WAR Totals

[Enscheff]

WAR totals 69 games into the season. Position payer WAR taken from FG, framing wins taken from BP (framing runs / 10).

Hitting/Defense/Baserunning
Freddie Freeman 2.6
Ender Inciarte 2.1
Tyler Flowers 1.6
Brandon Phillips 1
Matt Adams 1
Matt Kemp 0.9
Kurt Suzuki 0.6
Nick Markakis 0.2
Johan Camargo 0.1
Chase d'Arnaud 0.1
Danny Santana 0.1
Dansby Swanson 0
Jaime Garcia 0
Sam Freeman 0
Matt Wisler 0
Sean Newcomb 0
Anthony Recker -0.1
R.A. Dickey -0.1
Julio Teheran -0.2
Lane Adams -0.2
Bartolo Colon -0.2
Rio Ruiz -0.2
Mike Foltynewicz -0.3
Jace Peterson -0.5
Adonis Garcia -0.5
Emilio Bonifacio -0.6
7.4

Pitching
Jaime Garcia 1.1
Jim Johnson 1
Mike Foltynewicz 0.5
Arodys Vizcaino 0.3
Bartolo Colon 0.2
Sean Newcomb 0.2
Jose Ramirez 0.1
Luke Jackson 0.1
Sam Freeman 0.1
Jason Hursh 0.1
Matt Wisler 0
Chaz Roe 0
R.A. Dickey -0.1
Ian Krol -0.2
Jason Motte -0.2
Julio Teheran -0.2
Eric O'Flaherty -0.2
Josh Collmenter -0.5
2.3

Framing
Tyler Flowers 1.1
Kurt Suzuki -0.1
1

Total 10.7

A replacement level team should have won 20.4 games after playing 69 games. Add that to 10.7 WAR the Braves have totaled and it equals 31.1 wins. The Braves have actually won 32 games.

So....

Who is still doubting WAR as a way to evaluate players?

[msstate7]

I'm not doubting it all. In fact, fWAR is unquestionably the best measure out there. Bc of how great it is, we should change the KEMvP bet to fWAR haha

[msstate7]

How does BR add up?

[Enscheff]

I'm not doubting it all. In fact, fWAR is unquestionably the best measure out there. Bc of how great it is, we should change the KEMvP bet to fWAR haha

No chance mmstake7.

Not sure how BP stacks up because I didn't see a way to easily export to Excel on their site. Plus, I think FG does a better job evaluating fielding value haha. But I knew that when you let me pick which version of WAR we would bet on!

[thewupk]

How does BR add up?

Braves are at 9.9 bWAR at the moment.

[Enscheff]

Braves are at 9.9 bWAR at the moment.

Does that include framing runs?

[thewupk]

No chance mmstake7.

Not sure how BP stacks up because I didn't see a way to easily export to Excel on their site. Plus, I think FG does a better job evaluating fielding value haha. But I knew that when you let me pick which version of WAR we would bet on!

They add it up for you in the team stats page.

[Enscheff]

They add it up for you in the team stats page.

They have 7.2 for the position players (vs 7.4 from FG), and 2.7 for pitchers (vs 2.3 for FG). I am pretty sure catcher bWAR doesn't include framing runs, so the total bWAR is 10.9 (vs 10.7 from FG).

Seems legit all the way around to me.

Don't tell the derpi-Barves.

[thethe]

Yankees Hitting - 16.1
Yankees Pitching - 9.6
Pitch Framing - Not sure where to find so I'll assume 0

Total war 25.7

Number of games played - 67
Replacement wins pro-rated - 19.85
Replacement wins + total war - 45.55
Total number of Yankee wins - 38

Is that right?

[thewupk]

Yankees Hitting - 16.1
Yankees Pitching - 9.6
Pitch Framing - Not sure where to find so I'll assume 0

Total war 25.7

Number of games played - 67
Replacement wins pro-rated - 19.85
Replacement wins + total war - 45.55
Total number of Yankee wins - 38

Is that right?

Yankees are also 7-12 in 1 run games. A huge disparity like that will skew actual totals. Yankees pythag W/L which is based on actual runs scored and given up is 43-24. Yankees have been a bit unlucky.

[thethe]

Yankees are also 7-12 in 1 run games. A huge disparity like that will skew actual totals. Yankees pythag W/L which is based on actual runs scored and given up is 43-24. Yankees have been a bit unlucky.

The Braves have had some lucky come from behind wins this year though right? Don't we have the most runs scored after the 6th inning? So is our win total really true?

Either way - when you include enough players into the WAR equation you eliminate the variability risk. I'm not sure anyone has ever said that TEAM WAR SUCKS! But, there is a lot of noise in any value calculation when your data set is smaller.

[msstate7]

Yankees are also 7-12 in 1 run games. A huge disparity like that will skew actual totals. Yankees pythag W/L which is based on actual runs scored and given up is 43-24. Yankees have been a bit unlucky.

Braves are 7-10 in 1-run games

[thethe]

The delta for the Rockies with their WAR calc is 8-9 wins (negative - not sure where to source framing numbers).

Again - to just use one example as HEY THIS WORKS or HEY THIS DOESN'T WORK is just silly.

If you make any data set large enough its not very difficult for statisticians to create statistics that work.

[thewupk]

Braves are 7-10 in 1-run games

8-10. Accounts for the 1 game difference

[Enscheff]

Braves are 7-10 in 1-run games

They are 8-10...exactly what you'd expect a team with their winning percentage to be...more or less

[thewupk]

The Braves have had some lucky come from behind wins this year though right? Don't we have the most runs scored after the 6th inning? So is our win total really true?

Either way - when you include enough players into the WAR equation you eliminate the variability risk. I'm not sure anyone has ever said that TEAM WAR SUCKS! But, there is a lot of noise in any value calculation when your data set is smaller.

Pythag and WAR are based on the fact that ~10 runs equals 1 win in major league baseball. So if you score 50 more runs than you've allowed on the season you are going to be ~5 games over .500. The Braves can score a ton of runs late and they wouldn't be "lucky" because they all count regardless of when they are scored. The issue with 1 run games is because if you have a large disparity one way or the other then your actual record is going to be skewed by your runs allowed and given up. If you are 5 games under .500 in 1 game games then that's a 5 game swing with only a 5 run difference. That's going to skew pythag and WAR records.

[thethe]

Pythag and WAR are based on the fact that ~10 runs equals 1 win in major league baseball. So if you score 50 more runs than you've allowed on the season you are going to be ~5 games over .500. The Braves can score a ton of runs late and they wouldn't be "lucky" because they all count regardless of when they are scored. The issue with 1 run games is because if you have a large disparity one way or the other then your actual record is going to be skewed by your runs allowed and given up. If you are 5 games under .500 in 1 game games then that's a 5 game swing with only a 5 run difference. That's going to skew pythag and WAR records.

Wouldn't you then need to normalize all wins and losses based on that sliding scale whereby you have the following:

1 run win = .1 wins
2 run win = .2 wins
etc...

1 run loss = .1 losses
2 run loss = .2 losses

Then prorate those figures to 162 games? Why just 1 run games?

[Enscheff]

Wouldn't you then need to normalize all wins and losses based on that sliding scale whereby you have the following:

1 run win = .1 wins
2 run win = .2 wins
etc...

1 run loss = .1 losses
2 run loss = .2 losses

Then prorate those figures to 162 games? Why just 1 run games?

The hell?

We only adjust for 1 run games because they are mostly a function of unsustainable luck.

https://www.beyondtheboxscore.com/20...inable-orioles

[thewupk]

Wouldn't you then need to normalize all wins and losses based on that sliding scale whereby you have the following:

1 run win = .1 wins
2 run win = .2 wins
etc...

1 run loss = .1 losses
2 run loss = .2 losses

Then prorate those figures to 162 games? Why just 1 run games?

No because an enscheff said in 1 run games are primarily luck based and you'd expect teams to be around .500 for a given year.

So in the Yankees example I would expect them to have a better winning percentage going forward if they continue to perform at the same level.

Now teams like the rangers last year do exist. Teams can have sustained luck for whatever reason but it shouldn't be expected which is kind of the whole point of WAR and a runs based system.

The goal is to score more runs than you allow. Generally teams that do that win more games. Teams with higher war will generally outscore their opponents the most.

To fight against that notion is a losing battle imo.

[thewupk]

They have 7.2 for the position players (vs 7.4 from FG), and 2.7 for pitchers (vs 2.3 for FG). I am pretty sure catcher bWAR doesn't include framing runs, so the total bWAR is 10.9 (vs 10.7 from FG).

Seems legit all the way around to me.

Don't tell the derpi-Barves.

Yeah framing runs is not included in either fWAR or bWAR as far as I know.