How many playoff teams?

[Le Grande Orange]

Originally Posted by NoOne

regular season -- larger sample provides greater certainty that the best teams are at the top of the standings. its function, whether they understood it or not, is to weed out the lucky and the poor. (they = people responsible for SOP of mlb)

Except that, even over that sample size, teams can still end up being lucky or unlucky. See Phil Birnbaum's article "Which Great Teams Were Just Lucky?" from issue #34 of The Baseball Research Journal.

Luck was defined this way:

• The club's hitters have career years, playing better than their talent can support
• The club's pitchers have career years, playing better than their talent can support
• The club was more successful at turning base runners into runs
• The club's opposition was less successful in turning base runners into runs
• The club won more games than expected from its Runs Scored and Runs Allowed

These five items are demonstrated statistically in the article. At the end of the article a list of the fifteen luckiest and unluckiest teams from 1960-2001 is given. Below are selected luckiest and unluckiest clubs from that list, with teams from both leagues before 1969 and from 1979-93 in the AL and 1993 in the NL (since balanced schedules were used in those league seasons). The club's actual record listed first and its luck-adjusted record second; the difference in wins is in parentheses.

1960 PIT: 95-59; 76-78 (-19)
1992 OAK: 96-66; 77-85 (-19)
1962 LAN: 102-63; 84-81 (-18)
1961 NYA: 109-53; 91-71 (-18)
1993 SFN: 103-59; 85-77 (-18)
1963 NYA: 104-57; 87-74 (-17)

1962 NYN: 40-120; 61-99 (+21)
1979 OAK: 54-108; 74-88 (+20)
1987 CLE: 61-101; 81-81 (+20)
1961 PHI: 47-107; 66-88 (+19)
1962 CHN: 59-103; 77-85 (+18)
1964: KC1: 57-105; 75-87 (+18)
1960 DET: 71-83; 87-67 (+16)
1985 CLE: 60-102; 76-86 (+16)

A difference of sixteen to twenty wins from luck is substantial.

Originally Posted by NoOne

i like playoffs, but i don't want ~1/2 the league in the playoffs like hockey does it.. that's at total joke.

It's only half now because the NHL has expanded over the years. When it originally increased its playoff field to sixteen, there were only twenty-one teams in the league. That's 76% of the league.

Originally Posted by NoOne

playoffs are silly and just for fun as far as picking a "winner". it's far from conclusive. the regular season likely tells you with more certainty whom is better. logic ignored, people like a championship game. there is no logic for it, because it doesn't provide an answer to the question of who is better in any meaningful way.

In the minors, some leagues, for a time, differentiated between the first-place club and the playoff victor: the team finishing first was the league champion while the playoff winning club was the playoff champion. But as the years went on this differentiation evaporated and the league champion is now the playoff champion.

[daves]

Originally Posted by Brad K

So what would the playoff schedule be?

Worst 1st place record gets to play against the best 2nd place team in a single elimination.

[Brad K]

Win a division and get a single game playoff against the second-place team while the other division winners get a bye?!?!?!? Don't think that can be sold.

Actually both the 5 and 7 team playoffs are impractical for a 4 division league. Not picking on you here but saying what to do without considering how is rampant in this thread. And as a belated response to the slightly veiled old school criticism, I'll call it a new school sort of thing.

[Brad K]

Originally Posted by Le Grande Orange

Except that, even over that sample size, teams can still end up being lucky or unlucky. See Phil Birnbaum's article "Which Great Teams Were Just Lucky?" from issue #34 of The Baseball Research Journal.

Luck was defined this way:

• The club's hitters have career years, playing better than their talent can support
• The club's pitchers have career years, playing better than their talent can support
• The club was more successful at turning base runners into runs
• The club's opposition was less successful in turning base runners into runs
• The club won more games than expected from its Runs Scored and Runs Allowed

These five items are demonstrated statistically in the article. At the end of the article a list of the fifteen luckiest and unluckiest teams from 1960-2001 is given. Below are selected luckiest and unluckiest clubs from that list, with teams from both leagues before 1969 and from 1979-93 in the AL and 1993 in the NL (since balanced schedules were used in those league seasons). The club's actual record listed first and its luck-adjusted record second; the difference in wins is in parentheses.

1960 PIT: 95-59; 76-78 (-19)
1992 OAK: 96-66; 77-85 (-19)
1962 LAN: 102-63; 84-81 (-18)
1961 NYA: 109-53; 91-71 (-18)
1993 SFN: 103-59; 85-77 (-18)
1963 NYA: 104-57; 87-74 (-17)

1962 NYN: 40-120; 61-99 (+21)
1979 OAK: 54-108; 74-88 (+20)
1987 CLE: 61-101; 81-81 (+20)
1961 PHI: 47-107; 66-88 (+19)
1962 CHN: 59-103; 77-85 (+18)
1964: KC1: 57-105; 75-87 (+18)
1960 DET: 71-83; 87-67 (+16)
1985 CLE: 60-102; 76-86 (+16)

A difference of sixteen to twenty wins from luck is substantial.

It's only half now because the NHL has expanded over the years. When it originally increased its playoff field to sixteen, there were only twenty-one teams in the league. That's 76% of the league.

In the minors, some leagues, for a time, differentiated between the first-place club and the playoff victor: the team finishing first was the league champion while the playoff winning club was the playoff champion. But as the years went on this differentiation evaporated and the league champion is now the playoff champion.

I don't agree with a lot of the criteria for luck. However yes, a full season schedule still won't guarantee the best team wins. That is no defense for a playoff system that increases the chances of an inferior team winning.

[daves]

Originally Posted by Brad K

Win a division and get a single game playoff against the second-place team while the other division winners get a bye?!?!?!? Don't think that can be sold.

Actually both the 5 and 7 team playoffs are impractical for a 4 division league. Not picking on you here but saying what to do without considering how is rampant in this thread. And as a belated response to the slightly veiled old school criticism, I'll call it a new school sort of thing.

Point taken. I would have preferred 4 but that was not an option on the poll.

[Brad K]

Originally Posted by daves

Point taken. I would have preferred 4 but that was not an option on the poll.

The poll choices were not well selected. Due to the problems describe, 5, 6, and 7 shouldn't have made it to the poll.

[Le Grande Orange]

Originally Posted by Brad K

I don't agree with a lot of the criteria for luck. However yes, a full season schedule still won't guarantee the best team wins. That is no defense for a playoff system that increases the chances of an inferior team winning.

Except that the criteria for "superior" and "inferior" typically consists of comparing the clubs' regular season winning percentage. Which has already been shown to be a less-than-perfect measure.

As for the article referenced previously, you can read it for yourself. This page of the SABR web site contains all the issues of the Baseball Research Journal that are currently available, free of charge except for the most recent issue (Vol. 47), for download or viewing online.

Here is an article by Pete Palmer examining skill and luck in MLB that was published in BRJ #46 (Spring 2017). From the conclusion: "Most people think luck is a lot less important than it is. A team’s record from year to year includes a great deal of luck, and luck contributes about equally as skill to a team’s eventual regular season record. (And in the postseason, it’s nearly all luck.)"

[NoOne]

Originally Posted by Brad K

Meaningful in the sense it is needed to break a tie. In contrast to a playoff game between two wild cards who finished second their division which is totally concocted and not needed at all except for being a patch for inept expansion, division set up, or scheduling.

I agree neither is meaningful in a mathematical sense as far as determining the better team.

i think i agree? lol

if you look at it purely from a logical standpoint, playoffs in general are a silly concept and purely for the spectator's enjoyment -- whether wildcard or even just a 2 team playoff. it settles very little definitively compared to a larger sample regular season.

there's a site that attempts to guesstimate the chances of best team winning a WS.. it's not high.

same with divisions. they exist for spectators. they like them. rivalries blah blah blah. it doesn't make sense from a competition standpoint.. it's purposefully making the outcome more volatile and less certain.

in RL if this stuff helps the ratings, then i'm all completely indifferent about it. that benefit to the future of the game may outweight any minor % reduction that the best don't win. and, i don't think they really give 2 cents who wins in a general sense as far as how it is setup ... (either that or someone is bad at math in MLB offices)

based on how we treat the losing team it seems a bit masochistic, lol. (see: 90's bills). we make it near(er to) impossible for the best team to come out on top through the system that is used, and then fans call them chokers, or worse, when they don't succeed against the odds.

LAD didn't fail or choke. they likely only had a 25% chance or less to start. just getting to the WS is a win and against the odds. (better example when they won 114 games or whtever it was)... lol seattle! i forgot about their colossal "failure", which in reality is one of the greatest seasons ever... only a pinhead would reduce it just because they didn't win a WS that year.

[Le Grande Orange]

Originally Posted by NoOne

if you look at it purely from a logical standpoint, playoffs in general are a silly concept and purely for the spectator's enjoyment -- whether wildcard or even just a 2 team playoff. it settles very little definitively compared to a larger sample regular season.

Except that the 'larger sample regular season' isn't as definitive as you appear to think it is, as the articles by Birnbaum and Palmer indicate. Luck plays a considerable role in the regular season records of clubs.

[Brad K]

Originally Posted by Le Grande Orange

Except that the 'larger sample regular season' isn't as definitive as you appear to think it is, as the articles by Birnbaum and Palmer indicate. Luck plays a considerable role in the regular season records of clubs.

I don't know. You've listed 16 team seasons that are way off out of a sample of what, over a thousand team seasons?

[Brad K]

Originally Posted by Le Grande Orange

Except that the 'larger sample regular season' isn't as definitive as you appear to think it is, as the articles by Birnbaum and Palmer indicate. Luck plays a considerable role in the regular season records of clubs.

What is the point? Perhaps you would like to cancel the playoffs - even a well-designed system including only division winners who played in divisions where no team had a schedule advantage - and substitute that with a luck analysis at the end of the season to determine who really should have had the best record?

But then you have the issue of is a difference of one or two luck adjusted wins is significant enough to crown a champion.

[NoOne]

you can have a bye week for one team using "7". (3+1 works the following week.) i use that # currently for a 30 team league (no sub leagues).

"5" would require 3 bye week teams and not likely practical for anyone.

"6" with 2 bye week teams might make sense for some? (2+2 following week works)

bye weeks are not a problem for rust. assuming they didn't just come off dl with no games to knock rust off. Active roster maintains, playing knocks rust off, being off a active roster accrues rust. since they aren't rusty and not off any active roster == no problems.

[NoOne]

Originally Posted by Le Grande Orange

Except that the 'larger sample regular season' isn't as definitive as you appear to think it is, as the articles by Birnbaum and Palmer indicate. Luck plays a considerable role in the regular season records of clubs.

i didn't say they were absolutely perfect... i did say the conclusions that can be made are absolutely far more certain than whatever a few playoff series tell you.

also, adding layers of playoffs only reduces the chances of the 'best' team winning... so, what exactly is the goal if not the crowning the best team as the champion?

e.g. take 3 playoff series and a % chance of winning of some amzingly awesome team just to show a high end of the spectrum: like 80%, 70%, 60% in championship. .8*.7*.6= .336.. it's probably slightly higher because sometimes they play weaker teams? but that may be balanced by upsets of their own? meh call it a ballpark guesstimate at best. i bet the best team wins in current mlb system about 1/5 at best and no worse than 1/10.

regardless of that, i also said i don't mind playoffs and having more teams makes more money and makes the sport better in the future -- best guess is that positive is more important than crowning a somewhat arbitrary champion.

the mob likes more playoff teams, so the mob gets more playoff teams. this is the same group of people that burned women if they knew math and thought the world was flat, lol... just with fancier electronic toys. popularity may have the most benefits, but it doesn't in any way mean it is logical, rational, or reasonable.

[Brad K]

Originally Posted by NoOne

regardless of that, i also said i don't mind playoffs and having more teams makes more money and makes the sport better in the future -- best guess is that positive is more important than crowning a somewhat arbitrary champion.

Don't know that there's a correlation between more money and a better sport. And if there is, now long will it last?

If industry management creates a system where clearly undeserving teams can often win, then I see interest declining because the value of being champion is less. OTOH, if, although its known the system is imperfect, its the best that can be arranged, then being champion is tarnished by intentional inclusion of impostors to the crown.

[Brad K]

The purpose of having divisions is not to have an excuse to put more teams in the playoffs. The purpose is to provide a balanced scheduling situation when a league has grown to large for a balanced schedule involving all teams.

So, properly handing this means...

1. Keeping divisions large enough so that the chance of strong or weak divisions are minimized.

2. Keeping all the games a team plays within its own division or if not, the majority in and all division teams facing the same opponents in balance outside the division.

3. Number of divisions must be a power of 2 so there are no wild cards. ( I suppose it would be possible to design a reasonable multi round play all elimination tournament, but I haven't looked into it so am not proposing it.)

[Fyrestorm3]

I really have but one opinion here: ultimately, sports are entertainment, and therefore are supposed to be fun.

Playoffs mean elimination games. Wildcards mean more teams in the playoff chase. One-game play-ins mean drama. All of which equals more excitement, and therefore more entertainment.

Getting too caught up in always making sure the "best" team wins ruins the fun. Hell, I hate the manager's challenge, not out of traditionalism or anything, but because it was simply more fun to get pissed off at the umps than to ensure every play is called perfectly. No, that's not logical, but it's entertaining, and that's what I watch for.

[Brad K]

Hey, can't argue with "I like it because I like it".

[Brad K]

Despite this discussion continuing, I expect the OP made a decision on playoff configuration long ago. Curious what it was. (Guessing one of the 8 per league configurations.

[professor ape]

Originally Posted by Brad K

I don't agree with a lot of the criteria for luck. However yes, a full season schedule still won't guarantee the best team wins. That is no defense for a playoff system that increases the chances of an inferior team winning.

I recently did a mini-study about luck in terms of actual wins against their Pythagorean wins. What was interesting to me was how the luck observed in the league is much lower than what you would expect through true random chance.

Based on random luck you would expect to see 52% of teams within four wins of their Pythagorean wins. Over the last ten years we have seen 72% fall in that range. Over the same period you would expect to see 30 teams with a variance of more than ten wins off of their Pythagorean wins. There have only been five.

Overall, through random chance the average team would have a variance of 5.07 between their actual wins and Pythagorean wins. Over the last ten years the average “luck” has been 3.21.

[Brad K]

Originally Posted by professor ape

I recently did a mini-study about luck in terms of actual wins against their Pythagorean wins. What was interesting to me was how the luck observed in the league is much lower than what you would expect through true random chance.

Based on random luck you would expect to see 52% of teams within four wins of their Pythagorean wins. Over the last ten years we have seen 72% fall in that range. Over the same period you would expect to see 30 teams with a variance of more than ten wins off of their Pythagorean wins. There have only been five.

Overall, through random chance the average team would have a variance of 5.07 between their actual wins and Pythagorean wins. Over the last ten years the average “luck” has been 3.21.

That's interesting. I previously considered the possibility individual players real life performance varied less in real life than in OOTP and another baseball game I previously played (No data, just an impression.)