NFL 2015 Season Week 15 Picks

https://www.reddit.com/r/minnesotavikings/comments/3wuct0/this_would_be_a_dream_come_true/cxziffy

Where in I argue with a guy who thinks it's actually *probable* the >99% to make the playoff Bengals will miss the playoffs. This is the internet after all, where even basic reasoning problems can be debated as opinion.

http://espn.go.com/nfl/playoffs/machine/_/factor/powerranking
So here is how things end up if winners are based on the current power rankings: Bengals get the number 2 seed.

If everything plays out the same and the Bengals lose out: Bengals still win the division
http://espn.go.com/nfl/playoffs/machine/_/factor/powerranking/results/400791578~1~400791726~1~400791617~2

If Pittsburg wins out and Bengals lose out: Bengals are the 6 seed:
http://espn.go.com/nfl/playoffs/machine/_/factor/powerranking/results/400791578~1~400791726~1~400791617~2~400791723~1

If Bengals lose out, Pittsburg wins out (which is only one more win than the power rankings matchup would have them winning) and the Jets win one more game than predicted): Bengals miss the playoffs

In other words, There is a somewhat "reasonable" scenario that has the Bengals missing the playoffs. It would require the Bengals to lose out (which is possible), the Steelers to win out (which is only 1 game different from the power rankings matchup would predict) and the Jets to win the 2 games they are predicted to win plus beating New England at home (low probability, but not impossible). So while there are 100 scenarios involving all the games that have to go a certain way, the predictions based on Power Rankings suggests a somewhat simple variation that would have the Bengals missing the playoffs. But even this variation requires that every point be hit right now. If the Bengals win one of their games, they are in. If the Steelers lose one of their games, the Bengals are in. If the Jets lose one of their games, the Bengals are in. By the strict definition of "probable" I would say that no, it isn't probable. Possible outside of what someone might normal think of as being outside of 99% certainty of not happening? I would say yes.

(for the record, I'm not debating you. Simply playing with the Playoff Machine to pretend I'm proving your wrong when in reality I'm just finding the 2% option and pretending it's a complete refutation of your argument, when in fact it is actually the very argument you are making since you argument isn't that the Bengals are definitely in, but rather that the 2% probability option is somewhat reasonable based on my arbitrary definition of what is and isn't reasonable (and that it didn't take much playing around to actually find the scenario that matches the 2% probability that results with the Bengals missing the playoffs)).

Kansas City also has to win out.

It's just one of those things where yes, on a game by game basis it's entirely plausible, maybe even all tilted against the Bengals, save for a Jets upset over NE. On the surface it looks like a perfectly plausible scenario. The argument isn't really any sort of "it would take 5 monster upsets to make it happen". The reality is though that the sheer number of things the Bengals need to happen though to miss is the matter at hand.

If every game had a 8 in 10 chance of going against the Bengals there's still only a 7% chance they all do. It barely matters how likely each game is (and I agree, almost to a game it doesn't look good for the Bengals) because it's about the whole. Even assuming 8 in 10 AND that the Bengals never win again this season the odds only go up to something like 13%.

That's right, but according to the Playoff Machine's Power Rankings predictor, the Chiefs will win those games. So obviously it's a done deal. But the more I look at it, the more I see how many things have to go exactly the a certain way for the Bengals to miss. No matter what the Bengals do, if the Chiefs, Jets, and Steelers end the final 3 weeks with a combined total of at least 1 loss then the Bengals are in. I would bet a pretty large chunk of money that the Bengals make the playoffs. I get the feeling that the guy you were arguing with would give me pretty good odds, too!

To expand on the example I gave in reddit. Let's say I hand you a dice, and tell you "bet you $100 you roll a 1", it should matter greatly to you how many times I want you do roll that dice.

The odds are stacked against me on each distinct roll, and no roll effects the next, and so on, but the odds of you *never* rolling a 1 go down down down with every roll I add to our bet.

It only takes 4 rolls where something overwhelmingly stacked in YOUR favor individually actually tilts MY way as a whole. At 12 rolls (which is what we're talking about with Bengals games) I win that bet 89% of the time.

12 rolls being the number of games that matter to the Bengals making the playoffs? (3 each by the bengals, chiefs, steelers, and jets)?

That is something. To expand on my equation:

If (Bengal_Wins) + (chiefs_losses) + (steelers_losses) + (jets_losses) >= 1 then Bengals are in.

We're not flipping a fair coin in the case of NFL games, but I'd say it's a pretty safe bet that you flip Heads at least once in those 12 flips.

Right, and we'll never know the "real" odds of course, but that fact is sort of irrelevant. Because you can be extra generous about "loading that coin" against the Bengals, and demonstrate that they still make the playoffs almost always, so the "real" odds that the Jets beat the whoever is kind of moot.

The Bad-for-the-Bengals outcome would have to be something around 95% likely in all 12 games before it becomes "likely" that they miss the playoffs.

Wow. Even if every game went against the Bengals at 90% probability it would still only be a 28% probability that they miss the playoffs. If one of those games falls to .6 (like, say the Jets beating the Patriots), it falls to 18%. If 3 of the games are .6 and the rest at 90%, the odds fall to 8%.

And even assuming the Bengals lose out doesn't move the needle as much as people think.

Also the fact that vbullinger put the Patriots-Jets game at 30% is funny since if that game went the way he predicted then the whole thing is moot. So without realizing it, even he is at best only 30% confident that the Bengals will miss the playoffs. For from "probable".

You should just respond one last time that if ANY of the 3 teams go 2-1 then the Bengals are in, end of story.

He states later he "never claimed" everyone had win out only that they had to go 2-1 and maybe tiebreakers would step in and oust the Bengals.

As such I guess he implied I was straw-manning him by using what actually has to happen for the Bengals to miss out in a conversation of how likely it is.

So in the end he spent a lot of words on why I have no idea what I'm talking about when thinking it isn't probable, without even actually looking up the state of the situation.

For all he knew it would have taken 6 ties for the Bengals to miss. In the end he, literally, just didn't even know what he was talking about.


Edit: but that's an interesting pick up on the 30% thing none the less. Might have been a good thing to point out, but as it turned out he was talking out of his ass, and really didn't look into the actual scenarios the teams are in (and evidently just didn't believe me or something) beyond "Dalton is out so they're done for".

That's just your opinion.

I saw a link to this article earlier in the week. The headline could almost qualify as clickbait. http://www.cbssports.com/nfl/eye-on-football/25414477/heres-how-the-10-3-division-leading-bengals-could-still-miss-the-playoffs

Yeah, I looked to see if anyone in the Bengals subreddit was concerned about it and I saw someone posted that article and a bunch of people were calling it click bait.

Bengals are in. It took all of one week to junk the "likely" scenario of 12 games all going in the exact the wrong way against the Bengals that would have caused them to miss the playoffs.

And they did it with their own win,which is funny, as a lot of the breakdowns were of the "even if you assume they lose out" variety.

The Packers, by my estimation, have a 16.8% chance of getting the 2 seed.

Packers @ Cardinals - 40% confident in Packers win
Vikings @ Packers - 70% confident in Packers win
Seahawks@cardinals - 60% confident in Seahawks win

Multiply that all up, it's a 16.8% probability. In other words, highly likely.

Also, I can't figure out how the Vikings haven't clinched a playoff spot. If they lose their final 2 games, who is still able to overtake them?

The only option I can see as a possibility is Atlanta, and it must be some obscure tie-breaker involving total points scored, or something else that doesn't factor into the Playoff Machine's capabilities. If the Vikings lose their final 2 games, they are 9-7. If the Falcons win their final 2 games, they are 9-7. Right now, the Playoff Machine has the Vikings in the playoffs with that scenario, but their standings don't indicate that they have clinched a spot. I want to say that the Falcons are the only team remaining that could finish 9-7.

edit: scratch that, the Vikings beat the Falcons head-to-head, so the Vikings have essentially already clinched a spot ahead of the Falcons.

edit 2: I figured it out. If Seattle and Minnesota lose out and Atlanta wins out, then they all finish 9-7, and the 3-way tie break procedure applies. Although, I'd have to have it stepped through for me to figure out how exactly it works.

Edit 2 is correct, it's all about the 3 way tie making the head to head meaningless. Although most people say the Vikings are out in that situation, I don't know if that's right. I think as far as I can tell that's what it takes to make the Vikings vulnerable, but it goes to the strength of victory/schedule tie breakers then, and I don't see how that can be fully settled with 28 some odd other games in play swinging that around hither and tither.

Here is a scenario where the Vikings make it even if the "big 6" go against them.

http://espn.go.com/nfl/playoffs/machine/_/factor/home/results/400791565~2~400791569~2~400791538~2~400791610~2

The funny thing here is that it matters who wins the Titan @ Colts game.

Bengals postmortem:

Here are all the games, and what's his faces predictions. We assumed the Bengals, at the bottom, would lose out. Lets see how he did.

Den @ Pit - 60 - Pass
Pit @ Bal - 95 - Fail
Pit @ Cle - 90 - Pass
NYJ @ DAL - 80 - Pass
NE @ NYJ - 30 - Pass
NYJ @ BUF - 55 - Fail
KC @ BAL - 90 - Pass
CLE @ KC - 70 - Pass
OAK @ KC - 65 - Pass
CIN @ SF - ? - Fail
CIN @ Den - ? - Pass
BAL @ Cin - ? - Fail

In the end the "easiest" call went their way, the longest shot went against them, but they didn't even have to sweat it out a week in realtime. Only 2 of the non Bengal games went against them, bur again, it was never about how likely each one was, only that only a single one had to, and at least one was bound to.