## Thursday, October 15, 2009

### Playoff DIscussion

Mike "the bot" Hearn:

This morning, I calculated this:

[Sport/Full Season/First Round Playoff Length/Percentage] –
Football/16/1/6.25%
Baseball/162/5/3%

Conclusion: baseball has the shortest first-round playoff of the three major sports by more than 50%.

[Sport/Winning percentage of top quarter of teams/Winning percentage of bottom quarter of teams/Difference] –
Football/73%/21%/52%
Baseball/59%/40%/19%

Conclusion: baseball has more parity by far than the other sports, meaning odds of the better team winning are lower than in any other sport.

Together these two conclusions essentially mean that a five-game playoff series is a coinflip. After 162 baseball games, the good teams and the bad teams are only separated by 19% points, so imagine the parity of putting almost equally good teams together and having them face-off in a series that is not long enough to statistically determine who is better. It’s meaningless.

What if a prize-fight only lasted one round? Or if the Super Bowl were 28 minutes (which is 3% of a total season, by the way)? It would be bullshit, and no one would accept that the victor was the better team because there WASN’T ENOUGH TIME TO DETERMINE ANYTHING. In a five-game baseball series, your star pitcher could have a bad day or A-Rod could go 2-12, whatever, and your season is OVER. Randomness rules over consistency as time span decreases.

To properly choose the length of the playoffs, two factors need to be considered: parity, and boring the crap out of people. Now to TRULY establish which team is better, a series would have to be INFINITE games, and the wins/losses would slowly approach the true mean of how good two teams are. In 1997, maybe in an infinite series the Indians would win 60% of the games and the Marlins 40% (but the Marlins win 4 of the first 7, sucks for you guys). However, infinite series are obviously unfeasible. So you need to pick the highest sample size (i.e. number of games) that will reflect the quality of a team without boring the crap out of an audience –

This is when parity comes into play. If in an infinite series (the ideal), one team wins 51% of the games while the other 49%, in reality you would need A LOT of games to establish this. Whereas if the split was 90%/10%, you could probably establish it in a 5 or 7 game series. This is why football (differential: 52%) can survive on one-game playoffs – because the good teams beat the bad teams. Same for college basketball’s March Madness. The NBA is king here – they have high parity (45% differential), and a high season/playoff % (8.5%). This is why the same good teams tend to win the championship: Bulls in the 90s, Spurs, Lakers, Pistons. The good teams win the championship. Astonishing.

Alternative #1: 9 game series. Why the fuck do baseball teams take breaks in between games in the playoffs? They have like one off-day every week during the season, then all of a sudden during the playoffs they get like every other game off. Play 9 games in 11 days. Fuck it. Also, your good pitchers would go 3 times, your good batters would get more than 15 ABs in the series.

Alternative #2: Bring back the old East/West system. There are 162 games in a season; it’s a pretty good system for figuring out who are the best teams, so you don’t need an elaborate playoff. Why are 8 teams in the playoffs? Just put the two clearly best teams into the AL/NL Championship series and then the winner plays in the World Series. It worked for like 60 years, FUCK YOU BUD SELIG.

Alternative #3: Allow steroids. This would not fix the parity issue but I’d get to stop hearing about it and that is so much better than having a legitimate champion ever year.

Matt Glassman

Ok, first off allow me to say that I think playoffs in ALL sports are bad. They are outdated. Playoffs were originally intended to find out which of the two leagues (AL and NL), two leagues with different governing bodies, different rules, and different groups of players was superior. Now, you have a standardized canon of rules, one draft pool, and 162 games to figure out who the best team is. If it were up to me, the NY Yankees would be hoisting a trophy right now. This is why I think the much beleaguered college football system is the best right now. Each week, regular season games are really fucking important. Baseball games are irrelevant for the most part.

Second off let me say that of the 3 major sports playoff systems, I think baseball is the most entertaining and BY FAR the best judge of superior talent, and I will explain in a moment.

In his e-mail, the bot showed a very well-researched and incredibly shortsighted and closed minded argument. First off, his formula to find parity is a very rudimentary and simple one. It ignores the fact that as teams play an amount of games approaching infinity, they will approach a .500 winning %, so of course baseball has the most parity, they play 10 times more games than football. Furthermore, he ignored the middle half of teams, which is where parity is usually found. 2 years ago, the super bowl winner was outside that top quarter and yet crowned champion, a glowing example of parity, but one ignored by the Bots formula.

Furthermore, he spends a lot of time on the percentage of the regular season that the playoffs represent, and I just don't see how this is valuable. In fact, I feel like one could make a very strong argument than the SMALLER the % of the regular season the playoffs represent would be more desirable hence placing a higher premium on the 6 month, 162-game regular season. This argument is easily made by looking at the utter irrelevancy of the NBA regular season given the 2 month-long, up-to-28 game playoff season nullifies so much of the regular season that it becomes almost boring. But that is not the crux of neither the Bot's nor my argument. Just something that should be noted.

He also points out that the first series is a "coin-flip." Well, bot, consider these facts: In basketball, the number 1 seed plays the number 8 seed in the first round. Often this is the best team in the league against the 16th best team in the league. Considering there are 29 teams, the top team, the favorite to win it all, is playing a team that finished 2 spots closer to last place than they did first place. That's a considerable adv. In baseball however, the Dodgers are the 4th best team against the 7th best team. The Red Sox and Angels represent the 2nd against the 3rd. The Phillies are the 5th and the 6th, and the Yankees are the best against the 8th. All of these are fairer and hence lead to a more "coin flip" atmosphere. Obviously if infinite games were played, the better team would always win, but in this case, that is not possible, and 5 games will have to do.

It should also be noted that when the NBA did have 5 game series, that the 8 seed only beat the 1 seed twice in history - once in the strike shortened 1999, and once in what many consider to be the greatest playoff upset in league history, the 1994 Denver Nuggets. So 5 games was always enough for the better team to win in basketball.

As far as football goes, it is true that their one-game playoffs are exciting and proportional to the regular season, but a football game more than any other sport is a coin-flip. Football games (between two good teams) essentially come down to 5 big plays. The bot argues, "what if a pitcher has a bad game, or A-Rod goes 2-12?" Yeah, but what if a QB hurts his ankle on Friday and can't play? What if a star running back has his first fumble in weeks during a critical drive? What if a kicker, who hasn't missed all year, misses a 40-yard field goal as time expires (see 2005 Colts/Steelers.) Things happen, bot. It's the playoffs. And considering the season starts in trianing camp in May, and can end on the mistake of one man's foot in January is doubly as unfair as CC having two consecutive bad starts in the 2007 ALCS. The season ending because of a bad pass interference call (see 2008 Colts/Chargers) is wayyyyy worse than A-Rod failing to get an RBI in 4 straight games in the 2004 ALCS. Sorry, but that argument doesn't fly.

And then he makes the argument that the NBA consistently puts the best teams in the finals because of their marathon playoffs. Well, I have two responses to that.

1. If that were the case, than wouldn't baseball's regular season consistently put better teams in the playoffs year in and year out due to having twice the regular season? This year, 50% of baseball's 8 playoff teams repeated as playoff teams. Last year's top 8 in basketball repeated 62.5% of the playoff teams. Despite having 50% the games. Which brings me to point #2.

2. The nature of the games--basketball and baseball--are different. In basketball, having a super-duper star like Lebron or Kobe essentially guarantees you a top 8 spot, while in baseball, the Yankees finished outside the top 8 in 2008 with the most super stars. The sports are just different and hence the results of the champions should be different.

Now I will address his ways to make it besser:

1. 9 game series. . What you said about playing every day for 9 games and the best pitchers pitching 3 times. Assuming they play 6 games, day off, 3 games. The best pitchers would pitch twice. Games 1 and 6. The same amount of games the best pitchers pitch now: twice. In addition to this, the number 5 starter--the guy who under the current system doesn't make the roster would pitch twice as well--including game 9. So the World Series could come down to Paul Byrd against some musher on the Cards. That's fair. Joke. Not really.

2. I'd be all for bringing back the old East v West system, but what's the point? You still have a 7 game series to even play for the World Series. This is a small band aid that doesn't solve anything.

3. I'm all for bringing back steroids.

Also, it should be noted that the bot is basing this info on numbers alone. He doesn't watch the games. He doesn't remember the magical 95 Yankees Mariners series, or the 1999 Tribe Red Sox. Watch a baseball playoff game and tell me that it is not the best system. Look, unless you played an unlimited amount of playoff games, you're never going to figure it out. But consider this bot, 2 years ago; the New England Patriots went 16-0 in the regular season. First team to ever do that. They were the best team, maybe in NFL history. Then, they lost to the New York Giants in the super bowl who finished 10-6, the same record as the Cleveland Browns, who didn't make the playoffs. So somehow, to you, a system in which a team with the same record as a non-playoff team, beating the best team in NFL history is a fine way to judge, but 19 games in 4 weeks started by the best pitchers in the game, in the loudest stadiums where every pitch matters is bad.

The Bot:

Hey glass, thanks for the response.

I'll start off by clarifying the parity situation:

If I flip a coin 1,000,000 times it will come up 50/50, because both heads and tails have p=.5 chance of coming up. Not the case in sports. In a theoretical infinite series, the better team will win > 50% of the games. If the Yankees and my local little league team played an infinite series, the Yankees would still win 99.99% of the time despite the infinite nature of the series. It would never approach 50%, it would approach the true mean of their team's abilities. Baseball has an incredibly high sample size, 162 games, and yet their parity difference is only 19%. Football's difference was 52%, which I admittedly based only on the past season, but I expect if I took winning/losing percentages of the top/bottom teams from each of the last 10 years (meaning 160 games), the difference would still be roughly 50% (meaning the good teams win 70% of their games, the bad teams win 20%).

In these sports, it doesn't have to do strictly with the sample size, it has to with the odds that in ONE GAME a good team will beat a bad team. In baseball, there's a higher probability of this happening than in other sports -- I don't know why, but that's how it is. The best record of all time was only 116 wins for the Mariners (71% win percentage). Four NBA teams had a higher percentage of wins LAST YEAR -- Cleveland (80%), LAL (79%), Boston (75%), Orlando (72%). That's just how it is. In a baseball game, a worse team has a statistically higher chance of beating a good team than in other sports. In my first note, I demonstrated this with the parity calculation using win percentages of good and bad teams, but I'll just say it in plain English -- in baseball, worse teams win more often, and the randomness of any particular game is much higher than in other sports. This is objective. If the Cavs played last year, they had an 80% chance of winning. If the Mariners played, they had a 71% chance. Which game would you bet on?

So -- having said that: due to the closer parity, you NEED A GREATER SAMPLE SIZE to have a truer determination of the better team. That's all I'm saying. A five game series is utter randomness because ANY TEAM CAN WIN when the difference in win % between the two teams is minimal. If I have a coin, in a 5 game series heads might win 3-2, or 3-1 or even 3-0, but in an infinite series it will approach 50% for each. Similarly in a 5 game series between baseball teams, even when one team might win 60% of the games in an infinite series and the other 40% (which, note, is roughly the parity difference in baseball as a whole), the worse team has a good shot of winning a 1 game series, a slightly worse but still good shot of winning a 5 game series, same for 7 games, 9 games, etc -- but as the series gets longer, the worse team's chances of actively winning the series decreases. In a 99-game series, the better team will win about 60 games, the worse about 40. When I say "approaching the mean", that's what I'm referring to.

Other notes:

--The baseball season does a great job of putting the best teams in the playoffs. I'll never counter that. A 162 game season is great at that. It's the playoffs I take issue with.
--In a 9 game series -- c'mon, you know managers would throw their aces out there in games 1, 5 and 9. They try to pull that shit in 7 game series, usually to catastrophic results.
--In football it's true that one individual play can change the outcome of the game, yet despite this the good teams still win a higher percentage of their total games than good teams do in baseball (football: 80%; baseball: 60%). I don't know why this is, but football is apparently a much more controlled game than baseball.
--In this argument, I'm not taking into account how good teams are across seasons, because so many factors can change. Somehow the Yankees weren't in the best 8 teams in 2008, despite their stacked roster. The 162 game season is more than enough sample size to prove this. Something on that team was wrong that couldn't be accounted for in their lineup (wait, their pitching -- it was terrible).
--The east/west series works because it keeps the riffraff (i.e. the wildcard) out of the playoffs (a clusterfuck where anything can happen); basically, it prevents undeserving teams from benefiting from the randomness that extreme parity the playoffs results in.
--The 2007 Giants are an interesting example of the randomness of having a one-game playoff system. I can't really account for them, other than to say that if football had a 3 game series in each round, the Giants likely would not have won. Again, sample size is important, but I accept the one-game series in football because of the lack of parity (i.e. better teams usually beat worse teams).
--Basketball, again, is the gold standard of the best team winning the playoffs. I don't know why they got rid of the 5-game series for basketball's first round, frankly, because if ever there was a sport that could actually use LESS GAMES, it's them. Good teams win 80% of their games, and they're usually playing terrible teams in the first round. They don't need 7 games to figure this shit out. I think the Bulls swept their first-round series in the 90s, like, every time. They could also cut off the first round entirely; because again you don't see the #16 team every winning shit, but whatever. I'll let basketball do what it wants.

Closing argument: the baseball playoffs may be exciting, but the wildcard has won the World Series FOUR TIMES since 1994. That is bullshit. But because in the baseball playoffs anyone can win when the series is short, they have to either (a) make the series longer to make sure the wildcards get beat by the better teams, or (b) make the playoffs more exclusive and eliminate average teams from benefiting from the almost coinflip results of a baseball series.

Glassman:

OK, I'm going to re-address the bot's issue with the MLB playoffs. First off, allow me to say that I think that the parity question really is an example of how you want to measure parity. One way of looking at it, is to compare the top to the bottom and how far they are from each other (which is what the bot has done.) Another way would be to look at how many teams comprise the middle (or in this case, low-end playoffs teams) in which case football would far and away win. Another way to judge parity would be to examine how often the same teams repeat in the playoffs, in which case basketball would be the sport with the most parity. In any case, it is somewhat of an irrelevant question.

The bot points out that the wild card has won the World Series 4 times since 1994 (there was no WS in 94, so he means since 95). He uses this is as an argument to prove that too often, a bad team (what he refers to as "riff-raff" wins the world series over far better teams, hence proving that baseball's playoff system does not provide a solid enough foundation to put the better team as a champion). What he leaves out is how baseball's wild-card system works. He assumes that the wild card is the 4th best team in each league and represents the 7th and 8th best reg. season records. That could not be further from the truth – in fact, the wild-card team regularly is not in the bottom tier of playoff teams. This year's wild cards, the Red Sox and Rockies were the 3rd best and 6th best teams in the major leagues respectively. By the way, both of these teams were eliminated in the first round. Last year, the Red Sox (4th best team) and Brewers (6th best team) were the wild cards and while the Red Sox beat the Angels in the playoffs before losing to the Rays, the Brewers were swept in the first round.

He then mentions that 4 wild card teams have won the championship, and that that is BULLSHIT. Unfortunately, he declines to note that in that same time period, 4 NFL wild-card teams have ALSO won the championship, and ignores this significance.

The 4 wild-card teams that won the championship are the 2004 Red Sox, the 2003 Marlins, the 2002 Angels, and the 1997 Marlins. These teams were the 3rd best, 6th best, 5th best, and 4th best teams in the MAJOR LEAGUE'S respectively. I'm sorry, but in 15 years, having the 6th best team in baseball win ONCE, while the rest of the teams have come from the top 5 doesn't seem so bad to me.

The 4 NFL wild card teams that won the super bowl are a different story. These teams are the 2007 Giants, the 2005 Steelers, the 2000 Ravens, and the 1997 Broncos. These teams were 9th, 9th, 3rd, and 4th best teams in the NFL. Somehow, this fact is OK.

He also points out that the NBA's marathon playoffs are the best to determine the best team, and while I agree KIND of on that, I stick by my argument that it is too long and arduous and nullifies the regular season. I will expand on this argument more in a moment.

By the way, his main argument is against the randomness of the 5 game first round series. Yet somehow, despite coin-flip odds, the NLCS is a rematch of last year's NLCS and the team with the better regular season record won every series this year. In fact, let's look at something really quick.

In the last three years, the MLB's first round (the round of 8), the team with the besser regular season record is: 10-2. This is 5 game "coin-flip" series, over the course of three years. In that same time period, in the round of 8 in basketball (7 game series, better judge of the besser team), NBA teams with a besser record went 9-2, with one series pitting two even teams against each other. So by default, baseball did besser, but even if one says even, to call the MLB first round BULLSHIT, when in the last three years if has performed as well, if not besser than the gold standard NBA is bullshit in itself. In case you were wondering, the NFL's round of eight in the last three years, is 4-7 (with one game featuring evenly matched teams).

The point is this: playoff strength boils down to a philosophical question. The NFL is the league the best rewards its teams for a strong regular season, DESPITE the regular season meaning nothing. The top 4 teams in the league receive BYES, making a first round victory a literal certainty. Furthermore, the best teams need to play ZERO road games. However, despite this advantage, the top 4 regular season teams are 9-11 in the last three years of playoffs with one champion. I don't get HOW the fuck the bot is fine with that.

The NBA, which obviously is the longest, and probably best, who also rewards the best regular season teams quite handsomely (Games 1 and 7 at home, in the sport with the greatest home-field adv BY FAR). The top 4 teams in the league in the last three years are: 24-10 in series with two champions.

The much-maligned MLB, with it's "coin-flip" playoff scenarios, and limited home-field adv. is somehow 14-10 with two champions. Also keep in mind that baseball teams aren't afforded the 1-8, 2-7 matchups that basketball teams are. Remove those from the NBA and they go down to 13-9, pretty even with major league baseball, with the same amount of champions.

The philosophical question is this: why do we have playoffs. Is it really do find out who the best team is? Because I feel as though 162 games tells us that. And if it really is, than anomalies like the 2007 NY Football Giants and the 2005 Steelers should NEVER happen, let alone twice in three years. Baseball is the best regular season because of its length and because of its symbolic meaning. But if you made the playoffs any longer, they would stretch into November, with rain, snow, and freezing temperatures, as opposed to the occasional game like Sunday's CO/Phi game, that would be the norm.

I agree that 5 games is a short series, but I disagree that it nullifies the regular season or that it is not fair. The team with the besser record has games 1 and 5 at home, as well as a 60%- 40% home field adv, as opposed to 56-44 in basketball or 100-0 in football; the best pitcher in the series will pitch twice, and if needed the second best pitcher will pitch twice. It allows the wild card, which despite the bot's RIDICULOUSLY short-sighted analysis ACTUALLY allows for a top tier (meaning top 5 MLB) team to get into the playoffs despite their unfortunate division placement. It gives hitters 20-28 at bats—more than enough for averages to matter—and exposes teams weaknesses (see: Bullpen, Rockies; lineup depth, Stl). Furthermore, these series have consistently been exciting, worthwhile, and competitive (as opposed to the NBA first round) and consistently allows the better team to advance (as opposed to the NFL).

Playoff baseball is DIFFERENT than reg. season baseball. Your top 2 starting pitchers become MUCH more important, and your bullpen and defense get exposed, but it was like this BEFORE the wild card, and before the fucking AL/NL CS. That's one of the beauties of October baseball.

Now, I think I'm done, I have to go get ready for the second round, featuring the 1st, 2nd, 3rd, and 4th best regular season teams in baseball. Bot, you have officially been served.