never rat on your friends, and always keep your mouth shut

Monday, January 23, 2006

the stats yahoo lets you choose

im sticking with 5x5 until the ship sinks. using these avaliable stats, if you can create a reasonable alternative i can be persuaded.

lets just try and have each aspect counted once instead of the same thing being counted multiple times (BA and hits). i know there is a difference so don't just go on about how they are not the exact same thing.


carter said...

Based on what Yahoo! makes available, i want to use the following...

Statistical Categories:
Position Players:
Slugging Percentage
On-Base Percentage
Total Bases

Pitching Appearances;
Games Started or Total Innings;
Earned Run Average
WHIP Ratio
Strikeouts per Nine Innings

Why? First, I am very excited about the possibility of positional player defense, but the stats available are very limited. If people refuse to use steals/runs/etc, we need something to create value for middle infielders, catchers and other players who do not have an OPS of .900+ - or else the league will simply come down to whose sluggers have the best season.

Second, I am swayed by the arguments that fantasy stats should largely be team-independent. These stats do a relatively fair job of accomplishing that goal.

Finally, these stats create some interesting tradeoffs in terms of drafting...

pedrag said...

"im sticking with 5x5 until the ship sinks. (but) (...) i can be persuaded." do i misunderstand the rhetorical implications of ship-sinking or do you mean that you just want 5x5 setup regardless of the scoring stats? minor point, but i don't know how to interpret this from the original omarian. i too am strongly in favor of 5x5, provided we get the right stats. i like the numerical symmetry -- ten teams, ten categories, a perfect score is 100, etc.

i don't know how i feel about defense, especially if it eats up two categories. it seems almost impossible to have defense be team-indepedent, because so many of your opportunities for putouts are tied to teammates (what kind of pitcher is he, how good is he) and opponents (are they slow baserunners, are they the royals, etc). same for assists, really. assists is basically the save/stolen base of defense; it's cool, but would probably be frustrating.

also, the value of middle infielders/catchers is that everyone has to have one, and you don't want to get caught with an especially bad offensive MI/C. that's oversimplified, but worth considering. also, i'm a fan of not having the vague CI/MI slots, because that lets people load up. i don't know if yahoo lets a league strictly police position eligibility but i hope we can do that -- i'm sick of the craig wilson/brandon inge loophole, unless i'm getting its benefits, in which case, fuck yeah that wide.

I'm OK wit SLG, OBP, TB, Steals (i'm a centrist on steals, i admit it) and another counting stat, can't think of a good fifth stat.

Omar said...

i meant 5x5 in all senses. classic 5x5 roto stats and yes i like how we now have 10 teams, 10 categories, striving to get 100 points.

im not a fan of defensive stats either. nor am i fan of pitching apperances or games started. maybe total innings but thats actually already counted via the max innings rule. in theory (but not necessarily practice) everyone would end up maxed out in innings.

seeing as how im a yankee apologist, im not behind this team-independent stats movement.

the only changes i might make is change AVG to OBP to account for walks. Im not a fan of saves either but i can't think of a better stat to replace it.

pedrag said...

we don't actually have to set an innings max, but i think there's no inherent contradiction in having IP capped and IP also being a scoring stat. unless IP is capped too low. then there would be, as omar says, a 10-way tie for 5th place in IP.

carter said...

i do not disagree that defense is not exactly team-independent. but the point about the speed,etc of the opposition is silly, since *all* stats reflect the quality of the competition, especially given the sample sizes with which we are working.

for example, say that over the course of a season, two players both have the same "true talent" BA of 30% (i.e. against league average pitching, both would have an BA=.3).

Say league average pitching gives up an OBA=.260; further say that the top 5 teams have OBA of only .245 and the worst 5 teams have OBA of .275.

If player A gets half of his ABs against average and the other half against the best pitching, his BA drops to .283. Conversely, if player B gets half his ABs against league average and the other hald against the bottom 5, his average goes up to .317.

These are pretty obviously extreme examples - although not that extreme, since for instance St. Louis, Pittsburgh and Cincy all faced three of the top six teams in OBA with their division last year, and the unbalanced schedule meant that these team's hitters may have suffered considerably thereby.

The broader point is that stats can be more or less independant of any given player's own team - but not really of the opponent's team.

Thus, i reject this objection to defensive stats.

pedrag said...

Well, granted the premise that true team-independence can never be achieve unless we start using stats like "bat speed in inches per second per second" or "time to first base" (even then, i think there's a lot of context - weather conditions, wind) (kidding).

i admit even that it's an arbitrary distinction to ignore unbalanced schedule, difficulty of opponent, etc, in saying that OBP or hits depicts talent more accurately than say, runs, or RBI, or assists. I suppose i have nothing against fielding stats, but i don't think putouts and assists are particularly good ones. first basemen, with a few exceptions, are already sluggers (in re your contention that fielding stats would make MIs more worthwhile), but they record 3x as many putouts compared to any pos but catcher. a good fielding catcher and a bad fielding catcher are going to have about the same amount of putouts per game/season/career. this really is team dependent. For instance, the cubs catchers cought 1250some strikeouts last year, all of which count as putouts to the catcher. The mariners catchers only caught 800some. Is this is anyway a fair arbitration of the defensive talents of mike barrett vs. miguel olivo? Not at all. Same goes for assists. Why should yuniansky betancourt get credited with an assist for firing a throw way off target towards first base, which was only caught because richie sexson is nine feet tall? He shouldn’t but he does. Now, obviously, if y. betancourt keeps throw balls in the dirt, eventually he will start actually making some throwing errors, but the fact remains, he gets to be incrementally shittier than all other middle infielders and record the same stats, just because richie sexson is nine feet tall (throws to first even arrive faster than a throw to a non-huge first baseman because the distance is shorter).

I figure that fielding would probably be kind of fun to have as a stat, because it would take us six months just to figure out how it affected player’s draft value. but we all remember what happened with holds. I’m against fielding stats because A) it is a bit more team-affected than say, just hitting (by no means am I saying hitting or pitching are truly individual) and B) the fielding stats themselves are flawed, and don’t lend themselves very well to the kind of multidimensionality that translates into fantasy fun. If that makes sense – maybe another way to think about is this: if you play 140+ games at a certain position, you wind up with a certain number of putouts and assists, pretty much without regard to your talent (relative or absolute). The actual difference between 10 pts in the standings and 1 pts in the standings for putouts and assists will wind up hinging on things like, air pressure, richie sexson’s ape-arms, ‘how hungover jeff weaver is,’ etc. which is really no more or less arbitrary than having it depend on “how well ichiro can see” or well, “how hungover jeff weaver is.”

mikecard said...

wait, why would we want slugging and total bases? isn't that like the maligned system last year when we had hits and batting average?
also, as always, i feel like runs are getting a bad wrap here. i will attempt a defense of runs if i have to, but be warned: bill james thinks runs are an UNDERrated stat.

mikecard said...

rap not wrap

Morcy said...

My main objection to defense was that they count too low. Then I saw that the league leader in POs (fittingly, Poo-holes) had 1597, and As had 122. In other words, the sample is bigger than HR, and no one would ever say "HR is too rare to count"

That said, Pooholes had over 200 more POs than second place (Teixeira), so, um, there seems to be a big drop off. 15th place had just over half as many POs as first. That doesn't seem to be very fair.

The other reason I don't like defense is that BP doesn't project those stats in the excel spreadsheet they make available to subscribers.