September 11, 2005

I can't make heads or tails of this chart. I hope I'm not the only one.

posted by mr_crash_davis at 10:41 PM on September 11, 2005

It's a poor representation of money vs results. The idea is that the top salaried team should have the top ranked team, and therefore their line should have no slope. If your team is doing worse than their salary ranking, then the line slopes up (left-to-right). If your team is doing better in the standings than their money ranking should indicate, then it should slope down (left-to-right). However, the difference between the highest (NY = ~$208million) and the lowest (TB = ~$30million) is so big that a simple "ranking" doesn't really describe it. Another option, $/win, is flawed because you'd end up having the best team be...the Tampa Bay Devil Rays, who would only be spending $503,035.03 per win (while the Yankees would be up there at over $2.6million per win). There is no way you can say that Tampa Bay is doing well, when they are 25 games below .500. One way that I've heard before is to calculate $/win over replacement level team. The minimum salary you can possibly field in MLB is $7,900,000 (based on 25 players on your roster all earning the MLB minimum of $316,000). And if you assume that a team with that barebones of a payroll would be the worse than the worst team in modern baseball history (40 wins by the 1962 NY Mets, so 35 wins sounds about right), then they should have 31 wins by now (based on playing 143 games). The ranking of $ spent (over the minimum) / wins (over the minimum) would be: $spent = actual payroll $min = $7,900,000 (replacement level salary) W = actual wins at this time in the season Wmin = 31 wins (expected replacement level win total at this time in the season)

Team 2005 salary W     W-Wmin $spent-$min    ($spent-$min)/(W-Wmin) CLE 41,502,500 81 50 33,602,500   672,050 TB 29,679,067 59 28 21,779,067   777,824 MIL 39,934,833 70 39 32,034,833   821,406 TOR 45,719,500 74 43 37,819,500   879,523 WAS 48,581,500 73 42 40,681,500   968,607 OAK 55,425,762 79 48 47,525,762   990,120 MIN 56,186,000 73 42 48,286,000 1,149,667 PIT 38,133,000 57 26 30,233,000 1,162,808 FLO 60,408,834 76 45 52,508,834 1,166,863 CWS 75,178,000 87 56 67,278,000 1,201,393 TEX 55,849,000 69 38 47,949,000 1,261,816 SD 63,290,833 71 40 55,390,833 1,384,771 STL 92,106,833 91 60 84,206,833 1,403,447 HOU 76,779,000 76 45 68,879,000 1,530,644 ATL 86,457,302 82 51 78,557,302 1,540,339 CIN 61,892,583 65 34 53,992,583 1,588,017 ARI 62,329,166 65 34 54,429,166 1,600,858 COL 48,155,000 56 25 40,255,000 1,610,200 ANA 97,725,322 80 49 89,825,322 1,833,170 DET 69,092,000 64 33 61,192,000 1,854,303 KC 35,881,000 46 15 27,981,000 1,865,400 BAL 73,914,333 66 35 66,014,333 1,886,124 CHC 87,032,933 70 39 79,132,933 2,029,050 PHI 95,522,000 74 43 87,622,000 2,037,721 BOS 123,505,125 83 52     115,605,125 2,223,175 LA 83,039,000 64 33 75,139,000 2,276,939 NYM 101,305,821 70 39 93,405,821 2,395,021 SF 90,199,500 64 33 82,299,500 2,493,924 SEA 87,754,334 61 30 79,854,334 2,661,811 NYY 208,306,817 79 48     200,406,817 4,175,142

Note: I didn't do an exact calculation of wins over replacement-level for each team. I just took their existing wins and subtracted 31 wins, which is the number a "replacement level" team would have after 143 games (which is about average # of games played so far for all the teams). So it looks like the Cleveland Indians are the best "bang for buck" team in MLB right now. However, to properly rate a GM, there should be a further breakdown of $ spent per player and how much each player represents to a team above (or, in some cases, below) a replacement level player. For example, a GM should get positive marks for spending the league minimum on a player that is contributing WAY above average (see: Felix Hernandez), but should be VERY penalized for spending big bucks on a terrible player (see: Sammy Sosa) (taking into account such factors as injury time (compared to previous injury time lost in career), replacement level for each position, etc). Accumulate the differences on each team, and you'd have a good idea of how they spend their money. Then use that as a reference for how they are doing as a whole, and that will give you the best performance by a GM. Or so I think.

posted by grum@work at 12:27 AM on September 12, 2005

Nice work Grum. I think these payroll numbers might be the 40 man roster and also the disabled list may skew your minimum wage calc. I agree that the salaries need a common benchmark to be measured but how about using the teams Gross Revenues (if they could be obtained). Salaries as a percentage of the teams gross revenues may work as a benchmark. It would be an eye opener if fans could compare the percentage of sales to salaries expense. Maybe the Yankees take in 5 times the revenue Cleveland does. If true then their win cost would be comparable.

posted by sandman at 04:36 AM on September 12, 2005

Not only is it a poor representation of performance and cost, but it's a poor graphical presentation as well. Grum's table tells you a lot more AND does it a lot more effectively.

posted by Amateur at 07:28 AM on September 12, 2005

Jesus Grum, that's fantastic. Only because it, of course, reinforces those things which I already believe - that by 2007 the World Series will be fought between the Jays and the Brewers. I'm convinced the Indians will find some way to fuck it up and not continue to get better. You can't just have your team represented all willy-nilly by Corben Bernsen in a series of films and not pay for it karmically for eons afterwards.

posted by WeedyMcSmokey at 08:32 AM on September 12, 2005

Good grief, grum. You sure you don't have a career doing this? Nice job!

posted by worldcup2002 at 09:35 AM on September 12, 2005

Nice work grum, I think sandman's right when he says revenues should be factored into that somewhere. Maybe percentage of revenue spent on salary. Weedy, Roger Dorn was a great shortstop in his prime.

posted by erkno11 at 10:31 AM on September 12, 2005

Oh damnit! The NYY and BOS figures are a bit screwy on the table. They looked fine to me, but that was after midnight so I can't trust my eyes. The idea of comparing to replacement players is something that a bunch of (much better) people have been using for a while (see BP and Hardball Times for examples). I'm not even sure if my calculations are correct (40 man roster might be better, replacement level wins might be off) but it was something I wanted to give a try. Interestingly, I did it because I hated the idea that TB would be represented as a "model" franchise. But even after my calculations, I have to admit that they are doing a lot with a little. The big test (for them, Toronto & Milwaukee) is to spend the little bit of extra money to get the extra wins to make it to the playoffs. That means not blowing a ton of dough on a mid-level player, or sinking the team with long-term contracts for older players. As a Toronto fan, I know that the Jays kept things very competitive this year, but were obviously missing a key part or two. If they spend another $15million/season, they might be able to get those key parts via free agency and make a run as BOS and NYY start to choke on their bloated rosters. Tampa Bay is another matter. They have an amazingly good young core of players in the majors and the minors. What they have to do is bring them into the key roles at the right time; too early and they'll struggle and be too expensive too soon (arbitration and free agency); too late and they'll be wasting away in the minors behind some lump of clay in the majors. Finally, if I had the time I'd like to do the same thing for the past 5 seasons and see if there is a trend among certain teams. I agree that the salaries need a common benchmark to be measured but how about using the teams Gross Revenues (if they could be obtained). Salaries as a percentage of the teams gross revenues may work as a benchmark. It would be an eye opener if fans could compare the percentage of sales to salaries expense. Maybe the Yankees take in 5 times the revenue Cleveland does. If true then their win cost would be comparable. Yes, that would take the calculations/comparisons to another level. It would then be ($spent/$available)/win, which would give a much better picture of the team's resources & ability to produce wins. However, the GM rarely has a say in how much a team could spend (that's the owner's job), so it would be more of a reflection of the team as a whole instead of the GM. somehow reading a comment that gets into baseball stat crunching I knew grum's name would be at the bottom. I like to wallow in baseball stats every so often, like a pig in mud. It's much more fun for me than discussing steroids, suspensions and character issues.

posted by grum@work at 11:29 AM on September 12, 2005

The Jays have as much stated that their extra money is going into Burnett's wallet (if they can get him). They've spent about $55 million of the $210 million they've been allocated for the next three years (including this one).

posted by WeedyMcSmokey at 12:56 PM on September 12, 2005

Grum, Follow the the line and you could see that. Huh? Follow what line? In the graph from the original post? As someone said, the criss-crossing lines and the fact that they don't line up exactly with the team/money values makes it difficult to spot the best bang-for-buck team. My post (while long) was just another way of examining the information provided.

posted by grum@work at 02:46 PM on September 12, 2005