Monday, May 30, 2016

Stanley Cup Final prediction

The model has gone 10/14 so far. For the final, it says the Pens have a 72% chance on winning and the Sharks have a 28% chance of winning. I think the model has underestimated the Sharks chances all playoffs, and I think this series is probably closer than the model thinks it is. In any case, let's go Pens!

Friday, May 13, 2016

NHL Playoff Predictions: Conference Finals

The first round and second round of the NHL playoffs are in the books. So far, the model is doing pretty well, going 9/12 and rolling along with the Pens and Blues. Here are the predictions for the conference finals:

The model still has the Pens and Blues as favorites, and it's probably underestimating the Sharks. The Sharks are a very good team; they've just had a difficult path. They had to beat the Kings, and now they're up against the Blues. And if they get past the Blues, they may have to get past the Pens as well. Anyway, let's go Pens!

Wednesday, April 27, 2016

Stanley Cup Playoffs: Round 2

Time for round 2 predictions! Check out the round 1 predictions here. So, round 2 will actually start before round 1 concludes. Game 7 of Anaheim-Nashville will be tonight, after Game 1 of Tampa Bay-New York Islanders. So I've included predictions if Anaheim wins tonight and if Nashville wins tonight. Here they are!

If Anaheim wins:

If Nashville wins:

And of course this is not dependent on the ANA-NSH game:

Wednesday, April 13, 2016

2016 Stanley Cup Predictions

It's time for my 3rd annual NHL playoff predictions! My model and methodology are basically unchanged from last year, so check it out here. In the two years I've been doing this, I've learned that it's pretty hard to predict the outcome of any one series, but if we look at it from a stratification standpoint, we can gain some valuable information. This is, instead of saying, "team x has a 60% chance of beating team y", it may be more constructive for us to say, "teams w, x, y, and z are more likely than the other teams to make deep playoff runs and win the Stanley Cup." Picking individual winners is hard, but we can see which teams rise to the top.

When I developed this model, I ran a tenfold cross-validation. That basically tells us how good the model is at predicting the outcome of new data. The past two years, the prediction accuracy is about 0.66. So the model should be able to correctly predict the series winner about 66% of the time. That's not great, but it's much better than just using point totals. Over the course of the past two seasons, there have been 30 playoff series, and the model has predicted the winner correctly 20 times, for an accuracy rate of 67%. So the model is performing as expected. So while 67% isn't great for an individual series, the model does seem to be able to tell us which teams are more likely to win 3 or 4 playoff series. Anyway, I ran the model again for this year, and here are the results:

The main takeaway is this: STL, PIT, LA, ANA, and WSH are the teams that rise to the top. They are the ones most likely to go deep into the playoffs and win the Cup.

I was a little surprised that San Jose, Chicago, Philadelphia came out as low as they did. These are good teams, but they have brutal paths to the Finals. San Jose would have to go through LA, then likely Anaheim and St. Louis just to get to the Final. Philly would have to go through Washington and Pittsburgh.

One other thing that stands out is that the model thinks that the Pens have a 91% chance of beating the Rangers. I don't think any team has a true probability of winning that is so high. The model thinks it's so high because there are only a few observations since 2007-2008 where the difference between the teams regarding score-adjusted Fenwick % (SAF) is so huge. In every series where the difference was this extreme, the team with the better SAF won the series. The Pens are favorites in this series, but probably not 91% likely to win.

Enjoy the 1st round!

Thursday, April 7, 2016

Spring Hypotheticals: an FTC Polling Exclusive

We asked a random smattering of FTC readers*, "after how many dates is it acceptable to try and sell someone car insurance?"

"Jeez, I'd say six? Even after any kind of hanky pinky, insurance is a commitment." - Greg, 32, kickball centerfielder

"I think the answer depends on how many more dates you want. If you want no more dates, then after as many dates as it takes you to figure that out. If you want more dates, no." - Josh, 76, some kind of sheriff

"Wow, I'm sorry to hear you new job didn't work out." - Cara, 40, writer

"..." - Sydney, 33, teacher

"Just after the last one." - Dom, 36, cars

"Not until you're married, for sure." - Samantha, 29, corporate lackey

"I think there's a sweet spot. You can't do it too soon, but you can't wait too long either. I'd say somewhere between two and four. But after four i think the likelihood of even bringing it up decreases the more invested you get in the relationship." - Nilesh, 36, statistician

"I'm going to say 'none.'" - Andrew, 31, author

"Erie or State Farm?" - William, 37, retired

"If one person expressed a need for insurance on a first date and the other happened to sell it, I think that would be okay. Otherwise, I'm going to say...never." - Jane, 32, academic

"Bone first, pitch later." - Gentry, 30-ish, karaoke enthusiast

"Several years. You've gotta play long-ball." - Asa, 25, computerist

"I would say about six months of regular dating at least with someone you could see a future with. Otherwise you're just being used." - Charles, 25, office manager

"I don't think it's acceptable for someone to try to sell you car insurance unless you inquire about procuring car insurance." - Annie, 30, female

"I would say that, if you are going to attempt to sell someone car insurance, it should happen on the first date." - Marty, 34, footballs

"Did you go on a date with someone who tried to sell you car insurance?" - Nichole, 34, cat owner

"You shouldn't sell product on any early dates, and if it gets to the point that that comfort level has been reached, you should hook that person up [with a good deal]." - Michael, 36, not a Ph.D. in epidemiology or federal employee of any kind

*A totally fictional population

Thursday, March 3, 2016

We're committed to committing ourselves to the possibility of finding a way to make you purchase this baseball cap for $40

On the possibility of extending Andrew McCutchen beyond his current contract, Pirates owner Bob Nutting said:

"We are committed to try to find an opportunity. There is no one who we'd like to have for a career in a Pirates uniform more than Andrew."

Implicit in this language is that an opportunity to extend McCutchen does not currently exist. The Pirates are not committed to finding one. The Pirates are merely committed to trying to find one. Andrew looks angry here because he's probably going to get traded next off-season.

Thursday, October 1, 2015

The Pirates' best chance to beat Jake Arrieta

We've known for the better part of the last month that the Pirates' most likely postseason scenario would involve a home game against the Cubs in which they'd have to face Jake Arrieta, who's been so frustratingly good since July that a lot hitters would sooner gouge out their own eyes with a spoon than have to deal with the ungodly late movement two of his three different fastballs. It also doesn't help that Arrieta is seemingly impervious to left/right splits, so what can the Pirates do to beat him?

Probably nothing.

Here's how the Pirates who are the most likely to be on the roster for that game have fared against Arrieta in their careers:
*Sean Rodriguez has never faced Arrieta.

But here's what they can do:

1) Outlast him. 

Outlast him and they have a chance, much as they did on September 16th. Arrieta went eight innings, gave up six hits and a walk in eight innings. He struck out only five, and the Pirates managed to score two runs while he was in the game (one unearned), and the Pirates went on to lose in 13 frames. Once Arrieta is out of the game, the Cubs turn to a series of hard-throwing, walk-prone relievers who are eminently beatable. The key to beating the Cubs is to get Arrieta out of the game. To do this, the Pirates must look to run up Arrieta's pitch count. 

One way to do this is to mandate no first-pitch swings, at least the first time through the order, maybe even the second. The Pirates should not make a single first-pitch out in at least the first five or six innings of this game. Give Arrieta the first-pitch strike if it means making him pitch to you. 

Another way to do it is to structure the lineup so that Arrieta has to face the guys who see the most pitches more often than anyone else. No Josh Harrison hitting second, no Aramis Ramirez hitting cleanup.

Here's a list of those same regulars and the average number of pitches they've seen per plate appearance this season (2015 OBP in parentheses):
Andrew McCutchen: 4.04 (.405)
Michael Morse: 4.01 (.419)
Francisco Cervelli: 3.91 (.375)
Gregory Polanco: 3.89 (.320)
Neil Walker: 3.88 (.329)
Pedro Alvarez: 3.87 (3.18)
Jordy Mercer: 3.71 (.295)
Starling Marte: 3.61 (.334)
Sean Rodriguez: 3.64 (.282)
Aramis Ramirez: 3.49 (.298)
Josh Harrison: 3.46 (.322)

Among hitters with at least 500 PAs this season, the average number of pitches seen per PA is 3.82. Cervelli is 3 PAs away from qualifying, and Walker, Marte, McCutchen and Polanco are the only other Pirates with that many appearances. Mash those numbers together and your batting order for the Wild Card game should come out looking something like this:


Notes on this:

  • The hard truth of it is that the Pirates don't have a true leadoff hitter and should probably have been hitting Cervelli in that spot for a while now. That he hasn't hit any higher than sixth in the order since Jung Ho Kang went down is inexplicable.
  • There is a reasonable case to be made for swapping Polanco and Alvarez. The top four hitters in the order, though, must be static.
  • There's an even more reasonable argument for hitting Mercer 7, Cole 8 and Harrison 9.
  • In all likelihood, the Pirates are not going to do any of these things.
2) Be judicious with the pitching.

This means three things:
  • The only pitchers who should be allowed to appear in this game are: Cole, Happ, Blanton, Soria, Watson and Melancon. That's it; that's the list. 
  • The Pirates should use one of the last three regular season games as a bullpen game. Give Blanton three innings just to stretch him out, then let Jeff Locke, Vance Worley and Rob Scahill go nuts. 
  • Happ won't be needed until Game 3 of the Division Series, so he should be up and available for the Wild Card game. The Pirates shouldn't hesitate to pinch hit for Cole early (in the fourth or fifth innings) if the table's set for them to score with Cole's spot in the order coming up. This allows the Pirates to effectively have two long men ready for this game, leave Jared Hughes, Arquimedes Caminero and Antonio Bastardo off the roster and carry extra position players who will be far more valuable should the game go into extra innings.
The Cardinals are hobbling into the postseason. They're still a good team, but they're more beatable now than they've been in the last three years. If the Pirates can get past Arrieta and the Cubs, they have a legitimate shot at knocking off St. Louis, its depleted rotation, terrible bullpen and loathsome fanbase. We'll explore that if an when it becomes necessary.