RYAN,

NOT TO BE DISAGREEABLE, OR TO SOUND LIKE A WALT APOLOGIST, BUT I WOULD ARGUE THAT YOUR ANALYSIS IS ONLY A PORTION OF HOW YOU SHOULD JUDGE GM'S...

1.) IN ORDER TO COMPARE WALT (OR ANY OTHER GM) TO A GM OUTSIDE THAT DIVISION, YOU'LL HAVE TO COME UP WITH A VERY DIFFERENT ANALYSIS.... YOU'RE CURRENTLY WEIGHTING A GM BY HIS DIVISION'S AGGREGATE BUDGET, WHICH IS DIFFERENT BY EACH DIVISION AND SO, FOR EXAMPLE, MARK SHAPIRO IS GOING TO LOOK STELLAR COMPARED TO THEO EPSTEIN BECAUSE THE CENTRAL SPENDS LESS THAN THE AL EAST!!!!

2.) GM'S WILL TEND TO SPEND THEIR BUDGET!!! EACH TEAM HAS A SPENDING CEILING, AND I WOULD ARGUE THAT IF BILLY BEANE COULD HAVE SPENT $80 MILLION, HE WOULD HAVE AND WITH A DIMINISHING AMOUNT OF RETURN.... TO TRULY MEASURE WALT'S EFFICIENCY WITH THE ALMIGHTY GREENBACK, YOU'D HAVE TO GIVE HIM THE SAME AMOUNT OF STARTING TALENT, YEARS, AND BUDGET AS THE GM YOU'D BE COMPARING HIM AGAINST!!! SO IT MAY BE SOMEWHAT VALID TO USE THIS METHODOLOGY TO MEASURE WALT AGAINST CHICAGO-N OR HOUSTON FOR THE PAST DECADE, BUT NOT REALLY AGAINST CINCY OR PITT....

4.) THERE'S NO ACCOUNTING FOR TALENT.... SOMEBODY (FORGIVE ME FOR NOT RESEARCHING AND LINKING) ONCE DID A STUDY OF WIN SHARES VS. ACCOMPLISHMENT, AND IIRC WALT WAS NOT STELLAR, BUT CERTAINLY ABOVE AVERAGE THERE!!!!

5.) ONE REASON OFTEN CITED IN PRO-WALT ARGUMENTS IS TALENT GIVEN IN TRADE FOR TALENT ACQUIRED!!! I DON'T KNOW "WIN SHARE" TOTALS OR ANY SUCH METRIC, BUT I WOULD ARGUE THAT THE TALENT HE'S GIVEN UP TO AQUIRE SUCH NAMES AS RENTERIA, MCGWIRE, ROLEN, AND EDMONDS HAS BEEN INCONSEQUENTIAL (DESPITE WHAT ROB DIBBLE THINKS ABOUT THE EDMOND TRADE BEING A GOOD ONE FOR ANA)!!!!

6.) YOU MIGHT ARGUE THAT ANOTHER MEASUREMENT OF A GM'S EFFECTIVENESS WOULD BE, GIVEN ENOUGH TENURE, TALENT PRODUCED!!! A KNOCK AGAINST THE WALONY LAJOCKETTY REGIME IS THAT THEY HAVEN'T PRODUCED ENOUGH YOUNG PITCHERS GIVEN WHAT THEY'VE HAD TO WORK WITH....

MY COUNTER, OF COURSE, IS THAT WHILE MIN DOES A GOOD JOB (OUT OF NECESSITY?) OF PRODUCING HOME GROWN TALENT, THAT'S NOT NECESSARILY A REQUIREMENT OF AN UPPER 1/3 MARKET TEAM!!!! FURTHERMORE, THERE ARE CERTAIN ORGANIZATIONAL STRENGTHS (DAVE DUNCAN, THE "BEST PLACE TO PLAY IN THE NATION", ETC) THAT A GM VERY WELL SHOULD PLAY TO, POSSIBLY AT THE EXPENSE OF DEVELOPING TALENT!!!

TOLAXOR, there's always room for disagreement, no need for apology!

I admit my model is crude, but it's about as fitting as many of the other stats I've researched. There are straight lines where I wish there were curves, but I can't always get what I want.

You make a lot of good points, but you quickly start going off into another area of interest and research. My study was done just to see how much bang for the buck the GMs are getting in this first 1/2 of the season, and by bang, I mean wins.

Essentially, I wanted to see the game as a financial function. If you take a $92 million club and plop them down in the NL Central, how well should they be doing now? The answer is they should be doing exactly as well as the Cardinals are doing.

If you took that $92 mill and put it into an ALC team, then the story changes, and it should. I think that's a fine way to compare GMs in different divisions -- you include their relative conditions in your analysis.

This is not a condemnation of Walt, in fact it's more a point that the Cardinals franchise is in financial balance, realizing a sort of point where spending a lot more money (if possible) wouldn't get us many more wins.

Mostly, I think my model disproportionately penalizes the BSOX and Yanks and rewards the poorest teams.... buuuut that's fine with me.

If you look at the GM Power Rankings, it's hard to disagree with the results. Most people are impressed with the White Sox and Nationals (and Cards) and recognize the Yankees as a giant disappointment.

Also, I don't mean to be too hard on Brian Cashman, I think he has by far the hardest GM position in MLB.

Ryan,

There's a good way to spend money (see Cardinals in the Walt Jockety era) vs. a bad way to spend money (see the Mets spending on Mo Vaughn). Maybe that's giving too much credit to the spending now that I can see the results, but I'd say there have been very fewer high $$$ moves that Walt and company have made that have not been worth the money.

Perhaps the above is hard to measure, but I'd say Walt has done a little better than just not hurting the Cardinals.

TSF

TSF, I certainly had the same impression as you do. Remember, this was a first glance attempt of mine to figure out how well teams are turning money into wins, not necessarily how well they evaluate talent, even though those are linked things.

It's 1/2 season of analysis, and it doesn't take into account things like injuries, etc.

You and I certainly remember and know of the great big signings Walt has accomplished. In regard to this team, Pujols, Rolen and Edmonds come to mind. However, I think he is certainly overpaying other big price tag players, Mulder and Walker come to mind. But again that's not what this study is looking at directly, I'm looking at a much higher level than that.

In terms of power rankings, Walt is evaluated as the 3rd most powerful GM in the game for this season to date. I don't think my conclusions hurt his status or reputation. He just doesn't lead the pack in efficiency, and he doesn't need to.

I may have been a little imprecise when I called him strictly "average".

To say he's done a little better than just not hurting the cards is a semantic argument, I guess. I'm saying he's spending $92 million dollars well and getting about the right number of wins out of it. I'd say congratulations are in order for that type of performance. Also, Walt's more efficient than any GM with a bigger payroll than his (managing a large market club is a different skill than a small market club).

I'll try to work more previous seasons into the model to see what happens. It should also be interesting to see what these numbers look like by the end of the season.

But I think people are just being thrown off by "Walt" and "average" appearing in the same sentence. It's like claiming the Cardinals are having an average 2005 so far by their own standards.

Ryan,

Great post. A question and a couple comments.

1) Could you explain the math for how you came up with GM Power Wins? The ((Team Wins - Div Avg Wins)/Div Avg Wins) - ((Team Pay - Div Avg Pay)/Div Avg Pay) equation comes out to a fraction; how does this factor into the GM Power Wins stat?

2) It's hard to use this as a measure for any GM that has been with their team for less than say... 5 years. Which isn't to say that another model could do it better; I don't think any formula can come up with an accurate measure for how a GM with short tenure rates against those with long tenure.

3) I'm just brainstorming and looking for discussion on this one... but shouldn't a measure of a GM's success account for injury? For example, should Paul DePodesta be penalized because Eric Gagne and J.D. Drew are on the DL? And I don't think the "Drew is fragile" argument applies here given the nature of his injury. So, perhaps the team's $$ lost to injury should be subtracted from their payroll.

But on the flipside, GMs should be accoutable for contracting oft-injured players... so Brian Sabean shouldn't get off the hook for building a team of senior citizens.

That is a very gray area, but I don't think it's OK to dismiss injury as "everyone GM knows the risk" or "it could happen to anyone".

Anyway, I really enjoyed your post and would love to hear your comments.

-Matt

Matt, thanks for your comments.

1) I computed two fractions, one the Division Power as described above, the other League Power as alluded to. They are fractions, but they can range past 1 and they can be negative. I scaled the range and then added these two power measures to actual wins to get Power Wins. The two Power measures are trying to capture the craftiness of a GM against his division and league and adding the actual wins reflects brute strength (payroll size), together they make overall GM Power.

2) Yeah, I think breaking down GMs in tenure groups would be interesting, as well as in large/small market groups. That's probably even a better comparison in some ways. I guess I think of the GM as the ultimate "buck stops here" position, so even those guys who've been there for less than 5 still shoulder the responsibility for their team. Most large, inherited GM situations are gone after a few years as long as the owners give the current guy some leeway.

3) I agree completely, that is to the effect those injuries affect wins. Some big-time players carry stigmas of injury with them, and sometimes they're warranted and sometimes not. Any GM needs to be considering some level of physical risk when investing, not just performance risk. This would even out and reflect better if i included more than this 80 games in it.

Keep tuning in and dropping your ideas and comments.

Forgive me but I still don't follow completely.

1) I'm going to guess on the Div and League average salaries and try to compute this for the Yanks. The average wins might be slightly off, too.

# = guess

YDW (Yankee Division Wins) = (14 - 16.8) / 16.8 = -1.67

YDP = (204 - 115#) / 115# = 0.774

So YDW - YDP = -2.44

YLW = (43 - 41#) / 41# = .049

YLP = (204 - 80#) / 80# = 1.55

So YLW - YLP = -1.50

I'm sure I'm missing something because I don't see how those numbers explain a 20 win difference between Power Wins and Actual Wins.

Could you humor me and go through how a team actually computes its GM Power Wins?

2) Answering #1 should answer this one, but I'm curious as to how you compute League wins. Does that mean you don't count Interleague play? Does it include play against your own division?

I think I have a couple other questions but they might be answered by these first two.

Thanks.

Matt, I think I see where you're going. Let me go through the Yankee example. First off, the Division distinction wasn't just games against teams in your divisions it was overall wins ranked against the overall wins of your division cohorts.

At the time, the Yankees had 39 wins, the league avg was 38.9, the ALE avg was 39.0, making the Yankees an average performing team by league or division standards. The ALE avg payroll is 96.2, league avg is 73.1, making the Yanks (208.3) a very above avg spender within their division and league wide.

Then do what you did above.

dwin = 39-39/39 = 0
dpay = 208.3-96.2/96.2 = 1.16
dwin-dpay = -1.16

This is the most extreme value of all of them. Most came out as small fractions, so I multiplied them all by 10 (poetic license to control the range) giving the Yankees a "Division Efficacy" of -11.6. Going through the same steps but multiplying the league results by 5 (to 1/2 weight the league measuring stick), they had a "League Efficacy" of -9.2. Add these two numbers to 39 wins and you get 18.2 GM Power Wins.

Like I said, I think this system unfairly penalizes the two biggest spenders, but I don't think it mischaracterizes league-wide results. Would anyone disagree that the Yankees are the least impressive team in baseball this year given the resources that go into them?

I'll publish these numbers after 162 go by. We'll see how it looks then.

Division Data Total
ALC Sum of Payroll ($Mil) 278.9 $46.5
Sum of Wins 202 33.7
Sum of Win Cost 139.6 23.3
ALE Sum of Payroll ($Mil) 481.1 $96.2
Sum of Wins 195 39.0
Sum of Win Cost 240.4 48.1
ALW Sum of Payroll ($Mil) 296.7 $74.2
Sum of Wins 161 40.3
Sum of Win Cost 149.6 37.4
NLC Sum of Payroll ($Mil) 395.9 $66.0
Sum of Wins 217 36.2
Sum of Win Cost 217.6 36.3
NLE Sum of Payroll ($Mil) 392.3 $78.5
Sum of Wins 212 42.4
Sum of Win Cost 188.2 37.6
NLW Sum of Payroll ($Mil) 347.1 $69.4
Sum of Wins 181 36.2
Sum of Win Cost 193.9 38.8

I apologize for the format, but above is the per division stats generated by my data. Thought it might be interesting.

Thanks dude. I think it would be interesting to include $$ lost to DL in the formula as well (Payroll = Payroll - $$ lost to DL). There's some more I want to discuss on this, but I think I'll save it for the full-season analysis.

ryan - i made a graph that charted my confusion, but oddly enough, i don't think anyone on this site could understand it. does that make me a bad person??

your post was better than raisin bran's.