins to HAL, regardless of their record. And that''s not just because they want to manage themselves.

As I said earlier, HAL's selection process would probably emphasize equal player distribution over team quality. That would be amplified in a league where he picks all the teams. That would also be exacerbated by HAL's unlikelihood to understand team "chemistry" in his formation. So, no, I don't think HAL would do significantly better than a manager with a 45% record.

Someone with considerable knowledge of probability, logarithms, and/or computer science could certainly chime in with evidence to the contrary.

Is it possible to be worse than .500 without TRYING to be?

sure.....its all about the draft.

A GM who only favors under performing over valued players will seldom win over .500

I could put together a team of assorted underachieving players with decent bats and above aver D that I feel would not crack .500

Example

OF: Upton/C Gomez/J Bruce

INF: Santana/Hosmer/Philips/Hardy/Seager

I dont think this team would win over .500( assuming 30M +/- on pitching)

There are hundreds of combinations like this one.

No, there are no easy answers, I agree.

That's why I'm asking for the speculation.

If you have played x number of teams and you are sitting at 45% wins, should you use the HAL picks 'em function only, and if you did, what do you THINK would be different than what you were doing previously?

Is it possible to be worse than .500 without TRYING to be?

I have had a number of teams that provide anecdotal evidence that the answer is "Yes."

I am always attempting to win games, and even when I do some crazy trades (i.e., I expect my team to be worse, but I am sick of looking at an under-performing card and want to try out someone I normally would never take/want), I am still hoping to win more than I lose. I don't always achieve that goal, however.

I would expect, a HAL-pick 'em roster, no drop/adds, no strategy or even lineups, should be able to win 40 games. Of course, it is untested speculation, but geez, 25% win percentage in SOMO universe should be do-able.

Is it possible to be worse than .500 without TRYING to be?

Edit: Is it possible to be consistently worse than .500 without TRYING to be?

For blue turtle,

We ALL have underperforming teams. We do not all have consistently underperforming teams. Why not?

For splinter,

That is a very intriguing observation. You argue you can pick the underachievers in advance. So would you say that there is a discrepancy between the quoted price and the actual value of the card, and the successful manager is better at exploiting the difference? Couldn't an underperforming manager learn these discrepancies and improve upon his record over time then? At least enough to get to .500?

I like the simplistic approach---An average manager would have a .500 record over several teams.


As for dropping and winning--this is the only one of mine that made a comeback dropping players--a guy in the division dropped Ethier and I just made wholesale changes at that point to get him--team was 82-80 in a weak division though, but did make up a lot of ground late.

http://onlinegames.strat-o-matic.com/team/1107709

I like the simplistic approach---An average manager would have a .500 record over several teams.

That is a simplistic approach, but statistics used for determining the average member of a group rarely bear out such simple results. Assuming the average player has a .500 record assumes there aren't more players with sub.500 records than post-.500 records, and vice versa. Either case is extremely unlikely. To effectively determine the "average" player, in terms of win-loss, one needs to first find out if most players have winning records or losing records. No one has determined that yet.

Is it possible to be worse than .500 without TRYING to be?

Edit: Is it possible to be consistently worse than .500 without TRYING to be?

For blue turtle,

We ALL have underperforming teams. We do not all have consistently underperforming teams. Why not?

For splinter,

That is a very intriguing observation. You argue you can pick the underachievers in advance. So would you say that there is a discrepancy between the quoted price and the actual value of the card, and the successful manager is better at exploiting the difference? Couldn't an underperforming manager learn these discrepancies and improve upon his record over time then? At least enough to get to .500?

Some of learn a Home Run on 3-12 is worth 6x less than a Home Run on 3-7 faster than others?