Tuesday, October 15, 2013

Red Sox Playoff Trivia

Win Probability Added (WPA) measures a baseball player's contribution to the probability that his team wins a given game. So it's a situational stat that takes into account only the context and the result. If you strike out in the first inning, you might decrease your team's chance of winning by 2 percentage points, and therefore get a WPA of -0.02 for that event. If you come up to bat in the bottom of the 9th with 2 outs, behind by 1 run, and hit a 2-run walkoff HR, you might increase your team's chances of winning from, say, 30% to 100%, and you'd get a WPA of 0.70 for that event. Similarly, if the pitcher strikes you out to end the game in that plate appearance, he would get a WPA of 0.30.

So here's the trivia question: What 2 Red Sox players led the team in WPA in games 4 and 5, respectively, of the 2004 ALCS?

Your only hint: The answer is NOT David Ortiz, for either game. In each game, Ortiz finished 2nd on the Red Sox in WPA.

26 comments:

Mark Bellhorn said...

Was it me?!

Luke Murphy said...

Nope, sorry Mark. You were up there in game 6, though.

Dad said...

Bill Mueller and Jason Varitek. That it's not Ortiz really makes me worry about the WPA.

No, wait. Mueller and Trot Nixon.

Game 5 has to be Ortiz.

Caleb said...

I'm familiar with the WPA leader for game 5, so I won't guess there.

My guess for Game 4 is...Kevin Millar.

Caleb said...

Follow-up trivia question - who was the Red Sox's WPA leader for game 4 (the clinching game) against the Rays last week?

Luke Murphy said...

Well, WPA takes sums up not just the positive plays, but the negative ones as well. Ortiz was 2/5 with a walk in game 4, so he made 3 outs, and those outs came in pretty big situations. His walk-off HR in the 12th came with 0 outs and a runner on first, so the win probability was already 73%. So that hit was worth 0.27. Ortiz's 2-RBI single in the 5th with 2 outs to give the Red Sox a 3-2 lead was worth 0.23. His walk in the 1st inning was only worth .02. So there's 0.52 in his 3 positive events.

However, Ortiz also struck out with no outs in the bottom of the 8th with Manny on 1st. That cost him 0.09. He popped up to end the bottom of the 9th with the bases loaded. That one really hurt him, taking the Red Sox chances from 67% to 50%, costing Ortiz 0.17. He also lost .03 with an out in the 4th inning.

That all adds up to Ortiz scoring 0.234 WPA for the game. We remember his walkoff HR in the 12th, but it was really his 3rd chance to win the game. So, somebody on the team contributed significantly to giving Ortiz that last chance, kinda like the Patriots' defense for Tom Brady this past Sunday.

Luke Murphy said...

Game 4 was not Mueller or Millar, and game 5 was not Tek or Nixon. Decent guesses though. Millar's 2 walks and hit in game 4 were pretty big, and he was 3rd among Red Sox batters with 0.212 WPA. Mueller actually didn't do that well, despite his game-tying single being the 2nd most valuable hit in the game, worth 0.23. He made too many outs (including a GIDP to end the 6th) in other key situations though, and ended with just 0.059 WPA.

Luke Murphy said...

Ortiz's game 5 was huge, and he totaled 0.411 WPA, but somebody else on the team still bested him. Nixon went 1/4, but his one hit was worth 0.22 WPA when he moved Dave Roberts over to 3rd with no outs in the 8th in order to set things up for Tek's sac fly. His game total was 0.078 WPA.

Tek actually had a pretty bad game. He was 0/4 with a walk. His walk actually drove in a run in the bottom of the 1st, so it did earn him .08 WPA. His 8th inning sac fly to tie the game actually DECREASED the Red Sox's odds of winning. That's pretty counter-intuitive, but it makes sense. A runner on 3rd with no outs already has a very high probability of scoring. Tek drove that run in, but stranded Kapler at 1st and made an out. A strikeout would've been even worse, but that sac fly actually costed Tek .04 WPA. He finished the game with a score of -0.215.

Dad said...

Your other hint is here:

"Similarly, if the pitcher strikes you out to end the game in that plate appearance, he would get a WPA of 0.30."

The answer is two pitchers. One was with the RS for many years, and the other for a brief time.

Luke Murphy said...

Yup. You got it.

Dad said...

I think it was the last game Leskanic ever pitched.

Luke Murphy said...

Pitching a scoreless top half of a 9th or extra inning is worth about 0.16 WPA. Also, coming in with 2 outs and the bases loaded in a tie game in a 9th or extra inning and getting the 3rd out without allowing a run is worth about 0.18 WPA.

Luke Murphy said...

Yes it was. Amazing.

So, just to be the clear, the answer is Curtis Leskanic in game 4, and Tim Wakefield in game 5.

Luke Murphy said...

Heh, I googled Leskanic's name and found this:

http://www.theonion.com/articles/red-sox-attempt-to-break-fabled-curse-of-relief-pi,2310/

Dad said...

"Pitching a scoreless top half of a 9th or extra inning is worth about 0.16 WPA"

Does it not matter what the score is?

Luke Murphy said...

It certainly would matter what the score is if it's the 9th, but in extras the score would necessarily be tied, of course. Trevor Rosenthal earned 0.148 WPA for allowing 0 runs in the top half of the 9th in game 2 of this year's NLCS, for example. The Cardinals' odds went from 85% to 100%. Koji Uehara only earned 0.037 WPA for pitching a scoreless 9th in game 1 of the ALCS with the Sox trailing 1-0. Basically, the Red Sox' odds went from bad (16%) to slightly less bad (%20).

Dad said...

Does this stat accumulate for a player? Do players have lifetime WPAs?

Luke Murphy said...

Yes. Ortiz has 39.6 regular season WPA in 1969 games, and 2.549 postseason WPA in 73 games. So that's .02 WPA/G in the regular season and 0.035 WPA/G in the playoffs. In other words, by this measure, Ortiz has helped his team win during the playoffs at almost double his regular season rate.

Dad said...

So, does that mean that for each game, the sum of Ortiz's at bats would increase the chances of the RS winning by 2% in the regular season and 3.5 % in the playoffs. Trying to get a grip on what this stat really means.

Dad said...

I mean, of course, for the AVERAGE game.

And how do those numbers compare to other players?

Luke Murphy said...

Yeah, I think that's about right. It's going to be relative to the average player, though, since the win probabilities are determined in the first place from historical outcomes.

Luke Murphy said...

http://www.baseball-reference.com/leaders/wpa_bat_active.shtml

Ortiz ranks 8th among active players in regular season batting WPA. He'll be 5th though starting next year, when I don't think Manny, Helton, or Berkman will still be considered active.

Luke Murphy said...
This comment has been removed by the author.
Luke Murphy said...

http://www.baseball-reference.com/leaders/wpa_bat_career.shtml

59th all time. Barry Bonds makes Ortiz look like Julio Lugo.

Luke Murphy said...

http://www.baseball-reference.com/leaders/wpa_bat_season.shtml

Ortiz's 2005 season was the 11th best season of all time for batting WPA.

Luke Murphy said...

I can't find or create a leaderboard for postseason WPA. I'd love to see what the best single postseason WPA of all time is. I know that David Freese's game 6 performance in the World Series a couple years ago in which I think he had both the game-tying and game-winning hits was the best single postseason game effort of all time. This guy does some silly weighting thing where he counts World Series and LCS WPA more than LDS WPA, but he at least has career totals for playoff WPA posted (from 2011). At that time Ortiz had 1.93 playoff WPA. Mariano Rivera looks like the undisputed champion of playoff WPA.

http://www.beyondtheboxscore.com/2011/10/17/2495296/wwpa-weighted-postseason-win-probability-added