Pages

12/7/17

Michael Freire - Run Differential - More The One Way To Skin A Cat



I left a comment earlier on one of Reese's posts and I wanted to explain what I was trying to say in a little more detail (I also miss contributing to this blog, beyond the comment section).

In short, run differential is the difference between a team's runs scored and runs allowed.  It can be used for any number of games played, but it is more accurate over a larger body of work.  Most websites will list these two figures somewhere within the team standings, usually to the right of the wins and losses.  The reason I focus on this statistic is that it is the end result of what a team has or has not done.   


There are no arguments once a game is over......your runs scored and runs allowed are what they are. 

Any casual fan with a calculator can use run differential as an accurate predictor of team success.

So, how do you figure this out?   By using the Pythagorean Win Expectation formula!


                                                                               Runs Scored (squared)
              Expected Win % =                          -----------------------------------------------
                                                               Runs Scored (squared) + Runs Allowed (squared) 



***You take the result of the calculation and then multiply it by the number of games played (162 for a full season), which will give you the "expected" win total based on the differential.

***There are many versions of this formula floating around and it has been tweaked over the years.  However, to keep things simple, the formula listed above will get you into the ballpark (see what I did there).


As a quick example, lets look at how the two best teams during this past season did and how the formula works (I saved you some time and I did not list the actual math, because no one wants to see that);

Houston Astros - 896 to 700 = 0.621% or 100.50 Expected Wins (2017 Record 101-61)

LA Dodgers - 770 to 580 = 0.638% or 103.36 Expected Wins (2017 Record 104-58)

Very impressive, as the formula pretty much hit the nail on the head with regards to the expected win totals.  What is equally important to note is that the two teams accomplished their impressive win totals in markedly different fashions.  The Astros scored 126 more runs the the Dodgers did over the course of the season which is 0.77 runs per game.   However, the Dodgers allowed 120 fewer runs over the entire season, which is predictably 0.74 fewer runs per game.

One team "clubbed" their opponents to death, while the other "suffocated" their opponents, but they both ended up as similarly successful franchises.  So much 
so that they staged one of the best World Series' in recent memory that could have gone either way.

Let's look at our favorite team and how they did during the 2017 season;

NY Mets - 735 to 863 = 0.421% or 68.11 Expected Wins (2017 Record 70-92)

Again, pretty close for the formula with regards to the putrid performance that the Mets put forth last year.   If you can keep your focus while reviewing the numbers, the Mets weren't too far off from what the Dodgers produced in the runs scored column (35 fewer runs over the course of an entire season is only 0.22 runs per game). 


 However, they gave up 283 more runs over the course of the year or 1.75 runs per game!

Teams can be built in many different ways, whether it is a balanced approach or being elite on one side of the equation or the other.  Taking this into account, the Mets' should focus on improving their "runs allowed" results if they want to have a chance at a successful 2018 season.  The good news is that the roster still features a potentially dominant pitching staff and that is precisely why they hired Mickey Callaway as the next manager.

The front office appear to understand what went wrong in 2017, so it is a step in the right direction.

6 comments:

  1. A great explanation.

    I always said... if you score one more run that the team you are playing, you just had a good game.

    One run more.

    ReplyDelete
  2. Mike Great Post.. looking forward to many more contributions

    ReplyDelete
  3. Thanks, Mack and I appreciate the offer to contribute (again).

    Mike

    ReplyDelete
  4. That is why I want the off-season investment to be in pitching, pitching and more pitching.

    I'm too lazy to look it up, but I'd be curious to see the run differential pre-July 31st trading deadline and post-deadline. When a lot of bats left, did the run production drop off significantly?

    ReplyDelete
  5. Ok, so a stats question is like the "bat signal", I am a secret stats geek.

    From 08-01-17 until the end of the year, the Mets played in a total of 58 games and their run differential was 241 to 318, which computes to an expected win percentage of 36.5% or a record of 22-36 for that period (worse then the first 104 games).

    On 08/01/17 the were 48-56 and they ended up 70-92, or a 22-36 finish to the year, as the formula predicted.

    ***They averaged 4.75 runs per game before the deadline and 4.16 runs per game after all the trades. Additionally, they also gave up 5.24 runs per game before the deadline and 5.48 afterward.

    The overarching theme here is that they pretty much sucked all year, with the last third of the season being a bit worse.

    Mike

    ReplyDelete
  6. Nice formula Mike - and I immediately saw what you saw: allow nearly 300 runs more than the Dodgers and the season W/L record will be a whole lot different.

    Amazing the Mets won 73 games in 1968 while scoring just 473; scoring 263 more runs in 2017 netted them 3 les wins. The 1968 pitching and the 2017 offense combined would give 100 wins easy.

    ReplyDelete