This Has My Attention: New Questions

A couple of months ago, the November Challenge caught my eye for some reason, so I set up the spreadsheets and mapped out two teams (one from 200x and one from ATG 9) in an attempt to gather some extra credits.

My 200x team, which had a dominant TTO% and had been in first place and in the black all season, finished 3-13 to end up five games below .500 and costing me three credits. Wow, some bad luck in the last week of the season. Odd, I had a positive pythag, but HAL clipped five games from me.

https://365.strat-o-matic.com/league/462784

Fortunately, my ATG team won the day and tonight I can still collect my three credits. Hurray for me. Odd again, I only had 89 wins despite a +213 pythag. HAL clipped eight games from me.

https://365.strat-o-matic.com/league/462692

Meanwhile, after a deep dive discussion with a friend about plate discipline, I concocted a team designed to exploit “plate discipline” by sorting the spreadsheets for hits and BB/K ratios. It’s been a very long time since I came up with an original experiment like this one, so I was kinda excited about watching it play out.

https://365.strat-o-matic.com/league/462980

With nine games to go, I’m going to most likely miss the playoffs, because despite a +154 pythag, HAL (as of game 153) has clipped 10 games from me.

So, this got me thinking a little more about pythag, and I discovered that according to the stat nerds, pythag is accurate to within +/- 3 games.

Curious, I checked out my very first 15 teams from 2002. 11 out of 15 teams were +/- 3 games, and of the other four, two were plus five and two were minus five, making for an extremely eye pleasing bell curve precisely as you would expect to find.

Let’s try the middle of the pack. I looked at my first 15 teams of 2012, and found 13 of 15 teams were +/- 3 games, and the other two were a plus five and a minus four. Those were woolly days, because…

Currently, I have three teams closing out their seasons holding -10, -6 and -5 pythags. Of my last ten completed teams, I had six teams finish +/- 3, BUT four teams finished -5, -8, -13 and -6. Zero teams finished better than +3 to balance those out.

Like so many others, I’ve had a super nagging feeling that something is not quite right with the current gameplay.

My record is a full twenty years long now. I’ve had a .530 winning pct over that span, and managed to win one ring for every five teams I’ve played.

So here’s some food for thought. A .530 win pct equates to an 86-76 season. In a strange way, playing a game 40 times a year for 20 years in order to be six games above a coin flip is pretty ridiculous. Funnier still is that .550 managers are only three games a season better. Truly, a game of inches. So if HAL has a thumb on the scale, which to me is becoming beyond obvious, he only has to “short” me six games a season to make me an elaborate coin flipper. Worse still, the league then picks up those six wins so I get slapped on each side of the face coming and going.

I’ve experienced the ups and downs of individual card sets, so I know how that goes. I know how to lose and I know how to win. I can’t fully explain why I feel the way I feel, but I’m finding these current results are making me pretty uneasy, and not in a “you can expect some streaks of bad luck” kind of way.

Playing out the string right now because of my general lack of "trust" for the game and the people behind it so I'm down to 4 teams.

Those 4 teams are:

-7 games
-11 games
-5 games
-6 games

Below their respective Pythagorean records. Two of they teams "should" be leading their divisions by 10+ games but are instead tied or up by only 1 game.

I can confidently say that easily 80%-90% of my teams UNDERPERFORM their Pythagorean projection. And, I can honestly say that I don't remember the last time one of my teams OVERPERFORMED their Pythagorean projection.

Something's rotten in Denmark (actually Glen Head).

Thinking this all through, I think the thing that shocks me the most is how easy it would be to sneak a game or two (or ten) away here and there. And if it was being done somewhat randomly (governor sometimes on, governor sometimes off) there is no way on earth you could ever detect it or isolate it.

But to me, the key piece of the puzzle is right here:

I can confidently say that easily 80%-90% of my teams UNDERPERFORM their Pythagorean projection. And, I can honestly say that I don't remember the last time one of my teams OVERPERFORMED their Pythagorean projection.

Something's rotten in Denmark (actually Glen Head).

Interesting. But, of course, "under-performance" for one team in a league means "over-performance'" for another, and any analysis/conclusions would have to account for that as well, unless you think that there are particular managers simply being "singled out". At any rate, just out of curiosity I went back and looked at my last fifteen teams, and came up with this in terms Pythagorean projection:

-8 games
+2 games
-4 games
-10 games
-3 games
0 games
-1 games
-3 games
+3 games
0 games
-5 games
-2 games
-4 games
-8 games
-1 games

Or a net of -44 over 15 seasons. And there's the fact that of the last 19 playoffs I've been in, I've won exactly 0 rings, which seems to defy the law of averages. On the other hand, I'm very reluctant to indulge in a persecution complex ("why me"?), and don't like conspiracy theories.... So one thing I'm wondering is whether teams with winning records *tend* by nature to under-perform their Pythagorean projection (because blow-out victories tend to exaggerate the projections), and losing teams tend to over-perform (for the same reason: blow-out losses tend to exaggerate the projection on the down side).

My last 25 completed teams (includes 8 champs), has a +11 , so I'm definitely not seeing it. Not a fan of conspiracy theories either. I've seen some huge differences over complete MLB seasons for teams i follow.
+6
+4
−1
−2
−1
−1
+8
+1
−4
+2
−3
−6
−5
+1
+5
+1
−3
+2
−6
+3
−5
+6
+4
+0
+5

I wouldn’t describe what I’m saying as a persecution complex. Again, after 20 years of this, I’m pretty familiar with the vagaries of the game. Also, you shouldn’t over or underperform the pythag projection. Your net runs simply reflect what the record should be. My losing teams underperformed the pythag as often as my winning teams. That’s the thing. It should be a very straightforward bell curve with the VAST majority of records within +/- 3 games (per the nerds). Then, for every outlier, there should be a (somewhat) matching outlier on the other tail.

However, I have ZERO offsetting outliers, only bad side outliers.

As for singling out particular managers, no, I don’t agree with that. What I would do is single out manager records. For every percent above .500, you could have the game engine dock you in some tiny fashion (say your OPS performs at only 98.5% or something. Just like the home field advantage gimmick. The results would be imperceptible to the naked eye. Or, vice versa, for every percent you are below .500, the game “helps you out a little.” Again, how you would recognize this against “unlucky dice rolls” would be impossible to see.

It’s strange that so many long time managers all started getting these weird feelings all within the past year or so.

BC15NY: You must be immune from this. My brain would break if I ever had a +8!

Without addressing the conspiracy stuff, just want to note that teams with good bullpens will overperform pythag and teams with bad bullpens will underperform pythag. This is true in Strat and real life because of the importance of bullpens in winning close games. I've never analyzed my performance vs pythag but I assume I underperform it because below $200 my bullpens are ass.

It did occur to me that a lot of the uneasiness does seem to coincide with the change in bullpen versions.

Max, in our last league together you underperformed your pythag by -11!

https://365.strat-o-matic.com/league/461821

And yes, your bullpen was completely ass!

My bullpen was no better, but I snuck away with a +3.

In the sabermetric stuff I’ve read, no where have I seen where relief pitching factors in to the pythag record. Having said that, it does make logical sense that $3 bullpens will lead to the more than occasional blowout, skewing the net runs numbers.

It would be nice to try the old bullpen version again and see if all the ghosts and goblins would simply disappear with the change back.

Pythag is useful because it's more predictive of future wins than actual wins are. The biggest reason for that is that there is lots of luck/random variance involved in close wins and losses. Extreme bullpens reduce the amount of luck involved in close games: teams with great bullpens are able to win close games more consistently than teams with average bullpens, and vice versa for teams with terrible bullpens. So I think it's less about bad bullpens leading to blowouts (since in ATG most of those involve the SP getting shelled as well), and more about close games being more reflective of true talent for teams with extreme bullpens than they are for teams with average bullpens.

Here's a study that seems to support: https://www.baseballprospectus.com/news ... of-relief/