Monday, November 17, 2008

Baseball Economics

Brewers GM Doug Melvin might be a very intelligent man, but I couldn't help but cringe when I read one of his comments on the NY Yankees' pursuit of CC Sabathia:

“It sounds like they’re overbidding. If the speculation is true that we’ve offered CC $100 million, why would you offer $140 million? Why wouldn’t you offer $110 million?”

Speculation was that the Brewers offered CC $100 million over 5 years. The Yankees' offer was a six-year deal. Making a 6-year, $110 million offer shouldn't sway Sabathia toward the pinstripes. That's less per year than the offer from the Brewers! Melvin might have been much better off giving a per year figure, perhaps stating that the Yankees don't need to offer $23.3 million per year when the Brewers' offer was for a meager $20 million per.

I see where Melvin is coming from, but I'm not sure he has enough perspective on this issue. His point is that the Yankees are outbidding the other teams by more than a few million dollars. But, does he fail to realize that the Yankees might have to do that to get Sabathia?

The Yankees are an attractive location for players who want to win championships, because the team is always in the mix. Sure, they haven't won the World Series lately (was the last time really the 2000 Subway Series against the Mets?), but they have the resources, and management desire, to try to win every year.

However, there are also a lot of issues working against them, especially with Sabathia. First, not everyone wants to be in the circus that is associated with being a member of the NY Yankees. Sabathia seems content to play in smaller markets with a little less spotlight. And, he's from the left coast ... so location, location, location isn't working for the Yanks. Additionally, Sabathia is a good hitter. If he goes back to the AL, he gives up most of his opportunities at the dish and has to pitch to the opposing DH's.

So, maybe the Yankees feel that they need to give Sabathia $3 million extra per year and a sixth year to get him to sign on for 6 years in NYC pitching in the AL. Maybe it's enough ... maybe it's not. But, I think the Yankees really want Sabathia and are willing to overspend a little bit for the guy they want - and if the Brewers are willing to go $20 million per, I don't really think the Yankees are overspending.

You can tie in the NFL? Really?

Apparently, Philadelphia Eagles QB Donovan McNabb didn't know you could tie in the NFL:

"I've never been part of a tie. I never even knew it was in the rule book. I was looking forward to getting the opportunity to get out there and try to drive to win the game. But unfortunately with the rules, we settled with a tie."

I was flabbergasted when I heard this. I'm not sure when I first became aware of that particular NFL rule, but I'm sure I had yet to finish elementary school at the time. For a QB in the NFL to not know the rule is incredible, and not in a good way.

And, the relative infrequency of this result leads me to believe that the current system is fine. If you don't score in the 15-minute period, you don't really deserve to win. The Eagles had three full possessions in OT: (1) 4 plays, 19 yards, (5) 5 plays, 9 yards and (3) 3 plays, 9 yards. The Eagles turned the ball over 4 times and converted just 3 of 18 3rd downs (although, they were 1 for 1 on 4th down). They averaged less than 6 yards per pass and less than 4 yards per rush against the previously 1-8 Bengals.

If you want to argue that it's not fair if you don't get a chance to play offense in OT, fine, I'll listen. But, if you want to argue that you should just keep playing ... no way. Ties are a part of football. There isn't any reason they shouldn't be. It's not inequitable to only give teams 15 extra minutes to score. Get a few first downs and get in FG position!

All that being said, this was an awful result for the Eagles. They'd be tied with the Redskins and Cowboys at 6-4 if they had defeated the Bengals. Oops. Now, they're a half game back. They're 1.5 back of the Bucs and 0.5 back of the Falcons from the NFC South. So, rather than being 1 game back of one team and tied with 3 others in the NFC Wild Card race, they now sit in 5th place in the race for 2 spots.

I disagree with President-Elect Obama

Apparently, Barack Obama doesn't read "There's a Catch." In a recent interview, he said "We should be creating a playoff system. 8 teams, that would be three rounds to determine a national champion, it would add three extra weeks to the season - you could trim back on the regular season. I don't know any serious fan of college football who has disagreed with me on this ..."

And, while I don't like the BCS, I've heard a lot of serious fans of college football say that they don't want a playoff because it would devalue the regular season. I don't agree with them, but they're out there.

If Obama wants to sacrifice regular season games to implement a playoff, non-conference games will be cut. This will further reduce the information we have about which conference(s) are the best. Right now, Pac-10 teams play 3 non-conference games. If you cut two games out of the regular season to implement a playoff, you're left with 1 non-conference game. There will be even fewer chances to lose during the regular season under that scenario, so teams will be even less likely to schedule difficult non-conference game. In addition, they'll probably play the non-conference game first in their schedule to prep for conference play. While Ohio State has been playing huge non-conference games to kick off the season lately, not many other teams want to do that. Teams with just eight conference games would be left with just two non-conference games. What are the chances that Brown, Meyer, Miles, Saban, Stoops, etc. schedule top 10 foes for those contests?

It isn't as easy as just taking 8 teams. How do you pick the 8 teams? Right now, Utah, Boise State and Ball State are all undefeated in non-BCS conferences. We could have a three-way tie for 1st place in the Big 12 South at the end of the season. It's possible that USC will not win the Pac-10, despite having just one loss and being ranked in the top 8.

If you take the champs from the 6 BCS conferences and then the next two highest ranked teams, it's unlikely any of the non-BCS undefeateds would make the field of 8. Barring anything drastic (Tech getting steamrolled in Norman AND the Sooners getting taken out by the Cowboys in Stillwater), at least two of the Big 12 South teams will be in the top 6 at the end of the season. Also, if Florida beats Alabama in the SEC Championship game, those two will still both be ahead of the highest ranked non-BCS team, Utah. So, the two non-champs would be from the SEC and Big 12.

I don't like Obama's idea, obviously, and these are a few of the reasons:
1. It increases the penalty for any loss
2. We have even less information about how Conference X compares to Conference Y
3. Is it reasonable to expect some teams to go to bowls while others are playing in a playoff? And, if you lose in the first or second round of the playoffs, do you still get to go to a New Year's Day bowl?
4. With even less information about how the conferences compare, how do you pick the top 8 teams? What if 5 or 6 of the top 8 are from two conferences?

Sure, the current system is flawed, but the changes that need to occur aren't trivial.

Wednesday, November 05, 2008

The Tromble Plan for CFB

We're headed for another disaster with the BCS. Penn State appears to be headed for an undefeated season, which should allow them to get into the national title game without being one of the two best teams in the country. And, we might have a scenario where the other team in the national title game is taken from a one-loss crew including USC, Texas, OU/Texas Tech/OK State, Alabama, Florida. And, we could have multiple undefeateds from the non-BCS conferences: Boise State, Utah and Ball State that would be left out.

I've seen a clip of Barack Obama on ESPN multiple times today (the same clip, multiple times) calling for an 8-team playoff. My question is: which 8 teams get to play?

For kicks, I'm going to play out the rest of the season in a reasonable manner.
SEC: Alabama and Florida win out the rest of the regular season, but UF takes down the Crimson Tide in the SEC Championship game. Alabama and Florida each have 1 loss on the season.
Big 10: Penn State continues on undefeated.
Pac 10: USC rolls the remainder of the season, winning the conference with just the loss to Oregon State.
Big 12: Texas wins out. Texas Tech beats OK State and loses at OU. But, OU loses to OK State in Stillwater. So, Texas and Texas Tech have one loss each, with Tech having the tie-breaker of head to head. OU and OSU each have two losses. I set it up this way because I didn't want the worst-case scenario. It's entirely possible that we could have three teams from the Big 12 South with just one loss at the end of the year, but let's not go there for now.
ACC and Big East: Someone wins each of those conferences with just 2 losses on the season, let's call it WVU and Florida State.
Other conferences: TCU takes out Utah, but Utah beats BYU. TCU and Utah each finish the season with one loss (TCU's only loss is at OU early in the season). Boise State and Ball State run the table.

If you want an 8-team playoff with all undefeateds and the winners of the 6 "BCS" Conferences, you'd get the following field:
Ball State
Boise State
Florida
Florida State
Penn State
Texas Tech (via tie-breaker over Texas, presuming a win over the Big 12 North champ)
USC
WVU

If you leave out Ball State and Boise State, you could insert Texas and Alabama. But, is it fair to leave out undefeated teams?

We don't need a playoff added on to the current format, we need a better system for figuring out who the best teams are!

Who was the best team in the NFL last year? It was the Patriots. They didn't win the Super Bowl, but they were the best team. Does the best team in college basketball usually win the national title? That's debatable, but the 64-team single elimination tourney isn't the best way to have the best team win. The larger the single-elimination tournament, the more likely that the best team will lose to a team they shouldn't lose to. Maybe you hit a bad match-up ... maybe your team has a bad game ... maybe you get shafted by the refs ... who knows? The solution is a systematic way to determine who the best two teams are ... then having them play in the national title game, winner take all. How do you determine the two best teams? Well, you get intelligent people to write code that analyzes results from the entire season to rank teams.

The current system doesn't work because there aren't enough meaningful results. And, it's going to take a governing body controlling scheduling to fix the problem. Here's how you do it.

1. Have teams play two pre-season games to get ready for the season schedule.
2. The season schedule starts with conference play, with 8 games for most teams over 9 weeks (teams in conferences with 10 (or 11) teams would play all 9 weeks).
3. Conference championships would be the next week, plus a meeting between the winners of the Big 10 and Pac 10 (site rotating between the Rose Bowl and some Big 10 country site).
4. Also, during conference championship week, the non-conference schedule would start for teams that aren't in conference title games. Then, for the rest of the season (3 or 4 more games), teams would continue to be matched up.
5. At the end of the season, the rankings would list teams from 1-nt, where nt is the number of teams, whatever it happens to be that year. The top x will go to bowls (x is equal to 2 times the number of bowls that year). There are two options for the championship: i) the top 2 teams play for the national title or ii) the top 4 play a single elimination tourney with the national title game being played the week between the NFL Conference Finals and the Super Bowl.

How does the scheduling work? Has everyone out there taken the GRE? If the first math question is 1+1= ? ... and you answer 2, then you'll get a tougher question. If you get it wrong, well ... uh oh. Perhaps the next question is: 2+x=8 ... you say x=6 and get a more difficult question. The third question is x^2=4. If you say that x=2, you get it wrong (x can be either 2 or -2) and you get an easier question for the next one. In this way, the system can more accurately gauge your actual aptitude than with a set bank of questions that everyone answers. It would take some flexibility by fans, but I'm sure people would make accommodations on short notice for these games. Perhaps, you could even lock in certain weeks that were guaranteed home games, although the opponent would be up in the air. I would start by suggesting that match-ups are scheduled between 2-3 weeks in advance, based on the rankings at that point. The previous year's rankings could be used to help create the early season rankings, because non-conference results will be non-existent.

We think the Big 12 is the best conference, but we don't really know. Maybe it's the SEC. Wouldn't it be nice is we could line up the following set of games one weekend to help us figure things out (kind of like the ACC-Big 10 challenge in BB):
Texas v. Florida
TTech v. Alabama
OU v. Georgia
OK State v. LSU
Missouri v. Mississippi
Kansas v. Vandy
Nebraska v. South Carolina
A&M v. Auburn
K-State v. Kentucky
CU v. Arkansas
Baylor v. Miss. State
Iowa State. v. Tennessee

Based on the results of the games, the rankings would be reshuffled. And, by having lots of meaningful results, we could sift through the teams and actually figure out who the best teams are.

Or, we can continue to watch Alabama play Arkansas State and OU host Chattanooga. And, we'd be able to work in Boise State, TCU, Utah, BYU, and Ball State against strong teams from "BCS" conferences to see if they are legit or not. Tulsa was undefeated until they lost (barely) to Arkansas last week. Who else did Tulsa beat in non-conference games: North Texas, New Mexico and Central Arkansas. If Tulsa had beat Arkansas and finished undefeated, would we really know if they were any good?