Is Playing Time All That Important- Exploring WAA with R

[Omar 382]

We are all familiar with WAR, or Wins Above Replacement. A common complaint I've heard with it is that it looks at a player's worth relative to a replacement player. What if, instead of a replacement player, we looked at a player's worth relative to his peers? That's where WAA comes in. I used R to sort the players with the biggest gap between WAR and WAA and those with the smallest gap.

The worst player by difference between WAR and WAA is Hank Aaron, with Pete Rose a close second. Why? Because WAA does not take into account playing time. Through age 35, Pete Rose was worth (in 10,001 PA), 70.1 WAR and 39.8 WAA. He had 2,762 hits at that point. You can see what I’m getting at here. From that point on, Rose collected 5,889 more PAs. He added a bit to his career WAR (just 6.6 WAR in ten years). However, he destroyed his WAA, coming in at –10.7.

So basically, to dumb it down, WAA might be used to look at those annoying stat-padders. The best players by WAR/WAA gap were Charlie Keller (hampered by back injuries, but one of the best players of the '40s when on the field) and Harry Stovey, who I admittedly don't know a whole ton about. Here are the results.
Worst players
Best players (the bottom of the list is the best, and gets worse the higher you go)

 

[soxfan1468927]

I think you're missing some guys.

Sam Rice: 52.8 WAR, 15.6 WAA
Harry Hooper: 53.1 WAR, 15.3 WAA
Johnny Damon: 56.0 WAR, 19.1 WAA

Where did you get the numbers?

 

[Omar 382]

I think you're missing some guys.

Sam Rice: 52.8 WAR, 15.6 WAA
Harry Hooper: 53.1 WAR, 15.3 WAA
Johnny Damon: 56.0 WAR, 19.1 WAA

Where did you get the numbers?

I took players with 40+ WAR since 2009. That's all my database had. Unfortunately, BB-Ref sucks for exporting data (and I think you would even have to pay to get WAA), and Fangraphs doesn't have WAA as a stat (yet).

 

[Omar 382]

It's funny, because two of the most overrated players in baseball history in my opinion, Aaron and especially Rose, are at the top. Ty Cobb was a surprise, but he's another guy that tacked on at the end of his career.

 

[Omar 382]

Even though I love the guy- Pete Rose is a bum! #NoHOF4You

 

[Omar 382]

Actually, looking at the data, I noticed I fucked up by taking the biggest difference. That does penalize guys more than it should who played longer, because they were superior players and could afford to play longer. After all, there is some value to staying on the field. Now let's look at the players who had the least percentage of their WAA make up their WAR. This will highlight those who were able to stay on the field for a while, but whose actual production is boosted by WAR simply because they stayed on the field so long.

Top being the worst offenders. Surprise, surprise. Look who made the last at 19....

 

[Omar 382]

Here's how Rose's WAR/WAA shakes out:

• +368 runs for batting
• +13 runs for baserunning
• +5 runs for avoiding double plays
• –54 runs for his defense
• –91 runs for the positions he played

which gives you a total of +241 runs above average. If you convert those runs to wins, you get 29.2 WAA. If you add one more component though, based on playing time,

• +492 replacement runs

Rose then has a total of +733 runs above replacement. That converts to 76.7 WAR. Those 492 runs say “if a replacement player played as much as Rose, he would have been worth 492 runs below average“. Therefore, Rose is 241 runs above average, but a whopping 733 above replacement—mostly because of how long he played.

Bum!!!

 

[soxfan1468927]

I took players with 40+ WAR since 2009. That's all my database had. Unfortunately, BB-Ref sucks for exporting data (and I think you would even have to pay to get WAA), and Fangraphs doesn't have WAA as a stat (yet).

Not sure how much time you've spent chatting with the Hall of Stats guys. Adam Darowski wrote this article, with attached spreadsheet, at the beginning of the year:

Hall of Stats: An Exploration: Taking WAR Out of Hall Rating

 

[Omar 382]

Not sure how much time you've spent chatting with the Hall of Stats guys. Adam Darowski wrote this article, with attached spreadsheet, at the beginning of the year:

Hall of Stats: An Exploration: Taking WAR Out of Hall Rating

He provided me with the spreadsheet. I sorted it. I also took the Pete Rose figure from it. Also, I think the article is like 5 years old right?

 

[soxfan1468927]

He provided me with the spreadsheet. I sorted it. I also took the Pete Rose figure from it. Also, I think the article is like 5 years old right?

January 1, 2017 and all the comments are from 2 months ago.

So no, not 5 years old.

 

[MilkSpiller22]

Yea, i have said many times how much i hate WAR but I do llike WAR7...

But even WAR7 has a problem, especially for pitchers... olden day pitchers often pitched well more than 300 innings a season, while modern day pitchers barely get over 200 a season... WAR is too based on the how much you play- which is your whole point...

So i will agree to WAA being better, just for that reason...

But i will say this, i am not a fan of comparing to averages... this is what i don't like about adjusted stats... Problem with averages is that we are using everybody, even the players that suck... I would much rather use an average amongst a certain population of players... I would think this would be much more telling... Especially in comparing the old school vs the new school... problem with pre expansion, is that there were not as many players, so the averages dont normalize as much...

 

[Omar 382]

Yea, i have said many times how much i hate WAR but I do llike WAR7...

But even WAR7 has a problem, especially for pitchers... olden day pitchers often pitched well more than 300 innings a season, while modern day pitchers barely get over 200 a season... WAR is too based on the how much you play- which is your whole point...

So i will agree to WAA being better, just for that reason...

But i will say this, i am not a fan of comparing to averages... this is what i don't like about adjusted stats... Problem with averages is that we are using everybody, even the players that suck... I would much rather use an average amongst a certain population of players... I would think this would be much more telling... Especially in comparing the old school vs the new school... problem with pre expansion, is that there were not as many players, so the averages dont normalize as much...

Not as many players =! not normalized

There are two ends of the spectrum, my friend- the really bad and the really good.