A blog post on the site Not In HD about what this particular blogger would love to see removed from Major League Baseball got me to thinking about something statistically. It stems from his feeling that the All-Star game is a ridiculous way to determine who should have home field advantage in the World Series, and I couldn’t agree with him more. Unfortunately, because of the current nature of baseball, straight-up Win-Loss record isn’t fair either, the schedule isn’t balanced.
Since the advent of Division Play, the idea that you play every team in the league the same number of times has gone out the window, although even then your variability was only based on full divisions, not individual teams. Interleague Play has changed all of that. No longer is a league guaranteed to have a win-loss record of .500 among all its teams, and now schedules are even more unbalanced than they ever were. There’s the issue of the four team AL West compared with the six team NL Central too, which I am fully aware of as a Reds fan! This stuff is just harder to figure out than it was before.
So because of my renewed love for great statistical analysis sites, I thought I’d ask a question that’s probably been asked before: what offsets of win-loss can I use to better determine who the best team in baseball was against their own league?
I have two possible ideas on how to do this:
- Idea #1: Use a combination of team winning percentage and strength of schedule winning percentage to determine an “index number”. Anything above .500 in each is a positive number, below is negative. So, if you had a .600 winning percentage against .495 competition, you would have a .095 index (.100 – .005). The advantage is that it directly covers strength of schedule, the possible disadvantage is that it would be swayed in the favor of the league that won interleague play unfairly.
- Idea #2: Offset the team winning percentage by the league’s overall winning percentage. This would provide maybe a .005 max offset in one direction for one league and the other way for the other league, so you could try and put both leagues back on a level playing field. It would help the issue but I don’t think it would fully help explain the solution.
So which one is right? Let’s take a look. I was able to quickly find strength of schedule information for the 2008 season, so we will go back in time a little bit.
| TEAM | WIN | SOS | INDEX | RANK | RPI | RPI-RANK |
| Los Angeles Angels | 0.617 | 0.502 | 0.119 | 1 | 0.531 | 3 |
| Tampa Bay Rays | 0.599 | 0.517 | 0.116 | 2 | 0.538 | 1 |
| Boston Red Sox | 0.586 | 0.515 | 0.101 | 3 | 0.533 | 2 |
| Chicago Cubs | 0.602 | 0.499 | 0.101 | 4 | 0.525 | 4 |
| New York Yankees | 0.549 | 0.515 | 0.064 | 5 | 0.524 | 5 |
| Philadelphia Phillies | 0.568 | 0.493 | 0.061 | 6 | 0.512 | 8 |
| Chicago White Sox | 0.546 | 0.509 | 0.055 | 7 | 0.518 | 7 |
| Milwaukee Brewers | 0.556 | 0.498 | 0.054 | 8 | 0.512 | 9 |
| Totonto Blue Jays | 0.531 | 0.516 | 0.047 | 9 | 0.52 | 6 |
| Minnesota Twins | 0.54 | 0.503 | 0.043 | 10 | 0.512 | 10 |
| New York Mets | 0.549 | 0.49 | 0.039 | 11 | 0.505 | 13 |
| Houston Astros | 0.534 | 0.502 | 0.036 | 12 | 0.51 | 11 |
| St. Louis Cardinals | 0.531 | 0.497 | 0.028 | 13 | 0.506 | 12 |
| Florida Marlins | 0.522 | 0.493 | 0.015 | 14 | 0.5 | 16 |
| Los Angeles Dodgers | 0.519 | 0.487 | 0.006 | 15 | 0.495 | 18 |
| Cleveland Indians | 0.5 | 0.502 | 0.002 | 16 | 0.501 | 14 |
| Texas Rangers | 0.488 | 0.505 | -0.007 | 17 | 0.501 | 15 |
| Arizona Diamondbacks | 0.506 | 0.485 | -0.009 | 18 | 0.49 | 23 |
| Oakland Athletics | 0.466 | 0.506 | -0.028 | 19 | 0.496 | 17 |
| Kansas City Royals | 0.463 | 0.505 | -0.032 | 20 | 0.494 | 19 |
| Detroit Tigers | 0.457 | 0.502 | -0.041 | 21 | 0.491 | 21 |
| Cincinnati Reds | 0.457 | 0.502 | -0.041 | 22 | 0.491 | 22 |
| Colorado Rockies | 0.457 | 0.485 | -0.058 | 23 | 0.477 | 26 |
| Baltimore Orioles | 0.422 | 0.515 | -0.063 | 24 | 0.492 | 20 |
| Atlanta Braves | 0.444 | 0.492 | -0.064 | 25 | 0.48 | 25 |
| San Francisco Giants | 0.444 | 0.485 | -0.071 | 26 | 0.475 | 27 |
| Pittsburgh Pirates | 0.414 | 0.503 | -0.083 | 27 | 0.481 | 24 |
| Seattle Mariners | 0.377 | 0.503 | -0.12 | 28 | 0.471 | 28 |
| San Diego Padres | 0.389 | 0.485 | -0.126 | 29 | 0.461 | 30 |
| Washington Nationals | 0.366 | 0.494 | -0.14 | 30 | 0.462 | 29 |
Info from ESPN Site: http://espn.go.com/mlb/stats/rpi/_/year/2008
So the straight-up index, and really anything that just bases on strength of schedule shows bias (and probably fairly so) on behalf of the team with the better interleague record, in this case the AL. Their record this year was 150-102. While that does not seem like a lot of games spread out over 30 teams, it was enough to give the AL a total winning percentage of .511, and the NL .491. These are of course not evened out because of the different number of teams.
This is where Idea #2 comes into play! If you want to determine which team was the best in their respective league, you have to use an offset to basically eliminate Interleague play from the equation. So, for the second experiment, we’re going to just subtract .011 from each AL team’s winning percentage, and give each NL team a .009 boost. To add some spice to it too, I’m going to do the same thing for each team’s index number, so that strength of schedule is given proper emphasis too. This might allow the in-league strength (strong versus weak division) to come out as opposed to just overall strength of schedule. Here’s what we got:
| TEAM | WIN | OFFSET | NEW WIN | NW RANK | SOS | NW IND | RANK |
| Chicago Cubs | 0.602 | 0.009 | 0.611 | 1 | 0.499 | 0.11 | 1 |
| Los Angeles Angels | 0.617 | -0.011 | 0.606 | 2 | 0.502 | 0.108 | 2 |
| Tampa Bay Rays | 0.599 | -0.011 | 0.588 | 3 | 0.517 | 0.105 | 3 |
| Boston Red Sox | 0.586 | -0.011 | 0.575 | 5 | 0.515 | 0.09 | 4 |
| Philadelphia Phillies | 0.568 | 0.009 | 0.577 | 4 | 0.493 | 0.07 | 5 |
| Milwaukee Brewers | 0.556 | 0.009 | 0.565 | 6 | 0.498 | 0.063 | 6 |
| New York Yankees | 0.549 | -0.011 | 0.538 | 10 | 0.515 | 0.053 | 7 |
| New York Mets | 0.549 | 0.009 | 0.558 | 7 | 0.49 | 0.048 | 8 |
| Houston Astros | 0.534 | 0.009 | 0.543 | 8 | 0.502 | 0.045 | 9 |
| Chicago White Sox | 0.546 | -0.011 | 0.535 | 11 | 0.509 | 0.044 | 10 |
| St. Louis Cardinals | 0.531 | 0.009 | 0.54 | 9 | 0.497 | 0.037 | 11 |
| Totonto Blue Jays | 0.531 | -0.011 | 0.52 | 15 | 0.516 | 0.036 | 12 |
| Minnesota Twins | 0.54 | -0.011 | 0.529 | 13 | 0.503 | 0.032 | 13 |
| Florida Marlins | 0.522 | 0.009 | 0.531 | 12 | 0.493 | 0.024 | 14 |
| Los Angeles Dodgers | 0.519 | 0.009 | 0.528 | 14 | 0.487 | 0.015 | 15 |
| Arizona Diamondbacks | 0.506 | 0.009 | 0.515 | 16 | 0.485 | 0 | 16 |
| Cleveland Indians | 0.5 | -0.011 | 0.489 | 17 | 0.502 | -0.009 | 17 |
| Texas Rangers | 0.488 | -0.011 | 0.477 | 18 | 0.505 | -0.018 | 18 |
| Cincinnati Reds | 0.457 | 0.009 | 0.466 | 19 | 0.502 | -0.032 | 19 |
| Oakland Athletics | 0.466 | -0.011 | 0.455 | 21 | 0.506 | -0.039 | 20 |
| Kansas City Royals | 0.463 | -0.011 | 0.452 | 24 | 0.505 | -0.043 | 21 |
| Colorado Rockies | 0.457 | 0.009 | 0.466 | 20 | 0.485 | -0.049 | 22 |
| Detroit Tigers | 0.457 | -0.011 | 0.446 | 25 | 0.502 | -0.052 | 23 |
| Atlanta Braves | 0.444 | 0.009 | 0.453 | 22 | 0.492 | -0.055 | 24 |
| San Francisco Giants | 0.444 | 0.009 | 0.453 | 23 | 0.485 | -0.062 | 25 |
| Pittsburgh Pirates | 0.414 | 0.009 | 0.423 | 26 | 0.503 | -0.074 | 26 |
| Baltimore Orioles | 0.422 | -0.011 | 0.411 | 27 | 0.515 | -0.074 | 27 |
| San Diego Padres | 0.389 | 0.009 | 0.398 | 28 | 0.485 | -0.117 | 28 |
| Washington Nationals | 0.366 | 0.009 | 0.375 | 29 | 0.494 | -0.131 | 29 |
| Seattle Mariners | 0.377 | -0.011 | 0.366 | 30 | 0.503 | -0.131 | 30 |
Once these stats hit the mark, the best team in terms of how they played in-league, figuring or not figuring strength of schedule was the Chicago Cubs. Using this logic, I would be willing to make the argument that the team most deserving of home field in the World Series this year was the Cubs if they had made it. As it turned out, by luck I think it ended up right this particular year, as the Rays were the home team. Never mind that it didn’t help them much.
NCAA Tournament: The Sweet 65, not the Horrid 96!
A week into the tournament, I’ve had more of an opportunity to think about where the tournament is headed, and why the world enjoys the current bracket setup as much as they do. I’ve also come to the conclusion that my style on progress is exactly inbetween technology and tradition. You never mess with a good thing, but you use things you never had before to make it better.
This leads us into the latest talk of the tournament moving to 96 teams. I hated it when I first heard it, I hate it more right now after having seen a tremendous first two rounds. If you went to 96 teams, you wouldn’t have seen any of the drama you did, and if you saw upsets, they would be so much less compelling.
First of all, at the number it stands at now, the tournament is truly one of champions. You either get in by winning a conference tournament, sans the Ivy League (and they don’t mind, see Cornell), or by being a top team who in almost every case has a winning conference record. The move to 96 is unquestionably one to get more big conference big name teams in the mix, and it would take one look at the NIT to see some of the teams who would have made it. North Carolina at .500 overall?! UConn at 7-11 in the Big East? These are teams that do not deserve a shot at the title, and if they qualified, would be facing these smaller conference champions. If the small conference team wins (say, Ohio?), it’s so much less impressive. If the big team wins, it creates forgettable matchups. Trust me when I say that the big schools will get the 9-16 seeds, the small conference winners will still be at the bottom barrel.
The NIT is an appropriate place for these teams, especially considering that they reward any regular season conference champion with a post season trip. Good for them for doing this.
Maybe in my next post I’ll lay out what I think would have happened with Ohio if the tournament was 96 teams big. We’ll see.
The other reason we know the move wants to happen is for a money grab, and from the sounds of it, a chance to redo a TV contract to get another billion dollars. Did we mention the NCAA is a non-profit? They might get a little bit more short term, but with the charm of the tournament severely damaged, the tournament would not draw more money long term. I do not believe it’s worth it.
The last reason is one that sports pundits can talk about, and one that the NCAA profits from what will never completely acknowledge: gambling! Brackets are typically legal, even when money is pooled, and Vegas sports books are as good as anyone as policing the sport, even if the NCAA hates it. Can you imagine filling out a 96 team bracket?! How many people who casually follow the tournament because they could pick that cool upset would just stop caring? Tons, and that is a lot of sponsorship money down the toilet from sites who want in on the coverage, and from sponsors who support the NCAA greatly. This type of gambling absolutely matters to the NCAA, even if they would never admit it, and this would be dimished in a 96-team field.
I want to watch a tournament of great teams and winners, not a tornament of ho-hum and very good. I want to have fun watching the underdog get to play the big guy, I want those Northern Iowa moments, I want Cornell to play the best, not some has been under-500 conference team. I hope every year we get a team like Davidson was last year. 96-team tournaments lose too much of that, and the game will suffer.