Question about the totoal values

gehrenberg   6/17/2014 1:13 PM

It seems the total z scores for the players at the top seem to be inflated. For example Miggy is expected to be a monster with t total value of 2.46, but that seems to be an impossible number to reach. His total score from last year was 2.07 and his numbers from last year seem to be pretty far ahead of what he is projected to do for the rest of this season. As good as he has been this year he has a total score of .99 which would mean he would have to be about 2.5 times more effective to finish the year to reach his projected value. So my question is what am I missing out on and do you guys really believe his total value for the rest of the year will get to 2.46?

descender  6/17/2014 1:18 PM

Depends on what page you are looking at. If it is his "remaining projection" then he is expected to outpace everyone from this point forward by that value. It has nothing to do with where he is currently. His actual value should be moving towards his projected value at all times, but will never likely reach it.

gehrenberg   6/17/2014 1:21 PM

Yes which is exactly my point, which is why does it appear that the players at the top of the projections have total z scores that will likely never reach. 

mbuser   6/17/2014 1:31 PM

I took a look at his projection and scaled back his HR expectations a bit, but in general we're projecting him to essentially win the triple crown the rest of the way while staying completely healthy, and that's why he's got the astronomical value relative to the player pool. The question comes down to which individual parts of his projection you disagree with.

descender  6/17/2014 1:33 PM

Because... math!

He is being projected to achieve that 2.46 Z-score only in the games between today and the end of the season (ROS).  What he did yesterday shouldn't influence that.  The projections are what they are, and his 2.46 value means his ROS projection is way, way better than everyone else, even though that he isn't playing up to that level currently.

An "end of season" projected total value wouldn't really be useful.  You should only be concerned with how a player will play in the games from today until the end of the season vs the rest of the league.  You can compare that to what he has actually done, but you can't combine it in any useful manner.

gehrenberg   6/17/2014 3:18 PM

I think his actual numbers projections are fine, I was just wondering why it appears that some players have inflated z scores. It appears that even if they exceed their projected numbers the z score won't add up. Perhaps what I am saying isn't making sense because nobody appears to understand the point I am trying to make. My issue is not with any players individual number projections, but rather with how the z scores are calculated. Although this may just be due to more players at the bottom of the league performing to higher expectations. I fully expect Miggy to play to the hard numbers he was projected to. However, I am just asking if he hits .331 with 21 home runs will that actually end up working out to a total z score of 2.46. It seems to me just looking at the numbers in general he would from here on out (not counting what he has already done) he would finish with a z score in the 1.9 range even he hits those number projections. Does this make sense to anyone? If not my comments can just be ignored. 

mbuser   6/17/2014 3:24 PM

Sorry to gloss over that aspect of your original post. The differences in actuals vs projections will absolutely influence how well someone like Miggy scores for the remainder of the season. In this case, it would speak to the player pool as a whole performing a bit better than projected, which would depress the Rankings value of the top players.

gehrenberg   6/17/2014 3:59 PM

Okay great thanks this was exactly the answer I was looking for. Now this all makes complete sense to me. It seems a lot of the value in the top players is their consistency. And I also think this shows the value of working the waiver wire in roto leagues. A lot of crappy players (who should be projected to be stink) will outperform their value because of individual favorable matchups while sucking most other days. But those days where they really exceed expectations probably does a lot to bring players z scores closer to league average. Also one more question about z score, what number is considered an average player? I would think it be 0 but there are so many players below that line and the mean and median also seems to be below zero. Thanks for answering Matt this thread has been interesting for me. 

mbuser   6/17/2014 4:14 PM

Yes, a z-score of 0.0 would mean it's the average number for the stat based on the player pool being used for calculations.