Confusion in Carolina

No, this post is not about the woeful Carolina Panthers who are now 0-3 after a sloppy loss to the Bengals today. The Cats do looked confused on the field though. Things are not looking any better with upcoming games against the Saints and Bears.

Living in North Carolina and enjoying many different types of beers, I like trying the various local brews. There is one problem with beer from this area however, particularly if you are trying to keep track of what you are drinking. There are multiple brews available with "Carolina" in the name. In fact, there are two separate operations with "Carolina Brew..." in their name. The Carolina Brewing Company is located in the town of Holly Springs which is a short drive south of Raleigh. Their Spring Bock and Summer Wheat are on my Beer List, and I have previously had some of their other brews.

There is also Carolina Brewery which operates two brewpubs in this area. The original is in Chapel Hill. I have never been to that location, but I have been to their newer location in Pittsboro. Beers from Carolina Brewery are available in other places besides the two brewpubs.

In addition to the "Brewing Company" and "Brewery", there is the "Beer Company". Carolina Beer Company is located in Mooresville which is just north of Charlotte. They offer "Carolina"-branded beers plus specialty brews under the Cottonwood brand. Their Cottonwood Pumpkin Ale was the first pumpkin-flavored beer that I tried, and I still think it is the best pumpkin beer I have tasted. Try it in a half-and-half with Guinness or Murphy's Stout.

Finally, there is the string of Carolina Ale House locations. To be completely confusing, they offer their own special brews plus they have beers from the Carolina Brewing Company on tap. The Ale House's Hat Trick Red is on my Beer List.

Ordering a draft of the "Carolina" special around here is an iffy proposition. I was recently at a bar, and I had an oatmeal porter and an Oktoberfest from one of the "Carolina"-named companies. It took some investigation, but I was able to figure out that these two beers were from Carolina Brewery. First, I had the Oatmeal Porter. The Internet tells me that the addition of oatmeal to a porter or stout gives a smooth creaminess. That is definitely the case for this brew. It is a delicious beer, but one is really all I could drink. I followed the Oatmeal Porter with the Oktoberfest. Not quite as tasty as the smooth porter, but also a good beer.

While I am on the subject of local beers, I would like to mention a new brewery that recently opened in north Raleigh. According to their website, Roth Brewing Company was founded in 2008, but their beers have only started appearing in the past few months. I have only had two of their beers, and to be completely honest, I did not care for either one. Their FoeHammer is a big barleywine with a hefty 10.5% ABV. I thought this beer had a very strong alcohol flavor. It seems to be a high-gravity beer for the sake of the high-alcohol content. I have also tried their Forgotten Hollow which is suppose to be a cinnamon porter. It definitely was flavorful, but to me it did not taste like cinnamon. Even though I did not care for these two brews, I would still like to welcome Roth to the neighborhood. I am looking forward to trying more of their offerings, and I am sure I will find several that I enjoy.

That's a Great Idea

Today's post was originally going to be about beers from a couple of local breweries. I will get around to that post eventually. However, I was reading the most recent article from ESPN.com's Bill Simmons who has a familiar idea. In his article, he complains about the recent futility of the NFL's NFC West division. He blames the problem on the small divisions in the NFL and proposes to correct the issue via realignment. While he attempts to maintain the two conferences, his proposed alignment is very similar to one of my proposals. My first proposal eliminated the AFC and NFC and simply combined the directional divisions. My followup realignment proposal had the Baltimore Ravens, Indianapolis Colts and Miami Dolphins rotating divisions. The only difference between that proposal and his suggestion is that he has the Dallas Cowboys in the geographically-correct "NFC South" (my "South Division") which moves the Carolina Panthers to the "AFC East" (my "East Division"). Despite any excitement over this weekend's Lone Star showdown with the Houston Texans, I seriously doubt Cowboy's owner Jerry Jones would ever agree to give up the Cowboy's traditional rivalries with the other NFC East teams.

While our divisional alignments are similar, our schedule proposals are vastly different. He proposes a complicated 18-game schedule which includes only playing home and away games against some divisional opponents while only playing one game versus others. He also fits in eight games against teams from the other divisions. My proposal comes in 16- and 17-game flavors. Teams would play home and away games against each of the other seven teams in their divisions for a total of 14 games plus games against one team from either two or three of the other divisions. I just do not feel that the NFL will go to an 18-game schedule. I think the idea of expanding the regular season to 18 games is a negotiating ploy by the NFL team owners with the players union. If the idea does not get dropped completed in favor of some concessions by the players, I think the two sides will compromise by expanding to only 17 games.

Super Secret Formula

As I have previously discussed, I, like many people, play fantasy football. For the past few years, I have been in two leagues. As I think is common in fantasy leagues, both of my leagues release a weekly power ranking. Ever since I joined a 10-team league with some guys at work, the power rankings have been sent out each week as a large Excel spreadsheet. The spreadsheet tracks various pieces of data. First is the points scored by every team with the high and low scores from each week highlighted. Next, there is the number of other teams each team would have beaten for every week. Finally, there are the wins and losses for the season. The data in each category is cumulative as the season goes along. The ten teams are ranked in each of the three categories. The best team in each category gets 10 points and the worst gets 1 point. The teams are ranked based on the sum of the three rankings. Therefore, the best possible score would be 30 points and the worst would be three.

My other league has twelve teams. One of the other owners is Alan, a college buddy mine. His brother Matt is the league manager. The other owners are various friends of theirs. When the league first got started, Matt would post weekly power rankings based on his own personal opinions. While that was interesting and generated fun posts on the league's message board, I felt that I could come up with rankings that more accurately reflected the strength of the teams. I knew I would be using a spreadsheet, but I did not want to maintain data from previous weeks as is done on my other league. Therefore, I wanted to only use data available for the current week.

Both of my leagues are on ESPN.com. The standings page on ESPN.com includes two tables. The top table has wins (W), losses (L), ties (T), winning percentage (PCT), and games behind the first place team in the standings (GB). The second table has points scored for (PF), points scored against (PA), home record, away record, division record, and streak of either winning, losing or tying. (I have seen a tying streak of two). In my fantasy leagues, the home and away records are meaningless. My 12-team league only has one large division, so in that league, the division record is also meaningless.

I have been constantly evolving my ranking formula since I first began. I started by simply copying and pasting the standings data into a spreadsheet and then playing around with it. I eventually settled on using four of the available data categories. To the right is example standings from last season showing the four components. I think PCT, PF and streak are pretty obvious. Teams with the best PCT and PF have good arguments for being the best team. The streak component answers the "What have you done lately?" question. Regardless of the other data, would you rather play a team on a long winning streak or a long losing streak?

Probably the only questionable component is PA which I include in the formula as an adder instead of a negative. In fantasy football, the players on your team have no affect on your opponent's performance. Therefore, it is my opinion that PA represents the strength of a team's schedule. I think including PA as one of the components helps teams that have scored a good number of points, but do not have great winning percentages or long winning streaks because they have played teams that have also scored a lot of points.

If you thought things were geeky already, it is now going to get even geekier. For each component, I normalize the values. Normalization is a mathematical technique were a set of values are mapped between one and zero with the highest value being set to one and the lowest value being set to zero. All the other values are somewhere in between. To normalize a set of numbers, each number is subtracted by the minimum value in the group and the result is divided by maximum value minus the minimum value. Here are the example standings again with columns added for the normalized values of PCT, PF and PA. I have highlighted the highest and lowest values for each component.

Normalizing the streak value is a bit more complicated since there can be winning streaks, losing streaks, and even tying streaks. Before doing the normalization, the formula converts winning streaks to positive numbers, losing streaks to negative numbers, and tying streaks of any length to zero. Here are the streaks for the example teams with the plus-or-minus values and the normalized values. Unfortunately, my snapshot from last year does not include any teams with a streak of ties, but if there were teams on a tie streak those values would be set to zero which would fall somewhere near the middle of the range after being normalized.

Of course, the four components are not all equal. I think winning percentage and points scored are more important factors and should have more influence in the formula. Streak should have less influence, and PA should have the least. The weight assigned to each of the components can be easily changed in the spreadsheet. I have the weights listed as percentages. The weights for the four components should add up to 100%. My spreadsheet does not strictly enforce that particular rule, but does make it obvious when the weights do not sum up properly. I chose to give PCT a weight of 30%, PF a weight of 40%, PA a weight of 10%, and streak a weight of 20%.

The component values for each team are multiplied by their selected weights and are then added together to give a score that should fall into the range between zero and one if the weights were chosen properly. The teams are then sorted by their scores to give the weekly rankings. Here are the scores from the example sorted from highest to lowest.

Cruise Brews

Home Brew Update: Obviously, it has been more than two weeks since I bottled the latest batch of home brew. My opinion might be due to my lowered expectations, but the beer is not that bad. It is a little flat and seems a bit sour, but it is drinkable.

Moving on to other news, El and I just got back from a seven-night cruise of the western Caribbean aboard the Carnival Valor. Our "exotic" cruise departed from the Port of Miami. After a day at sea, the first stop on our itinerary was Grand Cayman. The Cayman Islands were part of the British West Indies and are now one of the British Overseas Territories. Grand Cayman was accessible from the ship via smaller tenders. El and I elected to skip Grand Cayman and stay aboard the ship. One of the drawbacks of cruise ships are the large number of people. The Valor has a capacity of 2,974 passengers plus 1,180 crew members. Spending the day on the ship with fewer people was a welcomed change of pace. We did see a fake pirate ship that roams the bay as part of a pirate-themed excursion.

The next day our ship docked at Mahogany Bay on the island of Roatán which is off the coast of Honduras. The area is known for a great strip of beach plus various water-based activities. Our ship shared the day in port with the Carnival Dream which is Carnival's newest and largest ship. El and I disembarked for a short time, but retreated back to the Valor due to the heat and the crowds.

The third stop on our cruise was Belize. El had us booked for a seven-hour tour of a Mayan ruins site. The morning started with a ride on a tender to the port of Belize City. After arriving in port, we were herded onto a tour bus for the ride out to Xunantunich. The main tour guide jokingly told us to pronounce the name of the site like "Tuna Sandwich". The ruins are accessible via a ride across the Mopan River on a hand-cranked ferry and followed by a quick van ride up the hill.

Prior to getting to the ruins, we stopped in a small welcome center which contained a scale model of the site. Then we proceeded to the ruins. There are multiple structures that are part of the site. The most impressive is the large pyramid known as "El Castillo". Our group climbed to the top of temple and were awarded with an amazing view of the surrounding area. The guides pointed out areas that were in Guatemala which is only two miles from the site.

After descending from the pyramid, we rode back down the hill and back across the river. Then we jumped back on the tour bus and headed off to lunch. We stopped at a restaurant in a small nearby town called San Ignacio. Lunch was the best red beans and rice that I have ever eaten, but that might have been due to our tired state. We also had chicken and fried plantains. To go along with lunch, I bought us two bottles of Belikin beer. Belikin is known as the beer of Belize. It is a run-of-the-mill lager that normally would have been uninteresting. However, it definitely hit the spot at the time.

After lunch, it was back on the bus for the ride back to Belize City. Outside of the compounds run by the cruise lines, you might expect the areas of Central America and the Caribbean to be in abject poverty and extremely rundown. What I saw of Belize during the ride to and from the ruins was better than that expectation. In fact, I really did not see much difference between outlying areas of Belize and many rural areas of the South.

The fourth stop of the cruise was Cozumel which is a Mexican island off the coast of the Yucatan Peninsula. It was our second early morning, since El and I had another excursion. After debarking and meeting our group, we were loaded into a taxi for a short trip to a zip-line center not too far outside of the port compound. The guides helped us into our gear, and after a short instructional briefing, we climbed up a wooden tower. I should note here that I am not a big fan of heights. It took some effort to just pick my feet off the ground and start trusting the cable. I finally took that first step, and after that, it got easier.

After returning from our excursion, we found the Carnival Conquest docked next to our ship. The Conquest is one of the Valor's sister ships, and it was pretty impressive walking down the dock with the twin ships on each side. It was reminiscence of walking around the downtown area of a large city with skyscrapers lining the street.

Other than the bottle of Belikin beer in Belize, I did not get the opportunity to sample any local brews. During the cruise, El and I spent most of our evenings after dinner in the piano bar. I downed several bottles of Pilsner Urquell and cans of Boddingtons Pub Ale while enjoying the entertainment provided by Randy on the piano.

Day seven was another day at sea. El and I had a very nice dinner at the steakhouse located on the tenth deck. On the eight day, we returned to Miami, and then flew back home.