Calculating $/WAR in the WBA

[Huckleberry]

Previously in the chat box this had been briefly discussed and I posted my rule of thumb that each WAR is worth about $1.4M in the WBA. This was based on some very quick and very simple math as follows:

There are 1000 WAR available every year IRL across all of Major League Baseball. It is essentially the definition of WAR that a team full of replacement level players will play approximately 0.294 baseball over the course of a season. If you scale that to the WBA and even further to each individual league, the ABL and the IBL, then you will see that there are a hair under 321 WAR per season in each WBA league. For some reason OOTP doesn't adhere to this standard as the WAR totals for the ABL have fluctuated between 340-360 WAR. That is somewhat irrelevant, though, for a rule of thumb (and for those of us who like spreadsheets and calculate our own WAR values using compiled stats when we want to see them).

Anyway, based on the calculated league total WAR of about 321, we then look at salary totals. Because I intended this to be a rule of thumb across the life of the WBA, I used $50M as the average team payroll in my calculation. I did this because WBA financials are set up so that teams average that much in revenue so if everyone broke even every year then that would be the average payroll. In reality, for example, the ABL teams did not spend $500M in payroll expenses in 2106, they actually spent around $450M. So I take the $50M team salary and then subtract $5M which represents the replacement player value that is built into team payroll via the league minimum salary. A team full of replacement players will have a team payroll of $200K times 25 players which comes out to $5M. So the napkin math says there is a total of $450M in discretionary player salary each year in each league which teams use to pay for WAR. $450M divided by 321 WAR comes out to roughly $1.4M per WAR.

So that is how I arrived at my rule of thumb of $1.4M being the dollar value of a WAR in the WBA. But recently I realized that due to our very rebuilding and team friendly rules, in particular the 5 years required before a player hits free agency, many WAR are tied up in younger stars who are still playing under league minimum contracts for up to 5 years. That means that while $1.4M/WAR is absolutely a valid rule of thumb for the league overall, it actually is not a valid rule of thumb for the free agent market. My next study is going to be to try to determine how many WAR are typically tied up in league minimum contracts each year, which will then determine the actual number of WAR that is purchased with the $450M in discretionary salary each year. This of course means that the final calculation will mean that each WAR is actually more expensive than $1.4M on the open free agent market.

[APMP]

Big fan of this. My guesstimate is that near half WAR is tied up in minimum contracts. So I've basically operated on the assumption that 2.5-3 million is a WAR in free agency.

[Huckleberry]

Well the first problem is that I don't believe OOTP keeps track of a player's salary history anywhere I can get to that data in bulk easily. And despite what I'd like, I simply don't have the time nor the patience to click through every single player's history screen in game in order to write it down. So I took a shortcut that will introduce some level of inaccuracy to the calculations. I used major league service time as a proxy for figuring out who had a minimum salary. There are cases where a player has been released during his minimum contract, although it's rare. Rare enough that I'm not going to worry about the fact that these players made more than the league minimum during their next contract, especially because again this is for a rule of thumb calculation and we should expect that these players would normally be playing on a $200K salary.

Now the first decision I made after that was that I was only going to use the 2105 and 2106 seasons. The early years of the league introduced a different wrench into the calculations as OOTP randomly assigned MLB service time at the outset of the league. But once we got to the 2105 season then there had been 5 previous WBA years so the service time shortcut made more sense. These calculations may change in another 7 years once pretty much all players have real service time from the actual WBA seasons, but this is good enough for now.

So the next thing I did was scale the WAR totals that OOTP calculated so that each league had the correct 321 WAR in each season out of the players I was looking at. For those that are curious at this point how I got to this data it was by utilizing the commissioner ability to export CSV files of historical stat and service time information. Typically when running my own personal analyses I use the HTML available historical stat registers:

http://www.worldbaseballassociation.com/reports/news/html/history/sl_batters_100_0_2105.html

which include a WAR column but because I am publishing this to everyone I went straight to the time-saving CSV generation.

So the bottom line is as follows:

2105 ABL -

Total Batting WAR - 194.3
Batting WAR by MinSal Players - 101.5
Total Pitching WAR - 126.7
Pitching WAR by MinSal Players - 58.9
Total WAR - 321.0
Total WAR by MinSal Players - 160.5

2106 ABL -

Total Batting WAR - 188.1
Batting WAR by MinSal Players - 125.4
Total Pitching WAR - 132.9
Pitching WAR by MinSal Players - 65.1
Total WAR - 321.0
Total WAR by MinSal Players - 190.5

2105 IBL -

Total Batting WAR - 194.2
Batting WAR by MinSal Players - 133.1
Total Pitching WAR - 131.8
Pitching WAR by MinSal Players - 37.3
Total WAR - 321.0
Total WAR by MinSal Players - 170.4

2106 IBL -

Total Batting WAR - 189.2
Batting WAR by MinSal Players - 126.7
Total Pitching WAR - 131.8
Pitching WAR by MinSal Players - 50.8
Total WAR - 321.0
Total WAR by MinSal Players - 177.5

There's some significant volatility in there so the small sample size of only two seasons makes it tough, but overall roughly 55% of WAR in the WBA is produced by players who should be on their initial minimum salary contracts. It appears, somewhat logically, that a higher percentage of bWAR is by young players than pWAR. As I said, there are some major fluctuations but at this point, neither league has ever had more than 50% of pWAR come from minimum salary eligible players while neither has ever had less than 50% of bWAR come from youngsters. The averages are 41% of pWAR and 64% of bWAR from young players.

Considering the discrepancy between bWAR and pWAR in this analysis I'm going to add a step to the study and try to determine the ratio of dollars spent in the WBA on pitching versus position players. That will help us see what the true cost of a WAR from a position player is versus a WAR from a pitcher. My initial thoughts are actually that this will yield the opposite result than what I expected - my instinct was that pWAR were more expensive on the WBA market. But if we have fewer discretionary bWAR available due to young players I'm not sure the salary information will make up for that. We'll see.

[Huckleberry]

So this latest plan is very seriously affected by the lack of player salary history kept by OOTP. Fortunately the almanacs I generate every year on January 1st include the team player salary pages like this:

http://www.worldbaseballassociation.com/reports/news/html/teams/team_6_player_salary_report.html

That's the good news. The bad news is those obviously don't include any players who sign between January 1st and when the season starts. The problem then is having to go through each one of them individually and manually entering the data. So I got lazy again and used the player contract info export, which (once again) has its shortcomings. It only has current contract information, but since I'm just looking at ratios of pitcher to position player salaries I'll go ahead and make the huge assumption that the ratios have been fairly consistent. I also eliminated all $200K contracts since those are year-to-year and therefore I couldn't even back out into old contracts. Here are the breakdowns in salary totals for all years in the current salary info:

2103 - 61% position players, 39% pitchers
2104 - 55% position players, 45% pitchers
2105 - 55% position players, 45% pitchers
2106 - 50% position players, 40% pitchers
2107 - 48% position players, 52% pitchers

Even though we can't trust the older info very well, the trend is clear in that teams have been spending more money on pitching than they used to. Based on the last three seasons there and again because this is all approximation for a rule of thumb, I will go with 50/50 leaguewide spending on pitching versus position players.

Also based on the WAR information above we will approximate 191.5 bWAR and 129.5 pWAR per league per year. Using the data above that roughly 59% of pWAR and 36% of bWAR is available from non-minimum salary players each year, we have 68.9 bWAR and 76.4 pWAR available on the market. Using the rough 50/50 spending split, the final results are:

$3.25M per bWAR
$2.95M per pWAR

I'm a little surprised, as I said, that bWAR is more expensive but this was a whole bunch of approximating anyway.  The short version is that I spend too much time just to arrive right where APMP's gut was - an open market WAR is roughly $3M in the WBA.

[squaredrive]

very insightful - thanks Huck

[Echo127]

Yeah, I think I've historically been shooting too low with my contract negotiations. Trying to be bolder about it, now. 

[APMP]

The short version is that I spend too much time just to arrive right where APMP's gut was - an open market WAR is roughly $3M in the WBA.

Haha yes.

Now if only some free agent stars would sign with Mumbai for fair market value.

[Huckleberry]

By the way, this means that Carlos Vargas' current market value would be roughly $40M per year. Hahahahahaha.

There's obviously a limit to the linear relationship once you get so big that you're affecting available payroll for the rest of the roster like that.

[Karachi_GM]

It would also say that Hashimoto's salary should be $26 million.  As it is, he is the highest paid player in the IBL at $17 million.