Jump to content

Providing independent coverage of the Minnesota Twins.
The same great Twins Daily coverage, now for the Vikings.

The Store


Photo

Where the Twins Games Turn

  • Please log in to reply
18 replies to this topic

#1 Teflon

Teflon

    Senior Member

  • Members
  • 239 posts

Posted 19 April 2012 - 10:16 AM

Baseball Reference features a chart of each game that tracks the probability of either team winning based on the outcomes of specific at-bats compared to historic averages of outcomes observed for the same game situations. For example, the 4-6-3 double play turned behind Jason Marquis last night in the bottom of the 1st inning with the Twins leading 4-3 shifted the Twins probability of winning from 45% when the play started to 60% when the play ended - a 15% shift. This was the highest probability-affecting play last night. (Morneau's two homeruns were 2nd and 3rd, each about 11%.)

I looked at the top five plays for each game the Twins have played so far and have assigned the following general credit or blame:

4 Wins
Credit: Morneau homeruns; timely hitting from Span/Carroll/Valencia/Parmelee;defense getting key DPs.
(Surprisingly, the 2 Willingham HRs in wins haven't had a huge effect in win probability)


8 Losses
Blame: Good opponent starting pitching (C.J. Wilson, Matt Harrison); Liriano being brutal; bad bullpen work by Gray, Maloney and Perkins; too many LOB and GIDP.


Commentary: The pitching has really been the real culprit so far. Liriano doesn't give us a chance to win and the middle relievers can't keep us in a game consistently.

* * * * * * *

Wednesday, April 18th, Twins 6 , Yankees 5 in New York
Why we won: Morneau’s homeruns, Twins defense turning two key double plays behind Marquis, top of the batting order produces early runs.

Tuesday, April 17th, Twins 3 , Yankees 8 in New York
Why we lost: Liriano on the mound.

Monday, April 16th, Twins 7 , Yankees 3 in New York
Why we won: Big 5th inning from the top of the order, Morneau homerun

Sunday April 15th, Twins 3, Rangers 4 at Target Field
Why we lost: Perkins gets shelled, Mauer GIDP in 8th.

Saturday April 14th, Twins 2, Rangers 6 at Target Field
Why we lost: Jeff Gray’s 7th inning, Doumit leaves the bases loaded in the bottom of the 5th. (Twins leave 15 on base)

Friday April 13th, Twins 1, Rangers 4 at Target Field
Why we lost: Mauer and Doumit GIDPs, Effective pitching by Rangers’ Harrison.

Thursday April 12th, Twins 10, Rangers 9 at Target Field
Why we won: Morneau homerun in bottom of 8th, timely hitting by Span and Valencia in 7th.

Wednesday April 11th, Twins 6, Rangers 5 at Target Field
Why we won: Timely hitting by Pamelee and Carroll in 7th, Willingham HR in 4th.

Monday April 9th, Twins 1, Angels 5 at Target Field
Why we lost: Blackburn surrenders early runs. C.J. Wilson shuts down Twins bats.

Sunday April 8th, Twins 1, Orioles 3 in Baltimore
Why we lost: Willingham DP in 5th, Maloney ineffective relief in 6th, Hughes takes called third strike with 2 men on in 8th.

Saturday April 7th, Twins 2, Orioles 8 in Baltimore
Why we lost: Liriano on the mound.

Friday April 6th, Twins 2, Orioles 4 in Baltimore
Why we lost: Pavano surrenders early runs, Mauer GIDP, Ploufe, Willingham fail to drive in key runs.

#2 TwinVike61

TwinVike61

    Senior Member

  • Members
  • 164 posts

Posted 19 April 2012 - 10:59 AM

Wow...thanks for this great evaluation. Not that I want to focus negatively on our $23 Million man, but (in this admittedly small sample size of 12 games) the only place I see Mauer's name is as a contributor to three "why we losts" due to GIDPs. Some commentators critized me for labeling Joe as an anti-clutch, rally-killer a while back but I think that's what these stats tell us...

Edited by TwinVike61, 19 April 2012 - 11:18 AM.


#3 Teflon

Teflon

    Senior Member

  • Members
  • 239 posts

Posted 19 April 2012 - 11:56 AM

C'mon - it's Joe's birthday today!

#4 CDog

CDog

    Senior Member

  • Members
  • 856 posts

Posted 19 April 2012 - 12:04 PM

fangraphs does these as well. When I found them, I was immediately a little addicted. They're great viewing, especially, I think, to use them to learn about how certain things affect the game more or less than you would have thought.

#5 CDog

CDog

    Senior Member

  • Members
  • 856 posts

Posted 19 April 2012 - 12:04 PM

Some commentators critized me for labeling Joe as an anti-clutch, rally-killer a while back but I think that's what these stats tell us...


OR...you could jump to incorrect conclusions.

#6 TwinVike61

TwinVike61

    Senior Member

  • Members
  • 164 posts

Posted 19 April 2012 - 12:33 PM

OR...you could jump to incorrect conclusions.


Yea...I guess it would be wrong to conclude that Morneau's clutch homeruns increased the probability of a win versus grounding into a double play.

#7 Ultima Ratio

Ultima Ratio

    Super Friend

  • Members
  • 1,746 posts
  • LocationAt my computer

Posted 19 April 2012 - 12:42 PM

Very interesting. We've talked for years about the GIDP of Mauer and Cuddyer killing rallies. It's nice to have some numbers/stats to show their real (quantitative) effect in a game. Don't worry TwinVikes, some commenters have welded homer glasses on and get snarky and upset if you opine that Mauer is not worth 23 million, which he certainly is not since the contract's inception.
Man is born free, but everywhere he is in chains.

#8 mattsaari

mattsaari

    Junior Member

  • Members
  • 16 posts

Posted 19 April 2012 - 01:22 PM

WPA is a silly stat. According to it, the Os were only .09 % likely to win the game where the starter took a no-hitter into the 8th based on their team's overall production.

#9 John Bonnes

John Bonnes

    Owner

  • Administrators
  • 4,954 posts

Posted 19 April 2012 - 02:20 PM

WPA is a silly stat. According to it, the Os were only .09 % likely to win the game where the starter took a no-hitter into the 8th based on their team's overall production.


??? I think we might be talking about something different. The Oriioles had a 92.6% chance of winning that game going into the 8th. http://www.fangraphs...h=0&season=2012

#10 CDog

CDog

    Senior Member

  • Members
  • 856 posts

Posted 19 April 2012 - 02:25 PM

Yea...I guess it would be wrong to conclude that Morneau's clutch homeruns increased the probability of a win versus grounding into a double play.


That was a Grand Canyon sized reach. Very nice.

#11 Buddy14

Buddy14

    Junior Member

  • Members
  • 12 posts

Posted 19 April 2012 - 02:34 PM

That's actually kinda neat to look at. But I feel like just like a lot of other stats, they are real neat but I'd rather just watch the game and scream at the TV.

#12 Teflon

Teflon

    Senior Member

  • Members
  • 239 posts

Posted 20 April 2012 - 07:11 AM

Thursday, April 19th, Twins 6 , Yankees 7 in New York

Why we lost: Swarzak mistaken for Yankees batting practice pitcher. (2 2/3 innings - 8 hits, 3 homeruns. Biggest Win probability plays off Swarzak: Granderson 2-run HR in bottom of 2nd (19%), Teixeira 2-run HR in bottom of first (16%), Jeter RBI single in bottom of 2nd. (10%)

Biggest Missed opportunities:Parmelee K in top of 9th with no outs and man on 2nd (10%), Morneau K in top of 7th with 1 out and runners on 1st and 2nd (9%), Span ground out in top of 8th with 2 outs and a runner on third. (9%)

Why we made it close: Doumit 6th inning 2-run HR off Hughes (-16%), Doumit 1st inning 2-run single off Hughes (-15%), Valencia 8th inning lead-off double off Robertson.

#13 powrwrap

powrwrap

    Senior Member

  • Members
  • 425 posts

Posted 20 April 2012 - 08:43 AM

Does anybody know what they base their conclusions for Win Probabilities on? Seems kind of arbitrary to me. How can they assign a -16% to anything?
[FONT=comic sans ms][COLOR=#000000]"Baseball is like church. Many attend, few understand." [/COLOR][/FONT]

#14 mattsaari

mattsaari

    Junior Member

  • Members
  • 16 posts

Posted 20 April 2012 - 09:26 AM

??? I think we might be talking about something different. The Oriioles had a 92.6% chance of winning that game going into the 8th. http://www.fangraphs...h=0&season=2012


I meant that all the efforts of the combined team only gave them a net positive percentage of that much. Its silly because it's trying to assign probabilities to an action that essentially happens in a vacuum. The wpa should add up to 50%, since both teams start out 50/50 in any individual game.

#15 Thrylos

Thrylos

    Yes

  • Members
  • 4,278 posts

Posted 20 April 2012 - 09:30 AM

Does anybody know what they base their conclusions for Win Probabilities on? Seems kind of arbitrary to me. How can they assign a -16% to anything?



There not subjective "conclusions". It is math, formulas and calculations (and computers running them :) )

If more interested, this is a good starting point and make sure that you read all the articles linked...
-----
Blogging Twins since 2007 at The Tenth Inning Stretch
http://tenthinningst...h.blogspot.com/
twitter: @thrylos98

#16 Seth Stohs

Seth Stohs

    Owner

  • Administrators
  • 7,429 posts

Posted 20 April 2012 - 09:44 AM

Does anybody know what they base their conclusions for Win Probabilities on? Seems kind of arbitrary to me. How can they assign a -16% to anything?


WPA uses like 80 years worth of data and every situation, and notes how frequently that score, runners on, etc has happened and what percentage of time that person had one. If a team has a 3 run lead with bases loaded and two outs in the top of the ninth, they have like a 99% chance of winning (based on all those years of historical data). If the next batter hits a grand slam, suddenly the other team's chance of winning jumps to like 95%. I'd say that's a scenario where the batter who hit the grand slam had a pretty strong affect on a team's Win Probability.

#17 John Bonnes

John Bonnes

    Owner

  • Administrators
  • 4,954 posts

Posted 20 April 2012 - 01:58 PM

The wpa should add up to 50%, since both teams start out 50/50 in any individual game.


Again, this is the beauty of it. That's exactly what it does. If you add up all the positive WPA and negative WPA for the winning team, it equals .5. If you add them up for the losing team, it equals -.5. And if you add up all the cumulative WPA for all the members of a team, it's equal to "how many games the team is over .500" times .5.

#18 Teflon

Teflon

    Senior Member

  • Members
  • 239 posts

Posted 21 April 2012 - 09:06 AM

Friday, April 20th, Twins 5 , Rays 4 in St. Petersburg

Why we won: Jason Willingham's biggest hit of the year - a 3-run double to CF in the top of the 7th off Peralta brings the Twins back from down 4-2. (+47%)

Why we were worried: A Twins starter again showed good distance on his pitches. Hendriks surrenders a 2-run HR to Longoria (-28%) in the bottom of the 6th after a bases-loaded double play gave us hope the Twins might get out of the inning still tied. Hendriks also gave up a solo HR to Matt Joyce an inning earlier.

Twins pitchers have now given up a generous 22 HR in 14 games.

Edited by Teflon, 22 April 2012 - 07:24 AM.
no "c" in "Hendriks"


#19 Teflon

Teflon

    Senior Member

  • Members
  • 239 posts

Posted 22 April 2012 - 07:23 AM

Saturday, April 21st, Twins 1 , Rays 4 in St. Petersburg Why we lost: The contrast of 6th-inning outcomes. Twins - Doumit lines out with bases loaded and 2 outs in the top of the 6th. (+11%) Rays - With bases loaded and 2 outs in the bottom of the 6th, Upton singles to deep CF, 3 runs eventually score as Span's throw to 3rd hits the runner. (+31%)