Saturday, April 28, 2007

What it takes to win: $ Values Edition

In response to my post about dollar earnings based on how expensive players are, sas4 writes:

First, I recall reading somewhere that the stats of 80% of all players decline from year to year.

I'm not sure if this is true or not. But, in terms of player valuation, it is irrelevant.

The statistics of all players drafted in your league are worth however much you paid for them. If you're in a 12-team A.L. only, the players you drafted were worth $3120. In a 13-team N.L. only, those players are worth $3380.

In reality, it is the denominators in determining what the stats are worth that are variable, not what the league as a whole earns. In other words, your league bought 2,200 HR last year but only buys 2,000 HR this year, each HR hit this year will be worth more. Using real life as an example, Brian Roberts's 18 HR in 2005 were worth $6.10, but Carl Crawford's 18 HR in 2006 were "only" worth $5.80. There were (slightly) fewer HR drafted in the American League in 2005, so each HR hit in 2005 was worth slightly more than each HR hit in 2006. Three cents more, to be exact.

It would be interesting to see how much performance in terms of $ it takes to win the average stage 2 or 3 league. $280?, $260? $240? or how about more like $220 which is my guess. No one expects to made a profit on a stud. You want last years stats - thats the most you can reasonably hope for with injuries and all

In my American League, the winning team last year hit 211 HR, drove in 832 runs, stole 127 bases, and batted .274. Using last year's Patton formulas, this team earned $207 on offense.

The pitching staff won and saved 81 games, threw up an ERA of 3.92 and a WHIP of 1.281. Again, using the Patton formulas, this team earned $140 in pitching.

This team's total earnings were $347.

This sounds about right to me. $350 in a 12-team league conservatively is the sweet spot. Keep in mind that a $260 team is a .500 team, which in 12-team Roto probably equates to about a 52 point team.

You can make an argument that the number can be lower. This team won with a total of 74 points, finishing 3 points ahead of the next squad. The first thing you could do is trim every stat away from this team so that it still wins with 74 points. For example, this team finished 6th in HR with its 211. The next team had 206. Subtract 4 HR or $1.33 from the winning team.

You're still not going to get to a $280 team, though, so I'm not going to continue subtracting these marginal amounts from the winner in order to figure out this new total. Besides, every Roto squad tries to maximize its value; who knows in June how many HR will be enough to win in October?

I would imagine that a $280 winning team would be in a very competitive league indeed. This is probably a team that would win the league with 58-60 points and a league where all 12 teams were fighting for the title until the last day of the season. Typically, the distance between the first and last place teams is anywhere from $150-200, depending on how early the bottom feeders dumped and how much the top teams benefited from these dump trades.

$220 worth of performance would be a losing team, since $260 is the bar for mediocrity. Even a team dumping power would probably still earn about $270 at a minimum assuming that team was competitive.

2 comments:

Anonymous said...

In my American League, the winning team last year hit 211 HR, drove in 832 runs, stole 127 bases, and batted .274. Using last year's Patton formulas, this team earned $207 on offense.

The pitching staff won and saved 81 games, threw up an ERA of 3.92 and a WHIP of 1.281. Again, using the Patton formulas, this team earned $140 in pitching.

This team's total earnings were $347.

This sounds about right to me. $350 in a 12-team league conservatively is the sweet spot. Keep in mind that a $260 team is a .500 team, which in 12-team Roto probably equates to about a 52 point team.






You've left me with some things to think about.


Maybe its like this: You spend $260 to get $260 in stats. By virtue of the players in your reserves, it is impossible to get $260 in stats unless you have more overachievement than underachievement. Then add in injuries, having the wrong player active one week, and things start going astray.

I understand what you are saying about SBs or HRs changing each year depending upon how many are achieved, but I have a feeling there is more to it that that. But intuition tells me that we never get what we pay for.



Here is what I am curious about: if you added up all the values based on each team in the league, what would the Patton total be. In other words, was the total value of all teams earnings in your league 12 X 260 = 3120 using Patton values. Maybe this is not relevant. I have to think about this some more. Not trying to make work for you but I am curious.

Toz said...

Well, its good to know that I was way off in my estimation.

I think the initial assumption is off base. I try to think of it this way. The key to winning is maximization of value. Assuming a keeper league, my initial goal is to maximize value on the keeper list.

My next goal is to try and reach a $ for $ match between auction price paid and anticipated value. This never works on a player for player basis, of course, since there is an inflation factor to worry about in combination with overpaying for top tier stats. As a goal, however, the initial assumption should be that, coming out of the draft, you want a team in excess of $260 in stats.

Drafting $260 in stats is a recipe to finish in the bottom tier each year, regardless of overachievement versus underachievement. Moreover, at the bottom end of the draft, you hope to be getting, if you did it right, additional value to make up for any overspends at the top end of the draft.

Then there is dumping, another way to achieve additional value. As May progresses, I am sure Mike and I will be devoting time to the dumping culture, so I won't beat it to death here.

I guess what I am having difficulty understanding is the assumption that a team devalues during the course of the year. I do not believe this is necessarily true.