Playoff Odds

[Frayed Knot]

As simple as I can make it, there are six different ways the Mets can play out the schedule [0-5, 1-4, 2-3, 3-2, 4-1, 5-0] and seven different ways each the Giants and Cardinals can play out theirs given that they have six games remaining rather than five.
So 6 x 7 x 7 = 294 different combinations that things could still turn out (ignoring the Marlins for the moment).

The way I figure it, 227 of those have us winning one or the other WC outright. Of the remaining there are 30 more scenarios where we'll need a play-in game because we either tied for the second slot or wind up in a three-way knot. And then there are 37 outcomes where we wind up on the outside looking in.
Problem is, not all of those outcomes are equally possible. It's far more likely, for instance, that the Mets go 3-2 or 2-3 than it is 1-4 or 4-1, and even those outcomes are more likely still than 0-5, 5-0. Those same 'squishing towards the middle' odds apply also to SF & StL. The odds therefore of each team running the table for instance is not 1/294 (0.34%) but a rather ridiculous 0.00076%
These odds are all generic, btw, and don't take into consideration the quality of opponent and pitchers, or the hot-ness or not-ness of each team, but it's the best we can do here.


I assume that places like BP take all those factors in and so it's probably just best to follow their daily updates.
At the moment they've got us at 83.3% in although 'IN' might still entail a 50/50 shot via a play-in game. Either way, with so few games remaining the odds have the ability to rise or drop fairly dramatically with each game played and the volume of possibilities will drop quickly.

Bottom line: The odds are pretty good as things stand right now that Sunday will not be our final game of the season ... but stayed tuned for further updates as they become available.

[Centerfield]

Great stuff FK.

It sounds very encouraging. With St. Louis losing yesterday, it takes a lot of the pressure off.

[Benjamin Grimm]

Magic Numbers: Mets over Cardinals: 5. Mets over Giants: 6.

Tragic Numbers: Giants over Mets: 7. Cardinals over Mets: 8.

Soonest that Mets can clinch a wild card: Thursday. The could potentially have three post-clinching games to setup their rotation and/or try to secure home field advantage for the October 5 playoff game.

Soonest that Mets can be eliminated: Friday. Their MWOM (Minimum Window of Meaningfulness) now extends through Friday, September 30. At most, they'll play two meaningless games.

[Frayed Knot]

Magic Numbers: Mets over Cardinals: 5. Mets over Giants: 6.

And the good thing about those magic numbers is that we only need Either of them come through.

It reminds me a bit of 1999. As we all recall (and if you don't you'll be banned) it was the Mets sweeping the Pirates AND the Reds stumbling twice against a bad Brewers team over the final weekend which forced that WC play-in game which in turn led to the Pratt win over Arizona which in turn led to the GS Single and the almost win over the Braves. 'We got lucky' and 'backed in' some claimed on account of the fact that we couldn't have earned it on our own and needed the unlikely result of the Cincy/Milw series.
Except what is often NOT so well remembered is that Cincy started that final weekend locked in a dead heat with Houston for the NL Central crown which means we merely needed EITHER team to stumble in their final series whilst we were taking care of our sweep. The Astros wound up winning twice vs the Dodgers (lost the first but then came back in games two & three) but the mere presence of having two teams in the mix doubled our odds for that year's 'miracle' post-season spot.

And now this year looks to play out in similar fashion: in running from the bear that is a post-season-less October, we don't have to out-run both SF & StL we just have to out-run one of them and let the bear eat the other.
I mean, sure we'd prefer beating both, but the difference between finishing ahead of neither and finishing ahead of one is a whole lot larger than the difference between one and two.

[Centerfield]

The cool part about that 1999 race was that if Cincy and Houston tied, we would have automatically gotten the Wild Card, no play in game.

Cincy and Houston would have squared off for the Division title. That being a regular season game, the loser would have finished a half game behind the Mets.

Definitely an odd quirk. And pretty unfair.

[TransMonk]

Frayed Knot wrote: in running from the bear that is a post-season-less October, we don't have to out-run both SF & StL we just have to out-run one of them and let the bear eat the other.

I like this analogy.

[Benjamin Grimm]

Given that the Cubs are looming as the opponent for whoever wins next Wednesday's game, even the team that outlasts the other two will still have ursine concerns.

[Centerfield]

BP has us up to 86.1% now.

St. Louis down to 42.9.

86 seems high for being just 1up in the loss column but then I remember even odds here are 66%.

[Frayed Knot]

BP has us up to 86.1% now. St. Louis down to 42.9.

And Giants at 71%

86 seems high for being just 1up in the loss column but then I remember even odds here are 66%.

Yeah, remember that these odds have to add up to 200% since two teams are guaranteed entry, and in this case (unlike in the AL) it's only split three ways with the Marlins going back down to 0.00% even though they still have a hail mary shot at forcing a play-in game.


Mainly the best thing that happened to us yesterday even though all three contenders won (and all three scored exactly 12 runs!!) is the removal of a day off of the schedule.
The total number of permutations for the three teams over the remaining week fell from the 294 prior to last night's action down to 180 [5 combos for us x 6 each for SF & StL] and only 20 of those 180 eliminate us.
We get in clear via 140, and at least tie for a spot in the other 20. Now again, like yesterday, these are unweighted odds so it's not exact but both my figuring yesterday and again today pretty much mimic the BP stuff: 140/180 = 77% then throw in around half of the 20% play-in odds and you're right there in the mid/upper-80s just like BP


Remember also that the Cards have to be at least two wins better than us and now have less time in which to do it (even with one extra game) and the Giants have to notch at least one more win - and if Either/or of those don't happen then we're in.

So plainly put:
- sweeping our four games obviously gets us in since we can't be caught
- but three wins virtually guarantees it also. There are 36 combos of us going 3-1 and we're in free and clear on 34 of them and in some sort of tie via 2 [a two tie for #2 slot w/StL if both SF & StL run the table; a three-way tie if Cards go 5-0 and SF 4-1]
- going 2-3 gives us 30 'IN', 4 'TIE', 2 'OUT'
- a paltry 1-4 still yields 24 'IN', 6 'TIE', 6 'OUT'
- and even 0-5 let's us sneak in almost half the time - 16 'IN', 8 'TIE', 12 'OUT'

[Centerfield]

I'm going to guess that our odds are pretty good now.

[Benjamin Grimm]

I'm now thinking more about clinching by Saturday than I am about clinching in general. I very much want the Mets to be in a position to start Ynoa or Montero on Sunday instead of Syndergaard.

[Edgy MD]

NYM 85.9%
SFG 69.4%
STL 44.7%

[Benjamin Grimm]

Magic number over the Giants is 3, over the Cardinals is 2.

[Centerfield]

Edgy MD wrote: NYM 85.9% SFG 69.4% STL 44.7%

I don't think the numbers are updated. They couldn't have gone down from yesterday.

[Frayed Knot]

NYM 85.9% SFG 69.4% STL 44.7%

Yeah, something's screwy there because those numbers are barely changed from the night before and Wednesday's results should have produced a huge shift with so few games remaining.



FWIW via the process of just seeing how many combinations put us in or out, the math is pretty simple as there are now exactly 100 possibilities and we're in on 92, out in just two, and in some sort of tie in the other six.
Following some intense calculations I estimated those odds at right around 92%, 2%, and 6%

- With a sweep in Philly we're not only in but in via the #1 slot

- With two wins in Philly it's the same deal because that would eliminate the Cards and we'd hold the tie-breaker w/the Giants even if they ran the table

- We can now even go 1-2 in Philly and still be guaranteed of playing past Sunday. We'd be in via 23 combinations of Cardinal & Giant outcomes but forced into a tie in two others:
1) Cards win their final 4 while Giants go 3-1 which means three-way tie
2) Cards win 4 and Giants win 4 in which case we'd tie the Cards and play-in vs them for the right to play the Giants

- If we take an oh-fer in Philly then we still get in free and clear via 19 of the 25 combos of SF/StL games.
There are the only two scenarios where we're out: 1) Cards win four, Giants win three; 2) Cards win four, Giants win four
and there are four combos which will throw us into a tie of one sort or another:
1) C = 3 Wins, G = 2 (3-Way T)
2) C = 4, G = 2 (Tie w/G for 2nd)
3) C = 3, G = 3 (Tie w/C for 2nd)
4) C = 3, G = 4 (Tie w/C for 2nd)

[Edgy MD]

Good work. Sorry to take the BP report at face value.

[Benjamin Grimm]

I just checked to see if the odds at Baseball Prospectus had been refreshed, and got a white page with a one-line text message:

Playoff Odds will return shortly - doing brief maintenance.

[Frayed Knot]

We should be at around 95% at this point, and if BP is giving any kind of weight to the idea that the Mets are closing out vs the weak Phillies while the Giants and Cards have the relatively stronger Dodgers & Pirates (as opposed to my simpler method which essentially treats every outcome as a 50/50 proposition) then our odds will likely be a few clicks higher still.



btw, both games of interest today are 6 PM local starts -- 7:15 EDT first pitch in StL; 9:10 EDT first pitch in SF -- so we've got a good chance of knowing the results prior to shut-eye tonight.

[Lefty Specialist]

Plus, the Phils will probably be pitching a 'bullpen game' on Friday.

[Frayed Knot]

OK, here's what I got:

IN = 95.06%
OUT = 1.95%
3-Way Tie = 1.76%
2-Way Tie for 2nd WC w/Giants = 0.29%
2-Way Tie for 2nd WC w/Cards = 0.94%


That may not be the only answer, but it's my answer and I'm stickin' to it!!

[Benjamin Grimm]

They're good numbers!

Hopefully we'll get one or two happy outcomes in tonight's games. (If I had my choice of one or the other, I'd choose a loss by the Cardinals.)

A Cardinals loss would eliminate the possibility of the Cardinals finishing ahead of the Mets.

[Centerfield]

Lefty Specialist wrote: Plus, the Phils will probably be pitching a 'bullpen game' on Friday.

Why do you say that? I thought they had that guy Asher lined up to pitch. He was pretty good in the early on against the Mets last week.

I think it would be ideal if the Cards lost the next two games, allowing the Mets to clinch, then won the last two, while the Giants lose all four.

Let's leave Madison Bumgarner at home.

And I realize this is bad karma, and that I'm inviting Carlos Martinez to throw a 2 hit shutout, but I've thought about it, and I want no part of MadBum.

[TransMonk]

538.com has the Mets' odds at 99%. I'm very comfortable saying it is likely that the Mets will get into the WC play-in game (99% scares the side of me that knows that anything is possible).

Everything went right last night.

[Benjamin Grimm]

Baseball Prospectus is back online. They have the Mets at 98.2% for a wild card, 48.2% to advance to the NLDS, and 2.0% to win the World Series.

Cardinals are 33.2%/16.5%/0.7% and Giants are 68.5%/34.6%/1.7%

[Frayed Knot]

Frayed Knot wrote: OK, here's what I got: OUT = 1.95%

BP's updated numbers: Mets = 98.2 (aka: OUT = 1.8%) ... and my numbers were two hours earlier and on my first try.


BP: Giants = 68.5%
Cards = 33.2%

[Edgy MD]

So, shorthand ... can the Mets clinch a Wild Card berth tonight?

[Benjamin Grimm]

No, they can at best clinch a tie if the Cardinals lose.

The earliest they can fully clinch a wild card is tomorrow.

[Edgy MD]

Team's don't celebrate "At-Least-a-Tie Night," right? Only "Outright-Sole-Possession-Guaranteed Night," right?

I just don't want to end up confused on the spot like Steve Zabriskie.

[TransMonk]

I'm rooting for a Cardinal loss tonight and a Mets win against the Phillies' bullpen outing tomorrow. That would provide a clinch of playing a game on Wednesday.

I'm pretty sure ties are not celebrated. We were discussing in some thread last week, but I'm of the opinion that champagne corks should only be popped with a win on Wednesday.

[Benjamin Grimm]

That's the ideal outcome: A St. Louis loss tonight and a Mets win tomorrow, allowing the Mets to clinch the wild card at the conclusion of their Friday night game.

But regardless of what the Giants and Cardinals do, if the Mets win the first two games in Philadelphia they'll have clinched the home field for next Wednesday's game. The two wins would guarantee they have a better record than the Cardinals and would mean that at best, the Giants could tie the Mets for the first slot. For home-field purposes, a tie between the Mets and Giants means a game a Citi Field.