Looking at the projected standings you hear the same complaints year after year. "This doesn't look like any standings I'm used to!" "These teams are all bunched together!"
While you might be tempted to say these are stupid people that don't understand how these sites are getting these numbers, really I place a good deal of blame on the guys making the standings themselves. When you attach a definitive win total to a team and put all 30 in standings form, what do you expect fans to do? Of course they are going to compare them to the usual standings and hey, these do look kind of weird in comparison.
The improvements that should be made to avoid this are simple ones. These win totals shouldn't be numbers, they should be ranges. 95% confidence is one standard but if that's too broad cut it down to 75% or whatever feels right. And they shouldn't be placed in standings form. Listing them by division is necessary, you want to compare your team versus its direct competitors, but ranking them would probably be the best. No, that isn't too different than standings but visually this :
Ranking the NL teams
1. WAS 85-89 wins
2. ATL 83-88 wins
3. NYM 73-81 wins
4. MIA 70-79 wins
5. PHI 66-81 wins
is different enough from this :
NL EAST
WAS 87 75
ATL 85 77
NYM 78 84
MIA 75 87
PHI 72 90
to get what I assume your true point is across. I hope you are saying "Here are the teams in the NL East and how I think they compare to eachother" as opposed to "Here are how the teams in the NL East will probably finish next year!" beacuse the former is all you really can say. The latter is mathematical snake oil.
Also, I wish rather than react with the knee-jerk "you just don't get it", there was more "we just don't get it". By that I mean admitting to the reader right up front that there are a bunch of season altering events that we just can't predict. Injuries are huge every season. There are breakout years for a couple young players, and years from veterans where everything seemingly goes wrong. In season deals can change the make-up of a team. Then there is luck. A few more one-run losses or wins than you probably should have and the whole year looks different. All this ranges from just barely predictable (you can somewhat account for injury prone players) to not predictable at all and the wrong injury coupled with bad luck and an offensive collapse from a key player and suddenly your playoff team looks like an also ran. I know people that do projections do admit this, but it's often in small type or after the fact. It should be first thing. We really don't know. Here's our best guess.
End scene.
Nothing really is going on Nats-wise as everything took a break for the Super Bowl (including the Broncos! Hey Yo!) so I suppose it's time to open up the mailbag. Oh wait. I have no emails because I didn't ask for any questions. Goes to show me.
So here you go : gmail account, natsoftheroundtable
Ask away and sometime later in this week. I'll get back to it.
Subscribe to:
Post Comments (Atom)
8 comments:
First of all, the ZiPS projections for the Nats are finally up and ZiPS seems to love the Nats. They've got Harper with a power boost and Strasburg bumping his K rate up half a point and his BB rate down half a point over 9 innings while retaining his low HR/9 from last season.
ZiPS has higher IP estimates for most of the starting pitchers, but lower for Strasburg, so those are probably pretty close.
Addressing your point in this post, Harper, I think the problem is that cautious and in depth doesn't sell. I remember the cover on SI last year stating that the Nats would definitely win the World Series and Strasburg would be Mr. October. A headline of "We can't know who'll make the World Series. We can't even know who'll make the postseason. Here are the teams we think will be good, why, and what we expect from them given that there will be huge error bars." probably would not lead to magazines flying off the shelf.
In certain places it'd be welcome, but probably not the big market media.
Oops, forgot to link:
http://www.fangraphs.com/blogs/2014-zips-projections-washington-nationals/
Yeah the preseason zips projections I think are best used as something to point you in the right direction.I generally don't like them, more so with players, because like the shredder on MLB Network (which i love the top ten right now series btw how about you?), they don't take certain things into account that I think should be.
I like projections for getting human bias out of the way. And because fans of the team tend to be overly-optimistic. So it's easy for me to think Healthy Harper = April 2013 Harper All Season Long, but that's not realistic.
I think we also expect young players to improve more than they generally do, though there will always be exceptions like Ian Desmond who suddenly figure everything out.
I totally agree with your point that projections don't predict breakout seasons or major drop offs. What's curious about the Nats is that I think they have more volatility than the Braves. I don't think anyone on their team has the upside of Harper who could turn in a Trout like year. But I don't expect anyone on their team to crumble the way Espinosa or Laroche or Storen did last year. This season could really go any way, with it being close all season to either team running away with it. That's why they have to actually play the games.
@Cass, I agree that we typically expect strides that are far too large for young players, I think Harper's case is different. Historical data about 19 year olds in the big leagues show that nearly all of them made HUGE strides in either their age 20 or age 21 seasons. Harper looked poised to make those huge strides in April of last year, but couldn't stay healthy. So I think it's a bit more reasonable to expect big things from him this year, although, somewhat obviously, you're looking at very small sample size here limited to a few players throughout history who got significant AB's at age 19.
I mean, look at AROD, he got a cup of coffee at 19, a few more AB's at 20, then put up a 9.2 fWAR at age 21. Griffey, put up 2.5 fWAR at 19, 5.0 at 20 then 6.9 at 21. I think Harper would be a lot more impressive if not for Mike Trout and all of the exposure he got in high school.
ZiPS has Harper pegged for 5 WAR season, but I see a 7 or 8 WAR season if he's healthy. Not just because I'm a Nats fan, but because the limited historical data of guys who played at 19 seems to suggest it. Also, because it's what my eyeballs tell me and with young players the eyeballs are as important as the numbers if not more so.
Donald: But isn't the point of a collapse that it's sorta tough to predict? There's a diff between a player collapsing because he's old (LaRoche) and a player collapsing because of injury (Harper) or headcaseness (Storen) or God knows what combination of the two (Espi). I'd think the Braves are equally vulnerable to the latter two. I mean, did anyone else BJ Upton or Heyward to have the 2013s they did?
@Chinatown
But that's the thing-- two of their core players put up sub-replacement level numbers and they still won 96 games and the division. Obviously the injury bug can hit anyone, but what realistic scenarios have 2+ of their players having complete collapses in 2014, since we saw in 2013 that their win total can withstand that. I feel much better about pegging the Braves to 91 wins than I do the Nats, since I think we have a considerably wider range of outcomes (somewhere between 86 and 98 wins, presumably).
Post a Comment