Thursday, August 5, 2010

Deconstructing Head-to-Head Part 1

So of course, for the first post after my pledge to restart the blog, I'm going to do something that wasn't mentioned as one of the ideas bouncing around in my head. Go figure.

Nope, I've decided to go with something that I've pondered about for a long time, but never really figured out how to quantify. As you might have figured from my very long posts 2 years ago examining my fantasy baseball draft halfway through the season, fantasy baseball is kind of a passion of mine. I'm no fantasy nerd, I don't obsess over every fantasy sport I play. Outside of baseball, it's merely just a casual thing with friends for bragging rights (except come March Madness time, when it's sometimes for money). But with baseball, it's kind of a point of pride. Yeah yeah, I know, that sounds awfully pathetic - that I invest a lot of time and even pride in my fantasy baseball teams. See, I view baseball as MY sport, and if I don't absolutely dominate in the fantasy version, then my being the resident baseball expert is in doubt. Yes, I also recognize that that's more of a self-imposed thing than anything else, but still. The fact remains that baseball is my favorite sport, and therefore I "care" more about fantasy baseball.

The leagues I'm in are what are referred to as Head-to-Head leagues, as opposed to the traditional Rotisserie (or Roto) leagues. Because each week is separate from each other, and much of your record is based on the luck of the draw as far as your schedule goes, the statistically best teams don't always end up at the top of the standings, or the statistically worst teams at the bottom. At least, that's what the popular belief is - but does it hold any weight? And if it does, then what is a possible way to better ensure victory from week to week? I plan, using the principal league that I'm in, to come up with some answers. I'll do it in three parts, starting with #1 today:

1) Do the teams as currently seeded match up with how they would be seeded in a Roto league?
2) Are teams that produce most consistently (no matter the level of production) as a whole more likely to be higher in the standings, and vice versa?
3) Are teams with players that produce most consistently (no matter the level of production) independently of each other more likely to be higher in the standings, and vice versa?

This league I'm in, luckily, archives statistics from the entire season, and I used to make practical use of them every so often. That is, the first year I was in the league, I would completely list out how everyone stacked up to each other statistically every couple of weeks - in order for me to see where I needed to improve my own team. Then it became far too time-intensive, and I never did it again...until now. So without further ado, here goes -

#1 - Do the teams as currently seeded match up with how they would be seeded in a Roto league?

As of Week 17 (last week), the standings are as follows (team names are just their seeding):

Team 1
Me - 2.5 GB (games back)
Team 3 - 8 GB
Team 4 - 8 GB
Team 5 - 21.5 GB
Team 6 - 22.5 GB
Team 7 - 22.5 GB
Team 8 - 27 GB
Team 9 - 27.5 GB
Team 10 - 28.5 GB
Team 11 - 32.5 GB
Team 12 - 34 GB
Team 13 - 34.5 GB
Team 14 - 67 GB

Three things to point out here - the clear delineation between the top 4 teams and the rest of the league, the muddled mess that is Teams 5-13 (an average of 1.44 games separating each team) and the obvious inferiority of Team 14. Remember these points later.

So what I did after this was to write down the place each team is in in each statistical category, and then added up the totals for the offensive categories (runs, hits, homers, RBIs, stolen bases, batting average) and pitching categories (innings pitched, wins, saves, strikeouts, ERA, WHIP). After finding the average of each (for my own curiosity), I added up the overall total and found the total average place each team was in across the board.

Take my team for example. I was 6th in runs, 4th in hits, 7th in homers, 6th in RBIs, 2nd in stolen bases, and 10th in batting average - which came out to an average of a place of 5.83 in each offensive category. I had an average place of 3.167 in the pitching categories, after being 2nd in innings pitched, 1st in wins, 6th in saves, 1st in strikeouts, 6th in ERA, and 3rd in WHIP. Overall, this came out to an average place of 4.5. This meant, then, that I was effectively in 4th and a half place in every category. Now all I had to do was do the same for every other team, and see where the stacked up relative to each others' average places. If we were in a Roto league, then, this is how the league would look:

Team 4 - 4.42
Me - 4.5
Team 3 - 4.5
Team 1 - 5
Team 5 - 7
Team 9 - 7.42
Team 11 - 7.75
Team 12 - 7.75
Team 10 - 8.58
Team 13 - 8.67
Team 6 - 8.83
Team 7 - 9.08
Team 8 - 9.75
Team 14 - 12.08

Note that the patterns that I pointed out above hold true here as well. There's a clear gap between the top 4 teams and the rest of the league - in fact, the jump of 2 average places from Team 1 (in 4th place here) to Team 5 (in 5th place here) is as much as the jump from Team 5 to Team 7 (now sitting in 12th place). Additionally, Team 14 is still far worse than everyone else.

And this is where we see something kind of interesting. Only myself, Team 3 and Team 5 would be in the same position in a Roto-based league as we are in our Head-to-Head league. In fact, Teams 9-13 all moved significantly up, while Teams 6-8 moved significantly down. In my biased view, this is good news - because those teams in line to make the playoffs now as the lower seeds are even worse statistically than their records might suggest.

So I guess what we can take away from here is that unless your team is far and away better statistically, or far and away worse statistically, where you end up exactly in the standings is not necessarily directly related to your statistical production. This is in line with the popular belief I explained above - therefore, what might directly influence the standings (other than who you play when, which is far too complicated for me to attempt to quantify)? I'll try to start answering that next time when I tackle question #2.

No comments: