A small experiment on WAR and aging

[Frobby]

I know more scientific analyses have been done on aging in baseball, but I got curious on the odds in today’s environment that Rendon’s $245 mm contract would pay off.    His contract runs through ages 30-36, so I looked at all active players 36 and older who are among the top 100 in position player WAR.    There are 13 such players: Pujols, Cabrera, Cano, Kinsler, Granderson, Zobrist, Molina, Ru. Martin, Cruz, Encarnacion, Choo, Pence, and Phillips.     For each player, I calculated the percentage of their WAR through age 36 that they had accumulated by age 29.    The average comes out to 51.2%; the median was 53.5%.

Applying those to Rendon’s case (27.3 WAR through age 29), the mean average outcome would be 26.0 WAR from ages 30-36.    The median outcome would be 23.7 WAR through age 36.   So, that’s an expectation of $9.4 mm/WAR if he achieves the mean outcome; $10.3 mm/WAR if he achieves the median.     

If anything, I think my sample is biased in favor of how players will perform at 30-36, since it excludes anyone who had fallen off so badly that they weren’t still in MLB at age 36.

It’s interesting that four players on my list produced a lot more WAR after age 30 than before age 30: Cruz, Zobrist, Encarnacion and Phillips.    Kinsler also produced slightly more after age 30 than before.    
 

[now]

Just for reference (to your figure of

2 hours ago, Frobby said:

So, that’s an expectation of $9.4 mm/WAR if he achieves the mean outcome; $10.3 mm/WAR if he achieves the median.   

Nice research! Just for reference, what's the going rate of $/WAR?

[99ct]

3 hours ago, Frobby said:

If anything, I think my sample is biased in favor of how players will perform at 30-36, since it excludes anyone who had fallen off so badly that they weren’t still in MLB at age 36.
 

Unfortunately it seems there's some serious survivor bias here. Not only does this fail to account for players who left MLB between the ages of 30-36, but, as I understand it, the sample is limited to players who are still top 100 players at age 36+. How do you know that Rendon will be a top 100 player at age 36+? That's part of the question your seeking to answer, so it really can't be used as an assumption also. 

[DrungoHazewood]

3 hours ago, Frobby said:

I know more scientific analyses have been done on aging in baseball, but I got curious on the odds in today’s environment that Rendon’s $245 mm contract would pay off.    His contract runs through ages 30-36, so I looked at all active players 36 and older who are among the top 100 in position player WAR.    There are 13 such players: Pujols, Cabrera, Cano, Kinsler, Granderson, Zobrist, Molina, Ru. Martin, Cruz, Encarnacion, Choo, Pence, and Phillips.     For each player, I calculated the percentage of their WAR through age 36 that they had accumulated by age 29.    The average comes out to 51.2%; the median was 53.5%.

Applying those to Rendon’s case (27.3 WAR through age 29), the mean average outcome would be 26.0 WAR from ages 30-36.    The median outcome would be 23.7 WAR through age 36.   So, that’s an expectation of $9.4 mm/WAR if he achieves the mean outcome; $10.3 mm/WAR if he achieves the median.     

If anything, I think my sample is biased in favor of how players will perform at 30-36, since it excludes anyone who had fallen off so badly that they weren’t still in MLB at age 36.

It’s interesting that four players on my list produced a lot more WAR after age 30 than before age 30: Cruz, Zobrist, Encarnacion and Phillips.    Kinsler also produced slightly more after age 30 than before.    
 

The single most important thing to consider in any baseball aging study is survivorship or the drop out rate. 

A number of studies were done years ago that suggested that aging was of little or no consequence because, for example, the OPS of all 37-year-olds is usually as high or higher than all 27-year-olds.  That's because there are only, like, eight 37-year-olds left in the league, and they were almost all among the very best players in baseball at 27.  95% of all players are sitting at home at the age of 37.  You have to include them in your sample in some way, and to be fair you should probably take their last year in the league and subtract something like half a win for each year beyond that.  So a guy who retired after a 0.5 win season at 33... at 37 you count them as a -1.5 or -2 win player or worse.

[weams]

5 hours ago, Frobby said:

I know more scientific analyses have been done on aging in baseball, but I got curious on the odds in today’s environment that Rendon’s $245 mm contract would pay off.    His contract runs through ages 30-36, so I looked at all active players 36 and older who are among the top 100 in position player WAR.    There are 13 such players: Pujols, Cabrera, Cano, Kinsler, Granderson, Zobrist, Molina, Ru. Martin, Cruz, Encarnacion, Choo, Pence, and Phillips.     For each player, I calculated the percentage of their WAR through age 36 that they had accumulated by age 29.    The average comes out to 51.2%; the median was 53.5%.

Applying those to Rendon’s case (27.3 WAR through age 29), the mean average outcome would be 26.0 WAR from ages 30-36.    The median outcome would be 23.7 WAR through age 36.   So, that’s an expectation of $9.4 mm/WAR if he achieves the mean outcome; $10.3 mm/WAR if he achieves the median.     

If anything, I think my sample is biased in favor of how players will perform at 30-36, since it excludes anyone who had fallen off so badly that they weren’t still in MLB at age 36.

It’s interesting that four players on my list produced a lot more WAR after age 30 than before age 30: Cruz, Zobrist, Encarnacion and Phillips.    Kinsler also produced slightly more after age 30 than before.    
 

Well we know what happened with Cruz and we suspect the Parrot carried the goods for Arn Car Na Cion

[Frobby]

3 hours ago, now said:

Just for reference (to your figure of

Nice research! Just for reference, what's the going rate of $/WAR?

Per Fangraphs, still about $8 mm/WAR.    

[Frobby]

2 hours ago, 99ct said:

Unfortunately it seems there's some serious survivor bias here. Not only does this fail to account for players who left MLB between the ages of 30-36, but, as I understand it, the sample is limited to players who are still top 100 players at age 36+. How do you know that Rendon will be a top 100 player at age 36+? That's part of the question your seeking to answer, so it really can't be used as an assumption also. 

It’s cumulative WAR, not single season.  Rendon is 52 on the current active list, so he’ll certainly be in the top 100 at 36.    Today the top 100 cutoff is 17.5 for active players, Rendon is at 27.3.