Why Playoffs are a Crap Shoot

Of course the playoffs are a crap shoot in Strat, just like real life, if not my Cardinals wouldn't be where they've been for years. I think one of the things that gets us 'stat geeks' is that the playoffs matter in the sense of what some call 'clutch' I think that all of us that even played high school, college or even little league baseball know what clutch means and those that can ramp it up in the playoffs are the true champs. Didn't we all watch Madison Bumgarner last year chuck those fastballs in that his opponents just watched or fanned at? Do any of us think the Clayton Kershaw isn't the better pitcher overall? But why does Bumgarner rock the playoffs? It's that other level. Unfortunately there is no way to account for it statistically, so we are all at the mercy of math in Strat, but we can pretend........

Of course it isn't a crap shoot. A crap shoot denotes the entire absence of contributing factors and presence of complete randomness. There are many such contributing factors in SOM preventing such randomness, factors reflecting each managers' choices and abilities:

1. The chosen hitters cards
2. The chosen pitchers cards
3. The managers' choice of lineups
4. The managers' selected pitchers preferences.
5. The managers' selected hitting preferences
6. The managres' selected strategy settings
7. The managers' selected pitching rotations

These are all significant contributing factors preventing the SOM playoffs from being an entirely random "crap shoot" as you say. If you would like to show how none of these factors have any impact on SOM playoff results, knock yourself out.

As to the real playoffs, the same applies, and moreso. Since the MLB playoffs are dealing with human behavior, such elements as performing well under pressure, or not doing so, definitely factor in. Also, your choice of Bumgarner as some anomalous proof of playoff randomness is odd. Even before those playoffs, Bumgarner had proven himself to be both an excellent pitcher and an immensely gutty performer. So, his playoff performance was just a remarkable affirmation of his prior regular season performance. it was hardly proof the playoffs inherently produce and engender anomalous performances.


P.s. Even your admittance there are certain "clutch" playoff performers who can "ramp it up in the playoffs" denies the playoffs are a "crap shoot." The presence of those players themselves negates the randomness required for a "crap shoot."

You can prepare for the playoffs by manipulating your starters or keep players from being injured, but in the end it is a crap shoot.

Agreed - but it sure is nice to win with an underdog team.

You can prepare for the playoffs by manipulating your starters or keep players from being injured, but in the end it is a crap shoot.

Agreed - but it sure is nice to win with an underdog team.

If it's just a "crapshoot," there is no such thing as an underdog team, since any team would have as much chance as the other. For the odds of that, refer to my last post.

Apart from managerial settings and strategies, some teams are better set-up because their 4-5 pitching slots are weak and you only have to use one 4 pitching start during a 7 game series. Some teams also match-up with other teams better than their records would indicate. Injuries can play a greater role in a short series than over an entire season. What seems unfair, though, is that random chance gets (well should) get canceled out over 162 games but won't be in a short series. So there are factors pertaining to differential results in play-offs pertaining to randomness but also to the fact that teams have a different competitive strength in a play-off as opposed to the regular season. At some point, if a team is strong enough, you would think it likely that they could overcome random chance and if they are losing it's because their play-off advantage is not as great as their regular season record would indicate.

Apart from managerial settings and strategies, some teams are better set-up because their 4-5 pitching slots are weak and you only have to use one 4 pitching start during a 7 game series. Some teams also match-up with other teams better than their records would indicate. Injuries can play a greater role in a short series than over an entire season.

Firstly, at least you acknowledge managerial settings and strategies, as well as team set-ups, are a factor in the playoffs. So, you acknowledge it isn't just a "crap-shoot." So, on that we definitely agree.

What seems unfair, though, is that random chance gets (well should) get canceled out over 162 games but won't be in a short series.

This, however, isn't true. Random chance is never cancelled out: not in the regular season, and not in the playoffs. This is a game based on simulated dice rolls with (theoretically) the same possible results with each roll. So, while random chance can have much greater effect in the much-shorter playoffs than in the regular season, it is equally proportionately present in both.

So there are factors pertaining to differential results in play-offs pertaining to randomness but also to the fact that teams have a different competitive strength in a play-off as opposed to the regular season. At some point, if a team is strong enough, you would think it likely that they could overcome random chance and if they are losing it's because their play-off advantage is not as great as their regular season record would indicate.

As to this passage, I again agree with you that both randomness and differences in competitive strength are in play. However, i would acknowledge that randomness is enough of a factor for an inferior team to beat a superior one if it rears its ugly head strong enough. So, while a "superior" team's losing could suggest it wasn't as competitive as its regular season suggested, it could also be a significant result of luck and randomness "over-exerting" itself...which it certainly doesn't do in every series.

I am not sure we are using the same definition of randomness. If I flip a coin a few times the number of heads and tails will oscillate back and forth but if I do it a thousand times the number of heads and tails is going to be very close to 50 percent each. If I roll on a player's card 30-40 times the results could oscillate substantially from his predicted value but over 300-400 times it should get a lot closer. Saying canceled out was perhaps inaccurate, but over 162 games the effect of randomness (defined as being departure from expected value) should be much less in a 162 game season as opposed to a 7 game season.

I am not sure we are using the same definition of randomness. If I flip a coin a few times the number of heads and tails will oscillate back and forth but if I do it a thousand times the number of heads and tails is going to be very close to 50 percent each. If I roll on a player's card 30-40 times the results could oscillate substantially from his predicted value but over 300-400 times it should get a lot closer. .

I, myself,am using the official definition of "randomness," which denotes "without definite aim, direction, rule, or method." That applies to the dice rolls; so, they are random throughout the season. In every single game, that randomness has a chance to produce an unlikely result (or results) that could significantly affect the outcome of the game, if not cement it. That possibility occurs within every single game of the season, and the results in previous games do not alter that possibility in future games. So, your coin-toss anecdote doesn't apply. And the 162-game season doesn't in any way cancel out randomness, it just suffers less from its effects.

Saying canceled out was perhaps inaccurate, but over 162 games the effect of randomness (defined as being departure from expected value) should be much less in a 162 game season as opposed to a 7 game season

Well, we're in complete agreement here. i even said in my last post the effect of randomness would be much less severe than in the playoffs. And we can definitely agree it isn't "cancelled out" during that regular season.

Ok now that we have your "official " definition of randomness perhaps we can get somewhere. The outcomes are determined by simulating (we'll assume accurately) the roll of 1-6 on a die and then the roll of two dice. The outcome of each roll is unknown before-hand but over a large number of rolls the outcomes will follow the patterns set by the probability of dice rolls. Exactly the same as what happens with flipping a coin, or more apposite, the throwing of dice. What happens is that over a small number of rolls there can be a large variance from expected values but with a large number of rolls the variance from expected (looking at all of the rolls) tends towards zero (like an infinite series if you recall your calculus). So if you have a hr on a batters card in a certain column with a two-sided roll of 7 there is 2.7% chance of that occurring. In a 7 game series, though, there could be a large variance between the expected 2.7% result and how many times that came up in the series. The fact that a prior game result does not affect the next game result does not change probabilities; each time two dice are thrown they are independent of prior dice rolls but if we throw the dice a million times we can very confident that the number of sevens is going to be 1 out of 6. Thus, over a 162 game season there is almost certainly going to be a lot less variance. And that is why a 162 season is more reflective of a team's performance because eventually the dice rolls will tend to their probability distributions. But if you want to believe that chance/randomness has the same proportion in a 162 season as in a 7 game play-off go right ahead--I am not stopping you. This gives a more visual demonstration. https://www.fourmilab.ch/rpkp/experimen ... stics.html

The more coin flips you have the skinnier the curve gets. In our terms, the less the curve goes to the left or right the less chance involved.

My definition is official; there's no need for scare quotes. As to the rest of your post, I'll be succinct:

1. The dice rolls in SOM do not entail the repetitiveness of your previous coin flip analogy, so that analogy does not apply. In SOM, teams will face different starting pitchers, different lineups, and/or different pitchers. So, there is no "evening-out" from repetition. There is only a tempering of randomness' effects from more games not so affected.

2. As I said in my last post, the rolls in previous games will have no effect on subsequent games. So, again, there is no necessary evening-out of results, particularly with shifts in line-ups, pitchers, and teams.

3. We definitely agree neither the regular season nor the playoffs are "crapshoots."

Very interesting article Freeman. Unfortunately, you're trying to use facts to sway a person who just likes to argue for arguments sake.