Minnesota Twins News & Rumors Forum
  • Twins 3, Tigers 4: The Value of Each Play

    On Monday night, the Minnesota Twins began a three game series in Detroit against the Tigers. It was Mike Pelfrey facing Max Scherzer. You likely watched the game or saw the box score already when you read this, so you know that the Twins fell by the final score of 4-3.

    It was interesting to watch the game while at the same time calculating the Twins Win Expectancy after each and every play throughout the game. Win Expectancy is a number that changes with every game situation. Someone who had too much time on his hands looked at every play in MLB games from 1957 through 2005. It assigns a percent likelihood of a team winning based on what happened over nearly 100,000 games.

    Here is a look at the Top 5 Plays from tonight’s Twins-Tigers game in terms of effect on Win Expectancy.

    #5 – Florimon Turns Double Play from Back Side.

    It was the bottom of the 5th inning and the Twins led 3-1. Mike Pelfrey had allowed a lead-off single to Alex Avila. With Avila on 1st, the Twins’ Win Probability was at 66.1% Omar Infanta blasted a one-hopper right at the Twins shortstop. Florimon fielded the ball cleanly but as he was throwing to second base, he was falling backwards. Despite his imbalance, he made a perfect feed to Brian Dozier who quickly turned and fired to first base to complete the double play. Now with two outs and no one on, the Twins’ Win Expectancy increased to 76.2% Win Expectancy Change – 10.1%

    #4 – “Hammer” Homers to give Twins 1-0 Lead.

    A visiting team’s Win Expectancy when the game starts is just 46.1% Brian Dozier and Joe Mauer flied out. When Josh Willingham stepped to the plate, that Win Expectancy was down to 42.1%. Scherzer didn’t give up many hard hit baseballs through the game, but Willingham did get a hold of this one for a solo home run that gave the Twins the 1-0 lead. Following the homer, the Twins Win Expectancy was 53.0% Win Expectancy Change- 10.9%

    #3 – Dirks Homers to Cut Twins’ Lead in Half.

    In the 2nd inning, the Twins added a second run on an Aaron Hicks’ ground out to take a 2-0 lead (which increased the Twins Win Expectancy by 4.3% to 69.8% After the Tigers and the Twins each posted "three up-three down" innings, Andy Dirks came up with two outs and no one on base in the bottom of the third. The left-hander’s solo home run cut the Twins’ lead to 2-1. It also cut the Twins’ Win Expectancy from 71.0% to 58.9% Win Expectancy Change – 12.1%

    #2 – Arcia Double Gives Twins 3-1 Lead.

    The Tigers (Dirks) had cut the Twins lead to one run in the bottom of the third, but the Twins were able to get two doubles to get that run back and reclaim a two-run lead. With one out, Chris Parmelee doubled to very deep right center field. After Trevor Plouffe struck out for the second out, Oswaldo Arcia fell behind Scherzer 0-2. However, he was able to fight off a fastball and hit a soft liner just inside the left field line for an RBI double. After the Plouffe strikeout, the Twins Win Expectancy was 57.8% However, after the RBI double, the Twins Win Expectancy jumped to 71.6% Win Expectancy Change – 13.8%



    #1 – Fielder Blast Gives Tigers 4-3 Lead.

    If you watched the game, or saw highlights, it was pretty clear which play of the game had the biggest impact on Win Expectancy. When the bottom of the 6th inning began, the Twins’ Win Expectancy was at 73.3% After an Andy Dirks bunt single to lead off the inning, it fell to 67.5%. Following the Miguel Cabrera walk to put runners on 1st and 2nd the Win Expectancy fell to 59.5%. However, after Pelfrey’s first pitch to Prince Fielder landed behind the wall in centerfield, the Twins Win Expectancy was just 23.1% Win Expectancy Change – 36.4%



    (Please note that this chart shows the Win Probability for the HOME team, the Tigers, throughout the game.)

    Pelfrey was very good through the first five innings. Unfortunately, the struggles in the sixth inning happened, and happened quite quickly. Win Expectancy is a nice, fun tool that can be used to see the impact of individual plays on the game. Like all statistics, it isn’t a perfect tool.

    For instances, when Fielder came up, the Twins Win Expectancy was 73.3% whether it was Prince Fielder or Don Kelly or Tom Kelly stepping to the plate.

    If you take a look at the Win Expectancy change for each player in the Twins lineup, it can show who had the biggest impact, positive or negative, in the game. It is extremely important to point out that this is a one-game sample, so performance (or clutchness), good or bad, has to be taken for what it’s worth. For this one game, here are the Twins hitters impact on Win Expectancy.


    • Brian Dozier -15.5%
    • Joe Mauer -9.2%
    • Josh Willingham +5.6%
    • Justin Morneau -11.4%
    • Chris Parmelee -4.4%
    • Trevor Plouffe -4.9%
    • Oswaldo Arcia +14.3%
    • Aaron Hicks -2.1%
    • Pedro Florimon -0.1%


    It is safe to say that Oswaldo Arcia was the Twins player who gave the team the best chance to win on Monday night.

    Win Expectancy is just one way to look at a game. It is kind of fun to see how much each play, from the first to the last, affects the game. On this night, one pitch completely altered the outlook of the game. Unfortunately, it was one pitch that meant a Twins loss instead of a Twins win.
    This article was originally published in blog: Twins 3, Tigers 4: The Value of Each Play started by Seth Stohs
    Comments 10 Comments
    1. snepp's Avatar
      snepp -
      If anyone else is interested in win probability stuff, Fangraphs provides all the dirty work for every game in near real time, as well as providing some glossary info and links to articles discussing it.


      Here's the link to last night's game.
    1. Wheresatom's Avatar
      Wheresatom -
      Nice write up. This Win Probability stat is kind of new to me but very interesting. I might keep an eye on the fangraphs page for a game or two for some extra info.
    1. savvyspy's Avatar
      savvyspy -
      This way of looking at games seems to place a ton of value on power hitting. If accurate, the front office insistence for putting together a roster that consistently finishes in the Bottom 5 in HRs seems counterproductive to assembling a winning team.
    1. sorney's Avatar
      sorney -
      Quote Originally Posted by savvyspy View Post
      This way of looking at games seems to place a ton of value on power hitting. If accurate, the front office insistence for putting together a roster that consistently finishes in the Bottom 5 in HRs seems counterproductive to assembling a winning team.

      Power hitting SHOULD be valued as it's the quickest way to put runs on the board...last I checked, more runs scored generally means more wins. What I don't get, is the game has changed from the 70's/80's way of playing ball, and the Twins just now are slowly starting to move away from that. What has taken them so long?!?! Oh well, more of a half hearted rant while enjoying a super surprising 500 team for as long as possible.
    1. jimbo92107's Avatar
      jimbo92107 -
      From this chart I gather that what the Twins really need is guys that make lots of great defensive plays and hit lots of home runs, especially late in games.
    1. IdahoPilgrim's Avatar
      IdahoPilgrim -
      Very interesting. Like all stats, of course, it has to be taken inside a larger context.

      According to fangraphs, Scherzer had a WPA last night of 0.88; Smyly had one of .193. Yet somehow I still think Scherzer added more to the Tigers win than Smyly. It also gives Pelfrey the lion's share of responsibility for the defeat, yet there were plenty of key Twins bats that were completely silent and could have made a significant difference.
    1. Seth Stohs's Avatar
      Seth Stohs -
      Yeah, I don't think WPA is a good statistical thing for pitchers.
    1. snepp's Avatar
      snepp -
      Quote Originally Posted by sbknudson View Post
      It also gives Pelfrey the lion's share of responsibility for the defeat, yet there were plenty of key Twins bats that were completely silent and could have made a significant difference.
      Huh?

      Pelfrey was credited with -.323 WPA

      The hitters were credited with -.318 WPA


      That's effectively identical responsibility for the loss.
    1. ThejacKmp's Avatar
      ThejacKmp -
      Quote Originally Posted by sorney View Post
      Power hitting SHOULD be valued as it's the quickest way to put runs on the board...last I checked, more runs scored generally means more wins. What I don't get, is the game has changed from the 70's/80's way of playing ball, and the Twins just now are slowly starting to move away from that. What has taken them so long?!?! Oh well, more of a half hearted rant while enjoying a super surprising 500 team for as long as possible.
      Well this is where the author's criticism of not taking into account who is up (and up after that) comes into play. Though Joe Mauer doesn't hit a lot of homeruns, I think most Twins fans would love to see him up with guys on 1st and 2nd in a two run game. He's not likely to hit a three run bomb but he's very likely to put the next batter in an advantageous position.

      It's not necessarily that you should get the 9 best home run hitters in your lineup - though that seems fun you'd have a lot of solo HR driven games. The tool seems like it needs some help - for example, the Tigers win % should probably have been higher given who was coming up. Having a higher OBP guy followed by their big bats is better than having Don Kelly coming up.
    1. ashburyjohn's Avatar
      ashburyjohn -
      Quote Originally Posted by sbknudson View Post
      It also gives Pelfrey the lion's share of responsibility for the defeat
      In Aesop's story, the lion decided how much everyone's share was. Usually it would be 100% for himself, but here I think he'd be generous and take zero.
©2014 TwinsCentric, LLC. All Rights Reserved.
Interested in advertising with Twins Daily? Click here.